Rules design: skill learning speed *span*?
Peter Knutsen
For some odd reason, I've dug up my old sketchy "RPG Engine", FFRE,
and resumed working on it. In the process, I've abandoned the old
skill learning speed formula and am searching for a new one, so I
thought I'd ask in here what people find is a reasoble span?
Some systems, such as FUDGE, refuses to link attributes and skills.
Thus a character with a low Dexterity can learn a combat skill as
fast and easily as a character with high Dexterity. Rather than
criticizing this I'll just point out that it represents one extreme
attitude towards how attributes makes people different from each
other (in FUDGEs case the attitude is that attributes makes no
difference - everybody learns at the same speed, so the span is
one).
A more moderate system, Quest FRP, turns out to have the learning
span be something like a factor of 3. A character with extremely
low relevant attributes (say, Strength and Dexterity of 6) needs
three times the training, to reach a given level of skill, as a
character with very high relevant attributes (Strength 18 and
Dexterity 18). Inept Ian needs 12 weeks to go from no skill to
competent skill, while Talented Tom needs only 4 weeks to go from
no skill to competent.
But please keep in mind that attributes of 6 are *extremely* rare
(the human average is 12). The real span is more like a factor of 2.
Or two-and-a-half since attributes over 18 are possible.
A system representing the other extreme from FUDGE is GURPS.
In GURPS, you cannot directly measure how long it takes for
a character to learn a given skill, but you can still gain a
vague impression. Which turns out to be hugely surprising.
A character with an attribute of 11 needs more than ten times
the amount of training, to reach a given skill level, as a
character with an attribute of 15. He learns at least ten times
slower. If you compare a character with Dexterity 9 (slightly
below average) with a character who has Dexterity 16 (which is
very good but not in the cinematic or mythical range) then the
difference is even larger. I'd guess a factor of one *hundred*.
Okay, I object to FUDGE's denial of the fact that talented people
learn faster. But I also object to the extreme learning speed
of talented people in GURPS. A span of 30 or 100 or maybe even
250 is absurd. It does not pass my reality check.
You all know that I'm fond of Quest FRP, but in the area of skill
learning I'd have preferred a slightly larger span. 2.5 or 3 isn't
quite enough, keeping in mind that the span goes from the most
inept and clumsy/stupid person to the most talented and dextrous/
intelligent person.
I would prefer a span of 6 or 10. But the reason I'm posting
in here is because I'd like some feedback. What kind of span do
you guys and girls think is suitable? What would pass your
reality check?
FFRE has an attribute scale going, for humans, from 1 to 8, with
3 being average and 1 being impaired. 0 is sometimes possible,
and for non-humans sub-zero is also possible (Hobbits will often
have a negative Size), but those attributes which can have 0 or
negative as a value are rarely those who governs skill learning.
Non-humans can have attributes exceeding 8 but very rarely
exceeding 10. I have a writeup of Morgoth who gives him an
Intelligence of 8. Saruman would have a Charisma of 8, Smaug
a Perception of 8. Thus the superhuman range is *very* rarely
utilized and you can focus on the human range.
The formula I use is (Aptitude*Aptitude)/9. Aptitude is the
average of six attributes (always with some doubles and triples,
for instance the Aptitude for Meelee Weapons is ((3 x Dexterity)+
(2 x Strength)+(1 x Agility))/6. The learning speed formula
yields the following results
Apt. Learn Speed
1 0.1
2 0.4
3 1
4 1.8
5 2.8
6 4
7 5.4
8 7.1
9 9
10 11
If we consider the human range, this gives a span of around 70.
It may seem as if FFRE is as extreme as GURPS, but in fact an
FFRE attribute of 1 translates to a GURPS attribute of 5 or 6,
and an FFRE attribute of 8 translates to a GURPS attribute of
20 if not 21. For GURPS, the span 6->21 means a learning speed
difference of several thousands, if not tens of thousands.
We can also try to narrow down the range a bit, excluding
attributes below 2 and above 6, because chances are you don't
know anybody who has an attribute at 1 or anybody who has an
attribute above 6. Now the span is only 15.
This sounds good to me, but I'd appreciate some input. Does it
sound good to you?
--
Peter Knutsen
Brandon Blackmoor
any realistic sense is dreadfully dull to play through, and thus, like all
things dull, it shouldn't be played through.
However, I am not entirely certain I understand what it is that your
"learning span" is measuring. Is it the difference in the amount of time for
a normal human and a talented human to both learn the same new skill to the
same degree of proficiency (in which case, in most game systems, the normal
human marksman is actually much more *skilled* with her weapon, even if the
end result is that she is on par with a more naturally gifted person who has
less formal training)?
Or is it the difference in the amount of time it takes for a normal human
and a talented human to both train to equivalent levels of competency if
their natural talents were *removed* from the picture (i.e., if the normal
human golfer were given a handicap to compensate strictly for having merely
ordinary hand-eye coordination, or if the more-than-human golfer were
wearing gear from the Harrison Bergeron collection).
bblac...@blackgate.net
15.february.2001
Warren J. Dew
those least competent to those most competent.
In terms of what's reasonable for a game, my own system has about a factor of 9
difference in speeds from worst possible to best possible in the human range,
so I guess I consider that reasonable. I actually think that less might be
better from a game balance standpoint.
From the standpoint of the player world, I've seen ballroom dancers get better
after a month than others are after ten years, with the people involved
spending comparable amounts of time per week, so that's a factor of 120 or so
difference.
I don't think that's the best representation of what actually happens, though.
Based on my experience teaching ballroom, I think people increase in skill at a
common rate until they reach some maximum imposed by their talent, interest, or
instructor. A less than fully competent instructor can slow their learning,
but the effect of attributes and interest seems to be to limit how far they can
get, not how fast they get there.
I'd also note that all three of talent, interest, and instructor can change.
Brandon Blackmoor:
Training in any realistic sense is dreadfully dull to play
through, and thus, like all things dull, it shouldn't be
played through.
Certainly it can be, making it a good candidate for abstraction.
On the other hand, I've had interesting training sequences both as a player and
as a gamesmaster; training, like any other activity, can be interesting to play
if desired.
Certainly in the player world, learning is often an interesting activity.
Warren J. Dew
Powderhouse Software
Peter Knutsen
Peter Knutsen wrote:
> We can also try to narrow down the range a bit, excluding
> attributes below 2 and above 6, because chances are you don't
> know anybody who has an attribute at 1 or anybody who has an
> attribute above 6. Now the span is only 15.
Oops! While I was falling asleep last night it occured to me
that I had made a mistake. The span in learning speed from 2->6
is not 15 but 10. A character with Aptitude 6 learns 10 times
faster than a character with an Aptitude of 2.
Also, it's worth nothing that characters will rarely have
Aptitudes of 6, since Aptitude is the weighted average
of several attributes. It is not uncommon for PCs and major
NPCs to have a single attribute of 6 or even 7, but the
Aptitude is the weighted average of several attributes, so
even such characters will mostly end up with Aptitudes in
the 4 to 5.5 range.
> --
> Peter Knutsen
--
Peter Knutsen
Peter Knutsen
Brandon Blackmoor wrote:
>
> On the one hand I think training time should happen offscreen. Training in
> any realistic sense is dreadfully dull to play through, and thus, like all
> things dull, it shouldn't be played through.
I probably didn't make myself clear, so I'll try again. Not your
fault.
I'm not talking about training speed but about learning speed.
A character has skill X, and then, through various means, he
accumulates experience points until he has enough to rise to skill
X+1. How much he needs to accumulate depends on his learning
speed.
Experience points can be accumulated through training, or through
skill use. Training is obviously the fastest method but as you say,
it is boring (both from a player and from a character perspective).
It is also costly, and may be time-consuming (if you're trying to
learn something that you don't have a high Aptitude for).
But the question that I need an RPG system to answer is this:
Joe with attributes X, Y and Z, and Billy with attributes X+4,
Y+3 and Z+2, both have no skill in tracking, but they visit the
Ranger School and spend two months there, learning to follow
tracks in the wilderness. When they come out, what skill levels
do they have?
> However, I am not entirely certain I understand what it is that your
> "learning span" is measuring. Is it the difference in the amount of time for
> a normal human and a talented human to both learn the same new skill to the
Actually it's the difference between an incompetent human and a
talented human.
> same degree of proficiency (in which case, in most game systems, the normal
> human marksman is actually much more *skilled* with her weapon, even if the
> end result is that she is on par with a more naturally gifted person who has
> less formal training)?
FFRE uses a direct skill level system. You train for X period of
time and this gives you Y skill levels.
Other systems, such as GURPS and Quest FRP, uses a system wherein
skill level equals (attribute + training).
This means that Joe and Billy might both train 3 levels or 2 DoM's,
but those levels or Degrees of Mastery would add on top of their
attributes, and since Billy has better attributes he ends up per-
forming much better than Joe even though they have the same number
of levels of training.
But in FFRE you simply roll a number of d10's equal to the number
of skill levels you have. This is why I badly need for Billy to
get more skill levels out of two months of training than Joe,
otherwise my suspension-of-disbelief is ruined.
> Or is it the difference in the amount of time it takes for a normal human
> and a talented human to both train to equivalent levels of competency if
> their natural talents were *removed* from the picture (i.e., if the normal
No, that's not what it means.
Skill level, in FFRE, is a flat measure of your ability to perform.
Joe gets to roll 3d10, Billy gets to roll 5d10.
I actually have some rules for talents, but Talent does not affect
your learning speed, instead is shifts the Target Number (or Task
Difficulty to use FFRE speak) down so that impossible tasks becomes
possible).
So if Joe I and Joe II both have the same attributes, but Joe II
has a Talent with the skill in question, then they learn at the
same speed (after 2 months of training, they emerge from the
school with skill 4 each) but Joe II ends up performing better
than Joe I, because he rolls vs a lower Task Difficulty.
But Talents are somewhat rare and should not be taken into account.
The default assumption is that nobody is talented.
> human golfer were given a handicap to compensate strictly for having merely
> ordinary hand-eye coordination, or if the more-than-human golfer were
> wearing gear from the Harrison Bergeron collection).
I'm sure you could set up the golf match such that the chance of
the less skilled character winning would be close to 50%, but that's
not really my question.
> bblac...@blackgate.net
> 15.february.2001
--
Peter Knutsen
Peter Knutsen
"Warren J. Dew" wrote:
>
> Peter Knutsen asks for thoughts on what the range is in learning speeds, from
> those least competent to those most competent.
>
> In terms of what's reasonable for a game, my own system has about a factor of 9
> difference in speeds from worst possible to best possible in the human range,
And worst possible is a person who is literally retarded (40'ish IQ)
and suffering from ADD and bunches of other conditions which severely
impairs learning speed?
And best possible is somebody like Zorro or Tarl Cabot or Sherlock
Holmes, or Batman?
> so I guess I consider that reasonable. I actually think that less might be
> better from a game balance standpoint.
I think I failed to mention that FFRE has 12 attributes. Not all
of them influences skill learning much (Size and Hardiness are
examples, and Fitness and Strength to a lesser extent) but I
still think that it's a lot better compared to GURPS which
only has 4 attributs.
In GURPS it is easy to create a character like Zorro or Sherlock
Holmes because you'd only need to give him a high DX/high IQ
(or both for Tarl Cabot or Batman). In FFRE Zorro would need high
Strength, Dexterity, Agility, Fitness and Perception, while Sherlock
Holmes would need high Perception (he's the archetypical 8)
Intelligence and Willpower.
So I'm not sure if game balance suffers much from the fact that
the span is 10 (or 70 if you encompass the entire range of human
possibility).
> From the standpoint of the player world, I've seen ballroom dancers get better
> after a month than others are after ten years, with the people involved
> spending comparable amounts of time per week, so that's a factor of 120 or so
> difference.
But that was game "physics", right?
A more interesting question is, did those game physics work for you?
Could you maintain the willing suspension of disbelief while seeing
that one character learned 120 times faster than another? Could your
character?
> I don't think that's the best representation of what actually happens, though.
> Based on my experience teaching ballroom, I think people increase in skill at a
> common rate until they reach some maximum imposed by their talent, interest, or
I'm inclined to suspect that this is because almost all ballroom
dancers are average. 95% of the population, in FFRE, has attributes
between 2 and 4. This "explains" why all your students learned at
the same rate. One might have had Agility 3 and Dexterity 4, giving
him an Aptitude of 3.3 which nets a learning speed of215% faster
than average, while another might have had Agility 4 and Dexterity 3,
giving a learning speed 52% faster than average, while the exceptional
student might have had *both* Agi and Dex at 4, giving him a learning
speed 78% faster than average (but this exceptional student is outside
the 95% of the population).
> instructor. A less than fully competent instructor can slow their learning,
> but the effect of attributes and interest seems to be to limit how far they can
> get, not how fast they get there.
That doesn't match my experience in the area of book learning.
I spent little time on homework, in the gymnasium yet I learned
more than my classmates even though they worked hard. Neither,
in fact, does it match my experience in the area of sports. I
trained as much as several of my friends yet they got better
than me quickly.
> I'd also note that all three of talent, interest, and instructor can change.
True.
> Brandon Blackmoor:
>
> Training in any realistic sense is dreadfully dull to play
> through, and thus, like all things dull, it shouldn't be
> played through.
>
> Certainly it can be, making it a good candidate for abstraction.
Except I'm not talking about training but about learning. You
can learn through training/instruction, but you can also learn
by doing.
[...]
> Warren J. Dew
> Powderhouse Software
--
Peter Knutsen
John Kim
A reply to Peter Knutsen concerning his question of what is a
reasonable ratio of learning speeds for a game system. His system
FFRE has average Aptitude stat of 3, with maximum of 8 (though he
suggests that this is highly cinematic -- comparable to a GURPS
stat of 20 or 21).
Apt. Stat 1 2 3 4 5 6 7 8 9 10
Learn Speed 0.1 0.4 1 1.8 2.8 4 5.4 7.1 9 11.0
Peter Knutsen <pe...@knutsen.dk> wrote:
>"Warren J. Dew" wrote:
>> In terms of what's reasonable for a game, my own system has about a
>> factor of 9 difference in speeds from worst possible to best possible
>> in the human range,
>
>And worst possible is a person who is literally retarded (40'ish IQ)
>and suffering from ADD and bunches of other conditions which severely
>impairs learning speed?
>
>And best possible is somebody like Zorro or Tarl Cabot or Sherlock
>Holmes, or Batman?
I think that using handicapped people as your basis is
probably a mistake, since the truly "worst" would be someone who
is all but brain-dead and thus gives you an infinite ratio of
learning speed. Also, very few PC's or major NPC's are seriously
mentally handicapped so the answer isn't very relevant to most
gameplay. Also, using fictional superheroes as your "best" is
also probably a little odd -- since fiction varies in just how
outrageously good the best can be (i.e. different versions of
Batman vary on how great he is, say).
I would tend to say that a better metric is the difference
between the "best" real-world human and an "average" real-world human.
Now, these are still pretty vague, but I think it is a little more
quantifiable. I think it is also a lot more relevant to the game
since very few PC's are seriously mentally handicapped.
Your system, then, has a ratio of x4 for average to Apt-6
(a "realistic maximum", perhaps) or x7 for average to Apt-8 (a
"cinematic maximum"). I would say that is pretty reasonable in
my view. Still, I'm not entirely sure.
-*-*-*-*-*-*-
Your assumption seems to be that there is a fixed ratio of
gaining skill between people based on their aptitude. I'm a little
doubtful of that. For example, I would say that there was a measurable
ratio between me and my classmates at learning high school physics
(maybe between x2 to x4). On the other hand, I don't think any
ratio applies to learning PhD physics. i.e. The average person
probably isn't going to get PhD-level skill no matter how long he
takes.
This is also true for many game systems. For example, you
compared FFRE to GURPS in your initial post. GURPS doesn't actually
have a fixed ratio of learning rate. Compare IQ-10 and an IQ-16
characters learning a Mental/Average skill. The ratio is x20 for
learning level 14 skill, but only x7 for learning level 16 skill,
and only x3 for learning level 18 skill.
This is actually the opposite of what I suggested above
for reality. i.e. I think that the ratio will be smaller for gaining
easy skills, but higher skill levels are increasingly difficult unless
you have the aptitude for it.
In my own vision of a realistic game system, the cost for
a skill would be linear up to your aptitude, and then maybe exponential
beyond that.
--
John H. Kim | Whatever else is true you
jh...@fnal.gov | Trust your little finger
www.ps.uci.edu/~jhkim | Just a single little finger can
UC Irvine, Cal, USA | Save the world. - Steven Sondheim, "Assassins"
Peter Knutsen
Peter Knutsen wrote:
>
> Peter Knutsen wrote:
> Also, it's worth nothing that characters will rarely have
^^^^^^^
Should be "Noting".
I make these slips disturbingly often :-(
> > Peter Knutsen
>
> --
> Peter Knutsen
--
Peter Knutsen
Peter Knutsen
John Kim wrote:
>
> A reply to Peter Knutsen concerning his question of what is a
> reasonable ratio of learning speeds for a game system. His system
> FFRE has average Aptitude stat of 3, with maximum of 8 (though he
> suggests that this is highly cinematic -- comparable to a GURPS
> stat of 20 or 21).
To clarify: An FFRE attribute of 8 has (should have) the same game-
mechanical effects as a GURPS attribute of 20 or 21. It should not,
however, be as rare or as inordinately expensive. GURPS states that
only one human being ever had an IQ of 20 ("smartest man ever").
FFRE makes no such claim, at any given time one or a score of
characters can exist with Dexterity 8, Willpower 8, Intelligence 8
or whatever.
> Apt. Stat 1 2 3 4 5 6 7 8 9 10
> Learn Speed 0.1 0.4 1 1.8 2.8 4 5.4 7.1 9 11.0
Yes, this is correct. Note that attributes of 0 are actually
possible, but player characters will rarely have them, since
an Agility of 0 or an Intelligence of 0 is crippling - you'd
be spastic (similar to a person suffering from a severe case
of cerebral palsy) or of sub-animal intelligence (a test IQ
in the range of 1-10).
> Peter Knutsen <pe...@knutsen.dk> wrote:
> >"Warren J. Dew" wrote:
> >> In terms of what's reasonable for a game, my own system has about a
> >> factor of 9 difference in speeds from worst possible to best possible
> >> in the human range,
> >
> >And worst possible is a person who is literally retarded (40'ish IQ)
> >and suffering from ADD and bunches of other conditions which severely
> >impairs learning speed?
> >
> >And best possible is somebody like Zorro or Tarl Cabot or Sherlock
> >Holmes, or Batman?
>
> I think that using handicapped people as your basis is
> probably a mistake, since the truly "worst" would be someone who
> is all but brain-dead and thus gives you an infinite ratio of
A person who is brain-dead would have an Intelligence of 0,
and probably Willpower 1 and Perception 1. (If you remove the
"all but" qualifier then WIL and PER drops to zero as well).
PCs will rarely have attributes of 1 but it should be both
legal and possible. For instance Raistlin Majere might be
seen as having Fitness 1 and Hardiness 1, although I think
it's more sensible to peg him down as 2/2.
> learning speed. Also, very few PC's or major NPC's are seriously
> mentally handicapped so the answer isn't very relevant to most
True for mental attributes, but what about others? Wizardly
PCs might have Strength 1 or Agility 1. Fighters could have
Willpower 1 or Perception 1 or Charisma 1.
> gameplay. Also, using fictional superheroes as your "best" is
> also probably a little odd -- since fiction varies in just how
I'm doing this deliberately because I'd like FFRE to be able to
simulate fiction as well as reality. Or rather: I'd like it to
be able to simulate a cinemaic realism.
> outrageously good the best can be (i.e. different versions of
> Batman vary on how great he is, say).
That's true. But one example, Zorro, is pretty consistently
good with the sword, as far as I know.
> I would tend to say that a better metric is the difference
> between the "best" real-world human and an "average" real-world human.
Okay, maybe you're right.
> Now, these are still pretty vague, but I think it is a little more
> quantifiable. I think it is also a lot more relevant to the game
> since very few PC's are seriously mentally handicapped.
True. Physical handicaps would be less rare, but rare nonetheless.
> Your system, then, has a ratio of x4 for average to Apt-6
> (a "realistic maximum", perhaps) or x7 for average to Apt-8 (a
> "cinematic maximum"). I would say that is pretty reasonable in
Yes, 6 is perhaps the realistic maximum. I based Intelligence
roughly in realworld IQ and decided that all attributes
(except possibly Strength and Size) should have the same
distribution. Intelligence 5 is the Mensa cutoff - you're
smarter than 98% of the population. Intelligence 6 is the
OATH cutoff - you're smarter than 99.9% of the population.
I'm not sure about higher values, although I do have some
old notes. Same distribution for the other attributes. If
you have Dexterity 6 then there probably aren't anybody
more dextrous than you in your town. There might be people
equal to you, but (probably) none better.
Also, we should keep in mind that learning speed (Aptitude)
is never based on a single attribute. (This is where, IMO.
GURPS fails).
A player might build a character who has Dexterity 8, intending
him to be a combat ace. But if this character has Strength 4
and Agility 5 and Fitness 5, and Perception 4, then the learning
speed for various combat skills will be
Meelee Weapons (DDDAAS) 4.5
Brawling (DDAAAS) 3.8
Archery (DDDPPS) 4
Ballroom Dancing (DDAAAA) 4
As you can see, he does not benfit overly much from his
cinematic Dexterity, when it comes to skill learning speed.
When it comes to direct DEX rolls, however, he shines!
If the player had skill learning in mind, then he would have
been better of by choosing a more balanced set of attribute
values such as DEX 7, AGI 6, STR 5, PER 6. (gives meelee 4.5,
brawling 4.2, archery 4.5, Dancing 4.5).
> my view. Still, I'm not entirely sure.
>
> -*-*-*-*-*-*-
>
> Your assumption seems to be that there is a fixed ratio of
> gaining skill between people based on their aptitude. I'm a little
> doubtful of that. For example, I would say that there was a measurable
> ratio between me and my classmates at learning high school physics
> (maybe between x2 to x4). On the other hand, I don't think any
> ratio applies to learning PhD physics. i.e. The average person
> probably isn't going to get PhD-level skill no matter how long he
> takes.
That's true, about the average person. But maybe that is because
it has never been tried? The average person lacks the drive and
motivation to spend 12 years getting a PhD which you could take
in 3 years. Could he do it if he tried? I'm not sure.
But, assuming you have a PhD, wasn't it possible that some of
your studymates learned slightly slower or faster than you did?
Maybe they had less spare time than you did, having to use more
time on study and homework? Maybe some had more spare time?
I'm talking differences like 10-30% percent here.
> This is also true for many game systems. For example, you
> compared FFRE to GURPS in your initial post. GURPS doesn't actually
> have a fixed ratio of learning rate. Compare IQ-10 and an IQ-16
This is true. I was writing based on a fudgey guess of how
much study a character with IQ X would need to reach skill
level Y.
A character with IQ 8 would need 4 points to reach skill level 10,
while a character with IQ 12 would need only one point. Increase
the IQ gap and you increase the time required dramatically. Even
if you move the span from IQ 8->IQ 12 to IQ 11->IQ 15, the difference
is still there.
> characters learning a Mental/Average skill. The ratio is x20 for
> learning level 14 skill, but only x7 for learning level 16 skill,
> and only x3 for learning level 18 skill.
Yes, this is because of the plateau effect of GURPS, yet
another feature that I'm not happy with.
> This is actually the opposite of what I suggested above
> for reality. i.e. I think that the ratio will be smaller for gaining
> easy skills, but higher skill levels are increasingly difficult unless
> you have the aptitude for it.
My initial idea for FFRE, way back in 1997, was to have each skill
level cost twice the amount of experience (=study/training/learning)
as the previous.
So skill 1 would cost 1 unit, skill 2 would cost 2 units, skill
3 would cost 4 units.
Later I thought that this might be too harsh, and make is very
difficult to reach higher skill levels. It might in fact "force"
the players to spread out and learn many skills, instead of
focusing on a few skills.
Now, I've gotten a different idea. The cost progression starts
out triangular. skil1 1 costs 1 unit, skill 2 costs 2 units,
skill 3 costs 3 units, skill 4 costs 4 units.
But there's a catch!
Once your skill exceeds your Aptitude, the cost starts to double.
So let us examine two characters, Billy has an Aptitude of 3.2
and Joe has an Aptitude of 5.1.
They first learn one unit each (although Joe learns 60& faster
than Billy), reaching skill 1. Then they learn two units more,
each reaching skill 2. Then they learn three units more, reaching
skill 3.
(Assuming that one unit is one 30-day month, Billy spent 158
days on this, while Joe spent 62 days. A normal person would
have needed to spend 180 days).
But to reach the next level of skill, Billy needs 6 units.
Joe only needs 4 units.
To each the next level of skill, Billy would need 12 units,
and Joe 5 units.
To reach skill 6, Billy would need 24 units, and Joe would
need 10 units.
To reach skill 7, Billy would need 48 units, and Joe would
need 20 units.
To reach skill 5, Joe has paid 15 units (1+2+3+4+5), while
Billy has paid (1+2+3+6+12=) 24 units.
To reach skill 7, Joe has paid (1+2+3+4+5+10+20=) 35 units,
while Billy has paid (1+2+3+6+12+24+48=) 96 units.
In total, Billy has to spend 7 times as much effort and time
as Joe, to learn the skill to 7th level. Joe learned 7 times
faster, on average.
To learn the skill to 5th level, Billy only had to spend
4 times as much effort and time as Joe (since the steepness
effect for Billy hasn't really kicked in yet).
To learn the skill to 3rd level, Billy only had to spend
2.5 times the effort (since the steepness effect hadn't
kicked in at all, until this point).
The steepness effect also has the nice simulated effect of the
less gifted eventually deciding that it's not worth the effort
to improve more. The character of average intellect realizes
that he would need a decade to climb from his hard-earned
Master's Degree to a PhD, whereas his classmates can do it
in 3 years, so he decides to not study physics any more.
> In my own vision of a realistic game system, the cost for
> a skill would be linear up to your aptitude, and then maybe exponential
> beyond that.
Seems we're thinking along similar lines :-)
> --
> John H. Kim
--
Peter Knutsen
John Kim
"http://www.lwithers.demon.co.uk/opes/ffre/index.html" However,
it is gone now. What happened to it?
I know you are planning to improve FFRE, but the old version
still has some useful ideas in it. It is as good as many of the
free RPG's out there (which is admittedly faint praise, but still).
Peter Knutsen
John Kim wrote:
>
> Incidentally, Peter -- FFRE used to be on the net at
> "http://www.lwithers.demon.co.uk/opes/ffre/index.html" However,
> it is gone now. What happened to it?
The Opes project was intended to be a CRPG engine, and in late 1997
I approached the programmer, Lawrence Withers, responding to his
request for ideas. I told him about the FFRE (Free Fantasy Roleplaying
Engine) task resolution system and the attribute system, and he liked
the ideas, so I became a member of the Opes team (initially the
LWAngband team). But since I was not a progammer (they were going
to use C or C++, and I only know six million versions of BASIC) I
had no influence on the pace of the programming, and after about a
year of little progress, Lawrence abandoned the project. A year
after that, he also removed the website.
In late 1998 or early 1999, I became enamoured of Quest FRP, so
I stopped independent development of FFRE, and started adding
tweaks to Quest FRP instead.
But now, I've become increasingly frustrated with the state of
roguelike computergames, especially Angband and variants (in
particular Zangband - Zelazny-Angband). I find many annoying
problems with those games, such as the combat system and the
attribute system. Some weeks ago I decided that maybe I should
program my own roguelike or CRPG (most likely a mix of the two).
And it was immediately clear to me that FFRE was the system to
use, since the action resolution mechanic is much more sophi-
sticated than what's used in any existing roguelike or CRPG that
I know of.
But FFRE was never developped into a "complete" system. Like FUDGE,
it has never progressed beyond the "build your own system kit". I
did specify attributes early on, but never skills, nor spells.
So in order to make this CRPG I first have to make a proper RPG
system out of FFRE.
Also, I've thought seriously about starting a hard science fiction
campaign, using the GURPS space ship creation system and many of
the GURPS tech books as references. My initial idea was to use a
heavily modified version of GURPS as the character creation and
action resolution engine. But then it occured to me that FFRE
might do the job. FFRE was initially intended to work with any
genre except really powerfull superheroes (like Superman), so
why not use it for science fiction, as well as for fantasy?
The only problem, in this department, are the attributes Psyche
and Faith, which are less useful in a science fiction setting.
Psyche normally governs Mage magic, which of course is non-
existent, but it also governs the Danger Sense skill, so it has
some use. Faith, similarly, governs Clerical magic, but it also
governs your Luck Point Pool, so it has some usefulness. But it
is not viable to build your character around a "high Psyche" or
"High Faith" concept, in a science fiction setting.
I don't think this will create huge problems, as I can just
advise the players not to take very high values in those
attributes, though. The other 10 attributes are useful regard-
less of the setting.
> I know you are planning to improve FFRE, but the old version
> still has some useful ideas in it. It is as good as many of the
> free RPG's out there (which is admittedly faint praise, but still).
None of the useful ideas will be removed. The skill learning
system will be improved but everything else stays. Attributes,
Spell Point Pool formulas (they might get tweaked slightly but
they stay mostly as they are) and the triple magic systems
(Mage magic, Clerical magic, Psionic magic). And of course the
action resolution system stays too, because I'm extremely fond
of it.
The biggest problem I have is whether I should change the dice
reading policy, so that a roll of 0 is read asa roll of 10, or
stick to the first policy which was to read 0 as 0.
For a CRPG, the ability to display the result of any single
dice as one character is a huge benefit, but it will confuse
human players who are used to the Storyteller system or just
to the standard convention of reading d10s. Then again it can
probably be overcome.
If you're extremely eager to get a glimpse of the FFRE system,
the stuff I've got at the moment, I can e-mail you the files
or post a short summary in here.
> --
> John H. Kim
--
Peter Knutsen
looking forward to Brian's review of FFRE v1.0, perhaps half
a year from now...
Bradd W. Szonye
>> From the standpoint of the player world, I've seen ballroom dancers
>> get better after a month than others are after ten years, with the
>> people involved spending comparable amounts of time per week, so
>> that's a factor of 120 or so difference.
Peter Knutsen <pe...@knutsen.dk> wrote:
>But that was game "physics", right?
Um, no -- I believe he was talking about the *player* (ie., real) world.
>A more interesting question is, did those game physics work for you?
>Could you maintain the willing suspension of disbelief while seeing
>that one character learned 120 times faster than another? Could your
>character?
I don't think this is a matter of suspension of disbelief, seeing as
this was his experience as a ballroom dance instructor in real life.
On the other hand, some aspect of the real world do require some
suspension of disbelief, so maybe ...
--
Bradd W. Szonye Work: br...@cup.hp.com
Software Design Engineer Home: bra...@concentric.net
Hewlett-Packard Cupertino Site, iFL Phone: 408-447-4832
Leszek Karlik
disseminated foul capitalist propaganda:
[...]
>I think I failed to mention that FFRE has 12 attributes. Not all
>of them influences skill learning much (Size and Hardiness are
>examples, and Fitness and Strength to a lesser extent) but I
>still think that it's a lot better compared to GURPS which
>only has 4 attributs.
OK. Why do you think that 'many attributes' approach is more preferable
to 'less attributes' approach?
In the mechanics set I am developing, Idee Fixe, there are 4 attributes
in the basic rules (Body, Will, Dexterity and Intelligence) which can
in turn be split into 2 sub-attributes each, first attribute costing
1/3rd of a full attribute and the second one costing 2/3rds of a full
attribute (the sub-attributes are, respectively, Muscles and Resilience,
Self-control and Willpower, Reflexes (initiative) and Coordination (DEX-based
skills use this) and Perception and Intellect (INT-based skills use this)).
Those attributes can be also modified slightly by using Ads and Disads,
but generally that's all.
I like it, as a matter of fact. The result is that there are two attributes
that govern 'strength' (of body and mind, respectively), and two that
govern 'coordination' (again, of body and mind). The second two are much
more expensive, since skills are based on them.
Also, the scale is much larger than in FFRE. Human range of attributes
goes from 0 (total atrophy, insanity, paralysis and coma, respectively)
to 8, which is the minimum for a non-crippled adult, up to 20, with 10
being average and 20 being the total perfection (not very cheap, though,
so 14-15 will usually be the max attribute encountered).
This mechanic is used for a rather gritty cyberpunkish setting, BTW. There's
also a cinematic version which has a scale compressed by a factor of four,
when the characters are not necessarily human, but rather cyborgs, psionics,
human-Alien hybrids, elves (Engineered Life Forms) and other pulp heroes
engaged in a cinematic behind-the-scenes struggle for the control of the
unsuspecting world. :>
>Peter Knutsen
Leslie
--
Leszek 'Leslie' Karlik; YASL(TM); SNAFU TANJ TANSTAAFL; KDSM FM; /^\ lk
Do you want to join the Ancient Illuminated Seers of Bavaria? / (*) \
Put $ 3,125.00 in a cigar box and bury it in your backyard. / \
One of our *Underground* Agents will contact you shortly. /_____________\
Peter Knutsen
Leszek Karlik wrote:
>
> On Fri, 16 Feb 2001 11:40:10 +0100, Peter Knutsen <pe...@knutsen.dk>
> disseminated foul capitalist propaganda:
>
> [...]
> >I think I failed to mention that FFRE has 12 attributes. Not all
> >of them influences skill learning much (Size and Hardiness are
> >examples, and Fitness and Strength to a lesser extent) but I
> >still think that it's a lot better compared to GURPS which
> >only has 4 attributs.
>
> OK. Why do you think that 'many attributes' approach is more preferable
> to 'less attributes' approach?
Because in GURPS it's easy to abuse the system by optimizing
your character. You spend a few points and in return you get
much optimization for a broad range of activities.
Optimization per se is not bad. It becomes bad when it's easy,
cheap and broad.
FFRE has many attributes. This means that by raising one
attribute to a high value, you optimize for a small subset
of the activities in the game. In fact it is not really
possible to optimize across the board.
Whereas in GURPS, you can optimize your character for skills
by taking a high IQ and a high DX.
In FFRE, you can optimize too, but you can choose to optimize
for meelee skills, or for ranged combat skills, or for know-
ledge skills, or for thief skills, or for wizard magic, or for
social skills.
So GURPS can be abused. This makes it necessary for the GM
to step in and survey each character sheet, approving or
rejecting each according to how optimized it is. I.e. not
all choices possible within the rules system are acceptable.
FFRE, on the other hand, can not be abused. All choices
possible within the rules system are by definition accep-
table. The GM should still check out the character sheet
but only because it is his duty to help the player. Maybe
the player overlooked a skill who is of vital importance
to the character concept. The GM stepping in and giving
the player a helpful game-mechanical suggestion creates
a pleasant and cooperative atmosphere.
Whereas in GURPS the GM stepping in, telling the player
that he can not make a DX 15 IQ 15 character, creates a
hostile atmosphere due to the implicit accusation of
abuse and munchkinism. A hostile atmosphere is not
conductive to roleplaying.
Also the GURPS system forces the GM to improvise rules
either before or after character creation. In practice it
is always (99.9% of cases) after character creation (only
a rulesmesiter like me would try to foresee possible and
undesirable player choices). And the Simulationist playing
style prefers such GM improvisations to be kept at an
absolute minimum.
> In the mechanics set I am developing, Idee Fixe, there are 4 attributes
May I ask what the name means, or where it's inspired from?
> in the basic rules (Body, Will, Dexterity and Intelligence) which can
> in turn be split into 2 sub-attributes each, first attribute costing
> 1/3rd of a full attribute and the second one costing 2/3rds of a full
> attribute (the sub-attributes are, respectively, Muscles and Resilience,
> Self-control and Willpower, Reflexes (initiative) and Coordination (DEX-based
> skills use this) and Perception and Intellect (INT-based skills use this)).
And which one of those has 1/3 cost and which has 2/3 cost?
It's not quite clear to me.
> Those attributes can be also modified slightly by using Ads and Disads,
> but generally that's all.
I have ads and disads too, but they're not supposed to be used
much and thus they are very costly. This means that players
are discouraged from using them, so they will only be used
if the character concept truly needs it.
> I like it, as a matter of fact. The result is that there are two attributes
> that govern 'strength' (of body and mind, respectively), and two that
> govern 'coordination' (again, of body and mind). The second two are much
> more expensive, since skills are based on them.
>
> Also, the scale is much larger than in FFRE. Human range of attributes
> goes from 0 (total atrophy, insanity, paralysis and coma, respectively)
> to 8, which is the minimum for a non-crippled adult, up to 20, with 10
> being average and 20 being the total perfection (not very cheap, though,
> so 14-15 will usually be the max attribute encountered).
Again, how then does a player determine whether his character
concept calles for a value of 12, 13 or 14? It seems to me
that the player will often either look at how many points
he has to spend (metagame concern) or else he will look for
a breakpoint or some other optimizing scheme (also a meta-
game concern), whereas in FFRE he will easily be able to
express the character concept (Simulation, free of meta-
game concerns).
My question is, why do you need all those intermediate steps?
I realize that it's nice to have an average of 10, but it is
really necessary to have all those microscale steps of
improvement? 11 is slightly better than 10, 12 is slightly
better than 11.
> This mechanic is used for a rather gritty cyberpunkish setting, BTW. There's
> also a cinematic version which has a scale compressed by a factor of four,
> when the characters are not necessarily human, but rather cyborgs, psionics,
> human-Alien hybrids, elves (Engineered Life Forms) and other pulp heroes
> engaged in a cinematic behind-the-scenes struggle for the control of the
> unsuspecting world. :>
Sounds interesting.
I hope, though, that FFRE can support both cinematic and
highly realistic campaigns without having to compress or
squeeze the scale.
> Leslie
--
Peter Knutsen
Leszek Karlik
disseminated foul capitalist propaganda:
[...]
>> OK. Why do you think that 'many attributes' approach is more preferable
>> to 'less attributes' approach?
>Because in GURPS it's easy to abuse the system by optimizing
>your character. You spend a few points and in return you get
>much optimization for a broad range of activities.
Well, yes, but that's because GURPS point system is broken.
(Which I have learned the hard way when my Starfighter Ace character
turned out to be on the same level of competence when it comes to
flying starfighters as the high-attributes, lots-of-half-point-skills
Space Commandoe character in an SF game I was playing in. ;)).
>Optimization per se is not bad. It becomes bad when it's easy,
>cheap and broad.
Yes, but again, IMO that's a feature ("it's not a bug, it's a
feature" ;))) of GURPS, not of 'low number of attributes' scheme
in general. (Though you do have a point)
[...]
>In FFRE, you can optimize too, but you can choose to optimize
>for meelee skills, or for ranged combat skills, or for know-
>ledge skills, or for thief skills, or for wizard magic, or for
>social skills.
OK. You do have a point there. In Idee Fixe, there's a broad optimization for
either physical (dexterity) or intellectual (intelligence) skills, and a
narrow optimization done by Ads/Disads.
However, I chose to avoid the 'GURPS Syndrome' by going in a different
direction - the price for attributes that are base for skills (dexterity
and intelligence) is high, and goes up rapidly (+10 points for 11, then
another +20 for 12, +30 for 13, +40 for 14 and +50 for 15 - character
points and attribute (as well as the low-to-mid values of skills) range
is comparable to GURPS). Thus, the infamous DX 15, IQ 15 character
would in IF cost 300 points for attributes alone - and a starting character
is based on 150 points).
Also, the second part of GURPS syndrome is the use of 3d6 for skill
checks. Thus, the difference between a skill of 12 and a skill of 16
is a very big one, between 16 and 20 is lower, and 20 and 24 is even
smaller yet. This is prone to abuse (as GURPS tends to be).
The testing system in IF is similar to Interlock, only without the
Murphy World rule of crit fail on 1 and auto success on 10 (the d10 roll
that is added to skill or attribute is open-ended both ways, and only
on negative final roll result does one get a critical failure).
So, it is possible to avoid the pitfalls of gurpsification ;) when
creating a system with less attributes.
[...]
>FFRE, on the other hand, can not be abused.
I think this statement is rather optimistic. :)
[...]
>> In the mechanics set I am developing, Idee Fixe, there are 4 attributes
>May I ask what the name means, or where it's inspired from?
It's from French, meaning 'an obsession'. Basically, characters are
supposed to be the obsessed ones, the ones that do not fit into
the 80% of the population that lives on the dole, staring into their
Activision sets or running VR porn programs.
For more explanations on the history of the name I'd have to ask the
main creator - I'm the mechanics and internal consistency guy, being
somewhat of an anal retentive nitpicker. :>
>> in the basic rules (Body, Will, Dexterity and Intelligence) which can
>> in turn be split into 2 sub-attributes each, first attribute costing
>> 1/3rd of a full attribute and the second one costing 2/3rds of a full
>> attribute (the sub-attributes are, respectively, Muscles and Resilience,
>> Self-control and Willpower, Reflexes (initiative) and Coordination (DEX-based
>> skills use this) and Perception and Intellect (INT-based skills use this)).
>And which one of those has 1/3 cost and which has 2/3 cost?
>It's not quite clear to me.
Sorry. The first of each pair is the cheaper one - in case of Body, it's
Muscles (since it's short term activities, as opposed to general hardiness
and ability to go on and not die yet), in case of Will it's Self Control
(which is the ability to control one's fears and obsessions now, as opposed to
the long-term Willpower stat that govern's one resistance to stress and
pain), and in case of the Dexterity and Intelligence stat's its those
parts that are not the basis of skill calculation.
>> Those attributes can be also modified slightly by using Ads and Disads,
>> but generally that's all.
>I have ads and disads too, but they're not supposed to be used
>much and thus they are very costly. This means that players
>are discouraged from using them, so they will only be used
>if the character concept truly needs it.
Well, this is a very good solution - I found that relying on a system
of advantages/disadvantages for character fine-tuning lengthens the
character creation process, which for some people is a put off.
[...]
>> Also, the scale is much larger than in FFRE. Human range of attributes
>> goes from 0 (total atrophy, insanity, paralysis and coma, respectively)
>> to 8, which is the minimum for a non-crippled adult, up to 20, with 10
>> being average and 20 being the total perfection (not very cheap, though,
>> so 14-15 will usually be the max attribute encountered).
>
>Again, how then does a player determine whether his character
>concept calles for a value of 12, 13 or 14? It seems to me
>that the player will often either look at how many points
>he has to spend (metagame concern) or else he will look for
>a breakpoint or some other optimizing scheme (also a meta-
>game concern), whereas in FFRE he will easily be able to
>express the character concept (Simulation, free of meta-
>game concerns).
>
>My question is, why do you need all those intermediate steps?
Ah. A good question, and one that I am able to answer.
Feel. You see, in cyberpunk (or post-cyberpunk, as the case may be ;))
there's this obsession with 'the Edge'. You need to get every possible
edge to survive. In case of very wide steps, all those tiny actions
you do to gain a slight advantage are lost in the big, blurry areas
between steps.
In case of a 0-20 scale, I have a resolution of approximately 5% (OK,
so it comes somewhere in between 5 and 10% - after all, a 10-sided dice
is used, but it's open ended, so the resolution is slightly higher than
10%, and even though the attributes will rarely go below 8, they might,
with chemical (combat drugs), cybernetical or, in some cases, psionic
enhancement go above 20). The combat drugs that characters use do not
have to be either extremely powerful (an Attribute raise of 1 in a 0-9
scale is much higher than in a 0-20+ scale) while stile having some
discernible game effect.
>I realize that it's nice to have an average of 10, but it is
>really necessary to have all those microscale steps of
>improvement? 11 is slightly better than 10, 12 is slightly
>better than 11.
It's not microscale - d100 is microscale ;)). That's why I dislike percentile
scale in RPGs - trying to model reality with a 1% accuracy is too much
for me.
[...]
>I hope, though, that FFRE can support both cinematic and
>highly realistic campaigns without having to compress or
>squeeze the scale.
Well, you do have a problem of impossible difficulties - TD 11 in
FFRE is just Plain Impossible (TD 13 for somebody with a Major Talent).
In Pulp IF, there's no limit to how high can one go. It's something
like GURPS Black Ops when it comes to power level, but without the
tedious seven-hundred-GURPS-points-character bookkeeping. :)
Since the scale is compressed four-fold, the entire 'normal IF' scale
fits between 0-4 (ok, 0-5 - skills in IF can get above 20, as a matter of
fact, they can get to any height, but the price gets kinda steep ;))),
and the average Pulp IF character (if there's anything like that) has
Attributes of 5 and skills in the 8-12 range. And can get to, frex, 20
or higher (which would correspond to a 'normal IF' skill of 80-something).
The sky's the limit. :)
But I like FFRE. It sure seems nice for fantasy campaigns, and I
like the magic system very much. It just won't scale indefinitively
as it is now.
However, you might consider having the d10's being open-ended. To avoid the
Shadowrun problem of 'difficulty of 7 is the same as difficulty of 6' you might
say that rolling 10 is 9+re-roll.
Make the capability to make open-ended rolls an Advantage that is available
only in cinematic campaigns, and voila. The difficulties above 10 will still
drop the number of successess sharply, but FFRE will now scale upwards
without limits (theoretically ;))).
I still wouldn't use it for gritty near-future (cyberpunk or SF), but
then, that's what I keep CORPS for.
Warren J. Dew
FFRE has many attributes. This means that by raising one
attribute to a high value, you optimize for a small subset
of the activities in the game. In fact it is not really
possible to optimize across the board.
Whereas in GURPS, you can optimize your character for skills
by taking a high IQ and a high DX.
Could you expand on this? It seems to me that taking two high attributes out
of four in GURPS would be equivalent to taking eight high attributes out of 16
in FFRE. You don't think a character with eight high attributes would be
strong in FFRE?
Or do you just think it would be more difficult to find the 'right' eight?
My question is, why do you need all those intermediate steps?
I realize that it's nice to have an average of 10, but it is
really necessary to have all those microscale steps of
improvement? 11 is slightly better than 10, 12 is slightly
better than 11.
I like having a lot of steps - or at least, I like having a lot of steps above
average. I like to be able to have small differences between characters - Fred
is "only a little stronger than" Bob, or whatever.
John Kim
Peter Knutsen <pe...@knutsen.dk> wrote:
>Leszek Karlik wrote:
>> OK. Why do you think that 'many attributes' approach is more
>> preferable to 'less attributes' approach?
>In FFRE, you can optimize too, but you can choose to optimize for
OK, I am going to start on a small rant here. The problem
I have seen with nearly all "many-attributes" schemes is that the
meaning of the attributes becomes increasingly obscure. i.e.
Rather than making your character good in categories like "ranged
combat skills" or "thief skills", instead you end up with characters
who are "very good Perception, fair Reasoning, and above-average
Intuition". The game designer then spends time explaining at
length what "Reasoning" is as opposed to "Intelligence", and even
more time making difficult decisions about whether X skill is more
Intuition-based or Perception-based.
I find that all this simply obscures making a character.
In general, when I make a character I think of what he is good at:
i.e. "he is good at science skills" rather than the abstract
categories like "Perception" as opposed to "Reasoning". In practice
I am forced to reverse-engineer the system. That is, first I imagine
the charcter and what she is good at, then I juggle numbers until the
attributes give me roughly what I am looking for (which is often a
pretty poor approximation).
For example, say I want to represent a sharp-witted urban lawyer
vs a keen barbarian tracker. They are both perceptive, but in different
fields. However, the way most attributes are set up, the lawyer can
easily become a great tracker -- whereas IMO his background would make
it very difficult.
As another example, I personally am very good at math and
science, but am not nearly as good in literature and humanities, and
am poor in foreign languages. Again, this is the sort of difference
that doesn't really show up in abstract categories like "Reasoning".
-*-*-*-*-*-*-*-*-*-
My two cents is this: Why not have stats based around the
general skill categories, like Peter's example of thief skills vs
knowledge skills vs melee skills, etc.? By having categories like
outdoors skills and so forth (as many logical skill groupings do),
you can represent a lot of characters much more simply.
In part, my annoyance is not that the abstract attributes are
that awful, but that virtually no one tries the alternative approach.
Peter Knutsen
Leszek Karlik wrote:
>
> On Mon, 05 Mar 2001 23:31:31 +0100, Peter Knutsen <pe...@knutsen.dk>
> disseminated foul capitalist propaganda:
>
> [...]
> >> OK. Why do you think that 'many attributes' approach is more preferable
> >> to 'less attributes' approach?
> >Because in GURPS it's easy to abuse the system by optimizing
> >your character. You spend a few points and in return you get
> >much optimization for a broad range of activities.
>
> Well, yes, but that's because GURPS point system is broken.
Why is it broken?
> (Which I have learned the hard way when my Starfighter Ace character
> turned out to be on the same level of competence when it comes to
> flying starfighters as the high-attributes, lots-of-half-point-skills
> Space Commandoe character in an SF game I was playing in. ;)).
GURPS has two faults. It has way too few attributes, and it
uses a system wherein:
Skill = (Attribute + Training)
FFRE will not use that kind of system, instead it uses a
system similar to Quest FRP, where the attribute determines
how much experience each level of a skill costs.
For instance, learning levels 1+2+3 of a skill might cost
360 SP for a normal person (Aptitude 3 gives a Learning Speed
of 1), but for a person who is better suited for learning the
skill in question levels 1-3 might cost only 257 (LS 1.4) or
200 (LS 1.8) or 164 (LS 2.2) SP.
In GURPS, if you have a high attribute, you get a "head start"
in all skills. This does not happen in FFRE.
> >Optimization per se is not bad. It becomes bad when it's easy,
> >cheap and broad.
>
> Yes, but again, IMO that's a feature ("it's not a bug, it's a
> feature" ;))) of GURPS, not of 'low number of attributes' scheme
> in general. (Though you do have a point)
Yes, but that's why I dislike systems with a low number of
attributes.
> [...]
> >In FFRE, you can optimize too, but you can choose to optimize
> >for meelee skills, or for ranged combat skills, or for know-
> >ledge skills, or for thief skills, or for wizard magic, or for
> >social skills.
>
> OK. You do have a point there. In Idee Fixe, there's a broad optimization for
> either physical (dexterity) or intellectual (intelligence) skills, and a
> narrow optimization done by Ads/Disads.
I don't like that. It becomes difficult to have a party of more than
2 characters, since you can't make them different on the most basic
and primal level (general attributes).
You have to resort to the secondary (advantages) or tertiary (actual
skills) to differentiate them, and that's ungood.
> However, I chose to avoid the 'GURPS Syndrome' by going in a different
> direction - the price for attributes that are base for skills (dexterity
> and intelligence) is high, and goes up rapidly (+10 points for 11, then
> another +20 for 12, +30 for 13, +40 for 14 and +50 for 15 - character
> points and attribute (as well as the low-to-mid values of skills) range
> is comparable to GURPS). Thus, the infamous DX 15, IQ 15 character
> would in IF cost 300 points for attributes alone - and a starting character
> is based on 150 points).
Do you use the Skill = (Attribute + Training) method, or something
else?
Anyway your solution doesn't solve the problem, as it then
becomes flat out impossible to play a character with an
exceptional attribute - for instance to play one of the ten
most dextrous men on the planet.
The reality is that attribute (dexterity, intelligence,
perception or whatever) doesn't give you so much advantage
that players should be prohibited from playing characters
with exceptional attributes.
But most systems gives unrealistic benefits to characters with
high attributes, which means that you are forced, as a GM (or
as a rules designer) to make the cost of high attributes some-
thing that the players can't reach.
> Also, the second part of GURPS syndrome is the use of 3d6 for skill
> checks. Thus, the difference between a skill of 12 and a skill of 16
> is a very big one, between 16 and 20 is lower, and 20 and 24 is even
> smaller yet. This is prone to abuse (as GURPS tends to be).
How is this part abused?
It's very useful to have a skill of 18 or 20, or higher, since
you can still perform well even under adverse conditions.
FFRE works in a similar way, but gives the added advantages
of a greater chance of spectacular successes. This does not go
up further than 6 on 3d6 for GURPS, no matter how high your
skill is. In FFRE, your chance of a Great Success(3S) can easily
reach 20% or 33%, or your chance of a Superb Success(4S) reach
5% or 10%.
> The testing system in IF is similar to Interlock, only without the
> Murphy World rule of crit fail on 1 and auto success on 10 (the d10 roll
> that is added to skill or attribute is open-ended both ways, and only
> on negative final roll result does one get a critical failure).
But doesn't this produce a flat critical chance for everyone?
FFRE is my attempt to get 100% away from the flat probabilities
of criticals. For each increase in skill, your chance of a
critical success should go up, and your chance of a critical
failure should go down. And it should continue and continue and
continue to go up (down), rather than hit a plateau beyond a
certain skill rating.
> So, it is possible to avoid the pitfalls of gurpsification ;) when
> creating a system with less attributes.
I don't see how that avoids GURPS syndrome.
All it does is to prevent players from making characters who
are exceptional attribute-wise.
A couple of months ago I tried to model myself in GURPS, but I
gave up again. Even with an IQ modestly pegged at 14 I got
ridiculously high levels in skills that I had invested microscopic
(default) or minimal effort(0.5 CP) in.
I knew before that that GURPS was a highly problematic system, and
one that I would never GM and probably never play, but that experiment
proved it beyond doubt.
It could be an interesting experiment, once FFRE is completed,
to actually finish my GURPS writeup and then do a similar writeup
in FFRE, in order to compare the two systems and show (prove?)
that FFRE is lightyears ahead.
> [...]
> >FFRE, on the other hand, can not be abused.
> I think this statement is rather optimistic. :)
Only because you're infected with the meme that all systems
are equally good and equally easy to abuse.
The reality is that some systems are better than others, and
that some systems may very well be abuse-proof. Quest FRP
certainly seems to me to be abuse-proof - I've never seen any-
body make an abusive character in Quest FRP.
And please note that I do not consider "optimization for a
specified area of endavour" to be abuse. It is not abusing the
system to make character creation choices such that your
character is optimized for meelee combat, or archery, or
sneaking, or wizard magic.
It is only abuse if you get "edge" too cheaply, which usually means
that you can gain much optimization for a broad area of endavour
(skills period) for a low cost (such as in GURPS).
Other people may have a broader definition of the term "abuse".
> [...]
> >> In the mechanics set I am developing, Idee Fixe, there are 4 attributes
> >May I ask what the name means, or where it's inspired from?
> It's from French, meaning 'an obsession'. Basically, characters are
> supposed to be the obsessed ones, the ones that do not fit into
> the 80% of the population that lives on the dole, staring into their
> Activision sets or running VR porn programs.
OK :-)
> For more explanations on the history of the name I'd have to ask the
> main creator - I'm the mechanics and internal consistency guy, being
> somewhat of an anal retentive nitpicker. :>
No problem, I was just curious.
It sounds like a good idea to make an RPG about characters who
are exceptional. Actually I think all roleplaying gaming should
be about exceptional characters because normality is so dull,
but it's interesting to focus on exceptional drive instead of
exceptional talent or exceptional skill.
> >> Those attributes can be also modified slightly by using Ads and Disads,
> >> but generally that's all.
> >I have ads and disads too, but they're not supposed to be used
> >much and thus they are very costly. This means that players
> >are discouraged from using them, so they will only be used
> >if the character concept truly needs it.
>
> Well, this is a very good solution - I found that relying on a system
> of advantages/disadvantages for character fine-tuning lengthens the
> character creation process, which for some people is a put off.
I guess I agree with you.
I haven't thought much about it, though, but it seems to me that
many of the GURPS advantages are only necessary because of the
low number of attributes.
So the more of the character you can express on the primary level,
the attribute level, the better.
I'm not sure how FFRE will turn out, in detail, but my guess is
that many characters will have no advantages.
In GURPS, you have to search low and high before you find a PC
(or major NPC) without any advantages. It's a very integral
part of the system.
FFRE will by no means discourage character advantages. It's just
that they won't be very necessary.
First of all there are more attributes. Secondly "Social Perks"
are a seperate category, you buy them with a seperate pool.
Examples of this is NPC Allies, NPC Contacts, Wealth, Fame,
Reputation. This is much of what's normally thought of as
Advantage. Many RPG characters who have Advantages turn out
to only have Social Advantages.
And since such aren't intrinsic to the character, I've decided
to treat them as seperate.
Characters with advantages will still exist, be legal and be
necessary. Anybody with above-average quality of appearance
(Attractive, Very Handsome, Mindbogglingly Beautiful) needs
an Advantage. Characters who are Ambidextrous, should I figure
out how to handle the advantage mechanically. nonHuman traits,
so simulate fantasy-genre humanoids, will be Advantages, as
will the benefits granted by cybernetic implants.
> [...]
> >> Also, the scale is much larger than in FFRE. Human range of attributes
> >> goes from 0 (total atrophy, insanity, paralysis and coma, respectively)
> >> to 8, which is the minimum for a non-crippled adult, up to 20, with 10
> >> being average and 20 being the total perfection (not very cheap, though,
> >> so 14-15 will usually be the max attribute encountered).
> >
> >Again, how then does a player determine whether his character
> >concept calles for a value of 12, 13 or 14? It seems to me
> >that the player will often either look at how many points
> >he has to spend (metagame concern) or else he will look for
> >a breakpoint or some other optimizing scheme (also a meta-
> >game concern), whereas in FFRE he will easily be able to
> >express the character concept (Simulation, free of meta-
> >game concerns).
> >
> >My question is, why do you need all those intermediate steps?
> Ah. A good question, and one that I am able to answer.
>
> Feel. You see, in cyberpunk (or post-cyberpunk, as the case may be ;))
> there's this obsession with 'the Edge'. You need to get every possible
> edge to survive. In case of very wide steps, all those tiny actions
> you do to gain a slight advantage are lost in the big, blurry areas
> between steps.
I can get into the mindset.
I also saw something like that in Brian Gleicman's "Age of
Heroes" system. All those tiny little things you could do
that would give you a +1% or +5% bonus.
But from a gaming/simulation perspective it seemed silly to me.
I know, intellectually, that in the long run a +5% or even a
microscopic +1% benefit, *will* benefit me. But emotionally,
intuitively and irrationally, I can't wrap my head around it.
I keep "going why bother".
Perhaps one reason I like FFRE is that the system is not
statistically transparent. In AoH, or in GURPS, you can look
at your character sheet and immediately find your chance of
succeeding at a skill roll. In AoH the skill ratings are given
as percentages (20%, 40%, 60% and so on) and the GURPS 3d6 scale
skills are easy to translate to percentages (rough ones such as
50%, 75%, 90%). FFRE can't be translated that way, because of
the non-linear curves in the multiple-d10 system.
This may cause problems with some players who end up fearing that
the actual probabilities are broken (as in Storyteller) but then
I'll just have to do my best to reassure them that no, the system
is not broken.
Anyway, I think that my inability to get interested in "microEdge"
is the time scale of roleplaying gaming. Roleplaying gaming spends
so much time on each little event and conflict that you rarely
get to resolve many conflicts per session. So a +1% benefit probably
turns out not to be useful.
Brain Gleichman, as a contrast, sometimes runs battles with
hundreds of participating characters (PCs and NPCs). Some of
the veteran Quest FRP GMs have done the same. Todd Richmond once
told about a skirmish he had simulated, with a couple hundred NPCs
per side.
In such largescale battles, the microEdges do become visible.
Maybe it's my playing style that means that there happens to
not be many instances of conflict per session?
It seems to me that I *do* care about microEdge as soon as we
move away from roleplaying gaming. Give me a computer game,
wherein one possible choices gives me a 5% increase in my
monthly or yearly income, and I'll take it. Because the game-
time resolution is so much lower. A gameturn (a month or a year)
is over in a very few minutes.
Real-life examples, where I can save a few dollars, works the same
way as gaming for me. I rarely bother. Because in real life, as
in my style of roleplaying gaming, "the long run" is far away.
Whereas in strategy games I often play 10 or 60 turns an hour,
so "the long run" is close.
> In case of a 0-20 scale, I have a resolution of approximately 5% (OK,
> so it comes somewhere in between 5 and 10% - after all, a 10-sided dice
> is used, but it's open ended, so the resolution is slightly higher than
> 10%, and even though the attributes will rarely go below 8, they might,
> with chemical (combat drugs), cybernetical or, in some cases, psionic
> enhancement go above 20). The combat drugs that characters use do not
> have to be either extremely powerful (an Attribute raise of 1 in a 0-9
> scale is much higher than in a 0-20+ scale) while stile having some
> discernible game effect.
I can understand that.
During play, such improvements can be relevant.
But from a character creation perspective, I fear that it
creates too much fuss. That the player spends billions of
brain cycles on deciding whether to give his character
DX 12, 13 or 14.
Also from an IC-point of view, it creates problems. My
character has DX 11, and your character has DX 12. I, as a
player, knows that your character is more dextrous than mine.
But does my character know this?
In FFRE, if my character has Dex 3 and your character has Dex 4
then it should become obvious after some days or weeks of
physican acquantance. Beyond doubt.
Your system leaves room for doubt and uncertainty.
> >I realize that it's nice to have an average of 10, but it is
> >really necessary to have all those microscale steps of
> >improvement? 11 is slightly better than 10, 12 is slightly
> >better than 11.
>
> It's not microscale - d100 is microscale ;)). That's why I dislike percentile
> scale in RPGs - trying to model reality with a 1% accuracy is too much
> for me.
d100 is nanoscale, 1-20 is microscale.
> [...]
> >I hope, though, that FFRE can support both cinematic and
> >highly realistic campaigns without having to compress or
> >squeeze the scale.
>
> Well, you do have a problem of impossible difficulties - TD 11 in
> FFRE is just Plain Impossible (TD 13 for somebody with a Major Talent).
Actually it's TD 10 and 12, since I use the nonstandard d10-
reading 0=0 convention instead of the standard 0=10.
And apart from that, I don't see a problem.
In other RPG systems, you simply never roll for TD 13+ tasks.
Such as trying to hit a jet aircraft with a thrown stone.
FFRE works the same way, so what's the problem?
> In Pulp IF, there's no limit to how high can one go. It's something
But that's because you apply task difficulty as a modifier to
the skill roll. Which results in Talented characters being
only quantatively better than normals.
In FFRE, a character such as Robin of Locksley would be
qualitatively better than a normal.
> like GURPS Black Ops when it comes to power level, but without the
> tedious seven-hundred-GURPS-points-character bookkeeping. :)
> Since the scale is compressed four-fold, the entire 'normal IF' scale
> fits between 0-4 (ok, 0-5 - skills in IF can get above 20, as a matter of
> fact, they can get to any height, but the price gets kinda steep ;))),
> and the average Pulp IF character (if there's anything like that) has
> Attributes of 5 and skills in the 8-12 range. And can get to, frex, 20
> or higher (which would correspond to a 'normal IF' skill of 80-something).
>
> The sky's the limit. :)
FFRE can't do Superman, I admitted that right from the start
in 1997 and I'll continue to admit it. But it could do Batman
or Spiderman or Doc Savage or Sherlock Holmes or Tarl Cabot
just fine.
> But I like FFRE. It sure seems nice for fantasy campaigns, and I
> like the magic system very much. It just won't scale indefinitively
> as it is now.
I don't need indefinite scale. If it can do a decent simulation
of a dragon or of a Morgoth-sized demigod, then I'm happy.
> However, you might consider having the d10's being open-ended. To avoid the
> Shadowrun problem of 'difficulty of 7 is the same as difficulty of 6' you might
> say that rolling 10 is 9+re-roll.
FFRE has no such problem. In no case will two difficultues yield
the same probability distribution for the same level of skill.
Not even impossible Task Difficulties.
A normal would prefer rolling 4d10 vs TD 12 to rolling 4d10 vs TD 13,
because his risk of Fumbling is higher vs TD 13.
> Make the capability to make open-ended rolls an Advantage that is available
> only in cinematic campaigns, and voila. The difficulties above 10 will still
> drop the number of successess sharply, but FFRE will now scale upwards
> without limits (theoretically ;))).
But if I allow that then Talented characters looses their
ability to be qualitatively better than unTalenteds. Then they
can only be quantatively better. Then any random person could,
with persistence and decades of training, become as good as
a Born archer such as Robin of Locksley.
> I still wouldn't use it for gritty near-future (cyberpunk or SF), but
> then, that's what I keep CORPS for.
I've got my mind set on using FFRE for that exact genre. It
creates a few minor problems with the Faith and Psyche attributes,
but I believe those can be solved by instructing the players not
to "waste too many points" on them.
> >Peter Knutsen
> Leslie
--
Peter Knutsen
Peter Knutsen
John Kim wrote:
>
> Peter Knutsen <pe...@knutsen.dk> wrote:
> >Leszek Karlik wrote:
> >> OK. Why do you think that 'many attributes' approach is more
> >> preferable to 'less attributes' approach?
> [...]
> >In FFRE, you can optimize too, but you can choose to optimize for
> >melee skills, or for ranged combat skills, or for knowledge skills,
> >or for thief skills, or for wizard magic, or for social skills.
>
> OK, I am going to start on a small rant here. The problem
No problem, I'll rant right back at you :-)
> I have seen with nearly all "many-attributes" schemes is that the
> meaning of the attributes becomes increasingly obscure. i.e.
> Rather than making your character good in categories like "ranged
> combat skills" or "thief skills", instead you end up with characters
> who are "very good Perception, fair Reasoning, and above-average
I find the opposite problematic.
Why should you be able to define your character according to
skill performance? That is, why should you be able to define
the direct competence of your character for various skill
groups?
It makes much more sense to define the character according
to general performance (attributes) and then derive the
skill category performance from the attributes.
If you allow character definition according to skill category
performance, then you make possible characters who are Excellent
at Combat but who Suck Deeply at Atlethics.
I have a major problem with that. It strains my willing suspension
of disbelief. Such characters should not be possible to create.
Secondly, if you define your skill category competences directly
then you never get any surprises. This means, most likely, that
your character never gets any surprises either.
Let's take my favourite Quest FRP example. You're making a
mage-type character, and you know that you should give him
very high Magic Talent, high Will and decent Intellect.
Depending on what types of spells you want to learn, Dexterity
can also be useful. Your kind of mage-type wants to learn some
of those spell lists where Dexterity is useful so he takes
Dexterity 16.
Now, these choices were made with spell lists in mind.
Specific spell lists, which the player wants the character
to be able to learn easily.
But the choices influences the learning rate of many other
skills. Security Systems (traps, lockpicking) is one such
skill, Sleight of Hand is another.
Maybe the player doesn't know that his character happens to
have an aptitude for these skills? The character almost
certainly doesn't know. At gamestart. But maybe he finds out.
There was a time when I didn't know that I was good at science.
The school I attended didn't take science seriously, to state
it mildly. Then I moved to a new school, and suddenly I
discovered that this particular subject was easy for me. That
was a lot of fun. It certainly affected my personality and
my general outlook on things.
Your system of directly defining skill category competences
denies characters the chance of such self-discovery.
> Intuition". The game designer then spends time explaining at
> length what "Reasoning" is as opposed to "Intelligence", and even
This is true. I'll have a bitch of a time writing good and
informative definitions of the various attributes (i.e. hard
work which I'm not fond of). But I'll only have to do this
once.
> more time making difficult decisions about whether X skill is more
> Intuition-based or Perception-based.
True except my system doesn't have such strange and hard-to-
grasp attributes as Intuition.
Figuring out the Aptitude formulae is farly easy. I've already
made the design decision that each AF consists of 7 attributes,
always with some doubles and triples and sometimes even
quadruples. (I started out with 6, then I decided that I wanted
7 instead)
For instance, physics is IIIIWWP
Ballroom Dancing is AAAADDP
Tracking is PPPPIII
Meelee Weapon is DDDDAAS
(Note that these keep changing, because I haven't written them
down in my "general notes" document yet. So I haven't defined
them in the final sense yet).
> I find that all this simply obscures making a character.
Some years ago, I decided that I like a little bit of obscuring.
Players doesn't need to, and shouldn't have to, know the exact
Aptitude Formulae for skills.
I'm not saying that they should be kept from the rules, prevented
from seeing them.
Merely that it is enough if a player has a general idea about
what influences what. Dexterity makes you better at meelee and
ranged and unarmed combat, for instance. Knowing such general
relations, they can build their characters.
> In general, when I make a character I think of what he is good at:
> i.e. "he is good at science skills" rather than the abstract
I don't.
I tend to start with basic attributes. Is he smart, strong,
perceptive, agile, large, charismatic, and so on.
Maybe I just don't see people as walking skills. Probably
because I'm not a walking skill myself, being very much a
(n undisciplined) generalist.
> categories like "Perception" as opposed to "Reasoning". In practice
> I am forced to reverse-engineer the system. That is, first I imagine
> the charcter and what she is good at, then I juggle numbers until the
> attributes give me roughly what I am looking for (which is often a
> pretty poor approximation).
But the advantage is that you can't make impossible characters
like "Exceptional at Combat, Sucks Deeply at Athletics".
Also, you can start with a character concept and derive
skills from there.
For instance, I might decide that I'd like to have a character
that is smart and perceptive, but not very charismatic, who
has average willpower and metaphysical stats.
Okay. I try to get inside the head of this young character and
look out through his eyes. What learning opportunities are
there? Which are accessible? Which are not? Which seems to be
promising the most fame, fortune and babes-to-bed? Chances are
he'll try out various tasks and discover that he is good at some,
bad at others, and suck deeply at a few. Maybe he'll find one
task he's really, really good at. Maybe he'll become depressed
because none of the tasks available suits him (I spent more
than half my teenagehood being mildly depressed because I was
in an enviroment where book learning and smarts were defined as
unimportant. I had nothing I was good at, so I lacked, or felt
I lacked, status in the community). He'll pick a career, follow
it for some time. Then either he will complete it (and start
working in it, or decide that he'd like to add a second career
to his portfolio), or he will leave it, either because it seems
to him that another career is more fun or because circumstances
(war, maybe) forces him to change (he gets drafted to the military
and has to learn how to fight).
What you advocate is a kind of top-down design system. You
directly design your skill aptitudes. My approach is bottom-up.
You decide your attributes, and from those aptitudes are derived.
> For example, say I want to represent a sharp-witted urban lawyer
> vs a keen barbarian tracker. They are both perceptive, but in different
The sharp-witted urban lawyer would have Perception 3 or at most 4.
Except if he has a prior career in the police, then he might be 5.
> fields. However, the way most attributes are set up, the lawyer can
> easily become a great tracker -- whereas IMO his background would make
> it very difficult.
The perception you're talking about is not the same as the
Perception attribute in FFRE.
It seems to be mostly a mixture of general smarts (i.e. high
Intellect) and some kind of "social perception" which is
simulated by an Empathy advantage (similar to the one in
GURPS).
Apart from that, Intellect does influence the learning speed
for the Tracking skill, so yes he would be able to learn the
Tracking speed faster than the average person, but not much
faster.
> As another example, I personally am very good at math and
> science, but am not nearly as good in literature and humanities, and
> am poor in foreign languages. Again, this is the sort of difference
Suggests to me that you're an exception from the norm. Most
people, according to psychometricists, tend to do averagely
in all such areas, tend to do well in al such areas, or tend
to do badly in all such areas (contrary to political correct
"psychology").
But fear not, FFRE can simulate you :-)
Just use the Talents system.
Actually, the concept of Talents and Incompetences is a
departure from the line of thought of mine which I've
described above.
A Talent is a skill (or skill-category) bonus, bought directly.
The player simply decides that he wants a Talent (or Major
Talent) with Archery, Wizard Magic, Fire Magic, or Sciences,
then he pays the points and then he gets the Talent directly
(all TDs for those skills are lowered by 1).
No obscurity, no indirectness.
Incompetences are available too. They increase the TD, making
everyday tasks difficult and difficult tasks near-impossible.
They also drastically increase the chances of a fumble. A
character who is Incompetent with Combat skills can learn
to fight, if he wants to, but he'll rarely hit and he will
fumble often, even as he continues to improve his skill.
(a kindly GM may eventually allow him to buy off his
Incompetence, but only in a very cinematic game).
Some skills, such as Languages, are rarely (or never) rolled for.
It doesn't make for a good simulation to roll when people are
talking, or writing, to see how well they do. Instead we (it's
my impression that this goes for nearly all roleplayers) prefer
to have the character always perform at the same level.
So whereas most FFRE skills are actually xd10 (although they
are written merely as x, the d10 being implicit), language
skills are simply x. One character might have Latin 4,
English 2 and German 2, another could have Norwegian 6,
Arabic 4, Latin 4, Russian 3 and English 2.
This is because almost all characters always performs at
the same level.
But characters with a Talent or Incompetence performs at
a range of levels.
Let's assume that you have an Incompetence with Languages
(sorry, FFRE can't let you have an Incompetence with
specific languages, that would be prone to abuse). This
means that if you buy your Spanish skill up to 4 (all
skills starts at 0) then you perform at 3-4, meaning that
you sometimes perform at 3, sometimes at 4 (the GM could
flip a coin).
Or if you have a Major Incompetence, you would perform at
2-4.
Someone with 4 levels of Spanish, and a Talent with
languages, would perform at 4-5. He performs adequately
some of the time, but at other times he performs really
well. He may commit witticisms, use poetic metaphors and
so on, stuff that a normal skill-4 character never does.
But a character who really has skill-5 would be able to
do this all the time, rather than only occasionally.
> that doesn't really show up in abstract categories like "Reasoning".
This sort of difference is mostly a litterary and cinematic
convention, rather than reality.
So originally I didn't feel a need to include such possibilities
in FFRE. Then, one day, the system for Talents struck me, and I
realized that this was a very cool thing. You could be good at
something in two different ways (for optimum effect, of course,
you ought to be good at it *both* ways). Very neat. The reverse
idea, Incompetences, followed a few days or weeks later (it was
years ago, I can't remember the details of it).
So even though it's mostly a cinematic concept, I decided to
include it.
Note that Talents can represent cultural traits, although this is
also cinematic.
Someone raised in a ninja clan could have a Talent for all
Combat and Sneak skills, representing the intensive and
prolonged training given since infancy.
Someone raised among woodsmen or hillsmen could have a Talent
with the Tracking skill, or with Wilderness skills in general.
> -*-*-*-*-*-*-*-*-*-
>
> My two cents is this: Why not have stats based around the
> general skill categories, like Peter's example of thief skills vs
> knowledge skills vs melee skills, etc.? By having categories like
> outdoors skills and so forth (as many logical skill groupings do),
> you can represent a lot of characters much more simply.
But then you open up for miss-correlated categories, which
means that the GM is required to check and approve each
character sheet (rejecting those who are abusive or
unrealistic).
IMO, the rules system itself should ensure that no such
abuses and unrealisms can happen.
> In part, my annoyance is not that the abstract attributes are
> that awful, but that virtually no one tries the alternative approach.
Prism, a freeware system, uses a mixture of the two. Aptitude
comes in part from attributes and in part from skill category
bonuses.
FFRE is the same, in a way, as Prism, except that with the new
Learning Speed formula, attributes have become more powerful.
But what about CORPS? And other skill-tree systems? CORPS
does have attribute-influence, but it's microscopic.
> --
> John H. Kim | Whatever else is true you
--
Peter Knutsen
George W. Harris
wrote:
:
:
:John Kim wrote:
:>
:> Peter Knutsen <pe...@knutsen.dk> wrote:
:> >Leszek Karlik wrote:
:> >> OK. Why do you think that 'many attributes' approach is more
:> >> preferable to 'less attributes' approach?
:> [...]
:> >In FFRE, you can optimize too, but you can choose to optimize for
:> >melee skills, or for ranged combat skills, or for knowledge skills,
:> >or for thief skills, or for wizard magic, or for social skills.
:>
:> OK, I am going to start on a small rant here. The problem
:
:No problem, I'll rant right back at you :-)
:
:> I have seen with nearly all "many-attributes" schemes is that the
:> meaning of the attributes becomes increasingly obscure. i.e.
:> Rather than making your character good in categories like "ranged
:> combat skills" or "thief skills", instead you end up with characters
:> who are "very good Perception, fair Reasoning, and above-average
:
:I find the opposite problematic.
:
:Why should you be able to define your character according to
:skill performance? That is, why should you be able to define
:the direct competence of your character for various skill
:groups?
Because quite frequently those are the terms
in which a character concept is defined.
:It makes much more sense to define the character according
:to general performance (attributes) and then derive the
:skill category performance from the attributes.
This only makes more sense if that's how
players envision characters in the first place; while it
may be true for you, I tend to visualize a character by
the tasks she can perform competently, rather than
general performance attributes. Forcing a player to
start one step further back degrades the ability to
create a character that matches the conception, and
not because the conception is abusive, but simply
because the tools are blunt.
:If you allow character definition according to skill category
:performance, then you make possible characters who are Excellent
:at Combat but who Suck Deeply at Atlethics.
Possibly, depending on the interaction among
skill categories. Is this necessarily a terrible thing? In
any event, it's certainly possible to define skill
categories in such a way that you can define skill
categories separately but nevertheless encourage
some broadening of competence (Frank Perricone's
PRISM skill system provides an (albeit
overcomplicated) example of this.
:I have a major problem with that. It strains my willing suspension
:of disbelief. Such characters should not be possible to create.
I think it's better for a system to allow an overly
broad range of possible character types and rely on GM
judgment than to have an overly-restrictive system that
throws the baby out with the bathwater.
:Secondly, if you define your skill category competences directly
:then you never get any surprises. This means, most likely, that
:your character never gets any surprises either.
I don't think that surprises in character creation
are what people generally game for, and I don't think
eliminating the unexpected and unwanted in character
creation has a negative impact on game play, either; the
same mode of reasoning can be used to justify completely
random character creation as well.
:Let's take my favourite Quest FRP example. You're making a
:mage-type character, and you know that you should give him
:very high Magic Talent, high Will and decent Intellect.
:Depending on what types of spells you want to learn, Dexterity
:can also be useful. Your kind of mage-type wants to learn some
:of those spell lists where Dexterity is useful so he takes
:Dexterity 16.
:
:Now, these choices were made with spell lists in mind.
:Specific spell lists, which the player wants the character
:to be able to learn easily.
:
:But the choices influences the learning rate of many other
:skills. Security Systems (traps, lockpicking) is one such
:skill, Sleight of Hand is another.
:
:Maybe the player doesn't know that his character happens to
:have an aptitude for these skills? The character almost
:certainly doesn't know. At gamestart. But maybe he finds out.
Maybe the player doesn't want his character to
have an aptitude at those skills. Maybe that specific set
of magical aptitudes shouldn't have a nearly perfect
correlation with security systems aptitude. Maybe a little
more flexibility wouldn't be a bad thing.
:There was a time when I didn't know that I was good at science.
:The school I attended didn't take science seriously, to state
:it mildly. Then I moved to a new school, and suddenly I
:discovered that this particular subject was easy for me. That
:was a lot of fun. It certainly affected my personality and
:my general outlook on things.
So maybe character skill aptitudes should be
rolled for randomly the first time they come up in play?
--
"Intelligence is too complex to capture in a single number." -Alfred Binet
George W. Harris For actual email address, replace each 'u' with an 'i'
John Kim
>> i.e. Rather than making your character good in categories like "ranged
>> combat skills" or "thief skills", instead you end up with characters
>
>Why should you be able to define your character according to skill
>performance? That is, why should you be able to define the direct
>competence of your character for various skill groups?
I guess it depends on your approach to character design. In
general, I have a defined character concept which I picture and then
I try to build that character in the rules system. Sometimes I will
base my character design on a real-world person or fictional character.
In this approach, I want to be able to directliy set my character's
competences in a way that fits her conception.
You're right that I hadn't considered the opposite approach,
which is letting the rule system dictate what your character is like
and playing the character that you get. Certainly if one isn't trying
to build a particular conception, then no, my objection doesn't matter.
-*-*-*-*-*-*-*-*-*-
>
>Secondly, if you define your skill category competences directly
>then you never get any surprises. This means, most likely, that
>your character never gets any surprises either.
[...example snipped...]
>Maybe the player doesn't know that his character happens to
>have an aptitude for these skills? The character almost
>certainly doesn't know. At gamestart. But maybe he finds out.
[...]
>Some years ago, I decided that I like a little bit of obscuring.
>Players doesn't need to, and shouldn't have to, know the exact
>Aptitude Formulae for skills.
Yes, that is exactly what I am talking about. As a player
and *especially* as a GM, I don't want to be surprised at what a
character I design is like. That is a flaw in the system, IMO.
I will state outright that I am biased against systems which are
confusing and obscure to the user.
I have frequently seen players who suddenly found out that
their character's competence was different than they imagined
because of obscure influences within the system (like, say,
_Rolemaster_). I have rarely, if ever, seen a player who was
pleased at this. More commonly, the players grumble and ask for
help understanding the system better so they won't make that
mistake next time.
I think the default should be that the player and GM have
accurate knowledge of character's capabilities. If you want the
character surprised, then either (1) the player can choose a
personality open to that, or (2) the GM can explicitly tell the
player that their PC may have differences from what is written.
-*-*-*-*-*-*-*-*-*-
>
>What you advocate is a kind of top-down design system. You
>directly design your skill aptitudes. My approach is bottom-up.
>You decide your attributes, and from those aptitudes are derived.
No, I don't think the analogy fits. My approach does start
at the competences -- but it isn't top-down in that there is no need
to go "down" from the top. Obscure and abstract categories like
"Reasoning" or "Intelligence" aren't (or shouldn't be) neccessary
for play.
They might in theory be useful for enforcing realism -- but
the problem is that the "science" of intelligence and aptitude testing
is extremely unreliable. Thus, I am extremely skeptical of any claims
to realism on this front.
-*-*-*-*-*-*-
>
>Then, one day, the system for Talents struck me, and I realized
>that this was a very cool thing.
[...]
>Someone raised among woodsmen or hillsmen could have a Talent
>with the Tracking skill, or with Wilderness skills in general.
>
>> Why not have stats based around the general skill categories, like
>> Peter's example of thief skills vs knowledge skills vs melee
>> skills, etc.?
>
>But then you open up for miss-correlated categories, which means
>that the GM is required to check and approve each character sheet
>(rejecting those who are abusive or unrealistic).
I don't think that you can enforce realistic characters,
for several reasons. First of all, your claim to realism is based
on the unreliable and problematic field of intelligence and aptitude
testing. Even if you take the "best" of this field, I am highly
skeptical of any solid claims.
Second, you admit outright that you allow a character's
competence to vary from the Attribute base by using Talents. Doesn't
this introduce exactly the sort of mis-correlation that you are
talking about?
Third, I can create an unrealistic character anyway, by
taking a mix of skills or other perks that don't make sense.
I don't think there is any way to get around this. Abusive
character can in theory be ruled out by the system, but that is
just as possible in a skill-category system as in an abstract-
attribute system.
--
John H. Kim | Whatever else is true you
Warren J. Dew
The problem I have seen with nearly all "many-attributes" schemes
is that the meaning of the attributes becomes increasingly obscure.
Agreed ...
In general, when I make a character I think of what he is good at:
i.e. "he is good at science skills" rather than the abstract
categories like "Perception" as opposed to "Reasoning". In practice
I am forced to reverse-engineer the system.
A bit of a tangent, but how is that related to the number of attributes? It
seems to me that it's more an issue of attributes being not directly related to
skills. Even with only the three Fantasy Trip attributes (ST, DX, IQ), if you
want a good archer, you still have to reverse engineer the system to figure out
whether you need high ST, high DX, or both (or for that matter, high IQ, I
suppose).
I think it's more a matter of approach. It seems that you - and Mary Kuhner,
in another thread - have a complete character concept in mind at the start that
you want to play. I, and it seems Peter, would prefer to start with a less
completely defined concept, and see how the character develops in play.
I suspect this may in part be a difference in preference regarding how much
game time passes in a campaign. I like many years to pass - so I'm expecting
my characters to change a lot over the course of play. I'm guessing that the
'complete character concept' approach is better when the characters are
expected to remain relatively static over the course of the campaign. On the
other hand, Mary at least is explicitly trying a long campaign, so I could be
wrong here.
To me, though, this looks like you'd be better off with a skill only system
with no attributes at all.
Why not have stats based around the general skill categories,
like Peter's example of thief skills vs knowledge skills vs
melee skills, etc.? By having categories like outdoors skills
and so forth (as many logical skill groupings do), you can
represent a lot of characters much more simply.
I don't because a lot of other characters, including most of those I know
personally, become impossible.
In helping to coach the MIT ballroom team, I've noticed that physical training
makes a huge difference. Points of technique and choreography, once the
students are physically capable of performing them, are trivial for them to
learn.
For a while, this caused me to think of ballroom as an almost entirely physical
activity. But this isn't true - as I notice when I observe classes in public
ballroom studios, where the students are of average rather than high
intelligence. The average person has quite a bit of trouble understanding what
they are intended to do, even when they are perfectly capable of doing it
physically - even though they are getting explanations, from more experienced
coaches, that are at least as good as mine. For the average student, mental
faculties are often as limiting as physical capabilities.
It's just that MIT students are smart and have good spatial reasoning, so they
can understand these things quicker. Yet they aren't selected for ballroom -
they are selected for ability in math, physics, and the like.
I can't see putting ballroom dance in the same general skill category as math
and physics, especially if you want that category to exclude literature and
humanities. But dropping the correlation would basically prevent players and
their characters from serendipitously discovering, during play, that their
engineers also happen to have an aptitude for ballroom.
John Kim
>> i.e. "he is good at science skills" rather than the abstract
>> categories like "Perception" as opposed to "Reasoning". In practice
>> I am forced to reverse-engineer the system.
>
>A bit of a tangent, but how is that related to the number of attributes?
>It seems to me that it's more an issue of attributes being not directly
>related to skills. Even with only the three Fantasy Trip attributes
>(ST, DX, IQ), if you want a good archer, you still have to reverse
>engineer the system
True, but it's a lot easier to reverse-engineer 3 attributes
than to reverse-engineer 12, I guess. :-)
-*-*-*-*-*-*-*-*-*-
>
>> Why not have stats based around the general skill categories,
>> like Peter's example of thief skills vs knowledge skills vs
>> melee skills, etc.? By having categories like outdoors skills
>> and so forth (as many logical skill groupings do), you can
>> represent a lot of characters much more simply.
>
>I don't because a lot of other characters, including most of those I
>know personally, become impossible.
[...]
>I can't see putting ballroom dance in the same general skill category
>as math and physics, especially if you want that category to exclude
>literature and humanities. But dropping the correlation would basically
>prevent players and their characters from serendipitously discovering,
>during play, that their engineers also happen to have an aptitude for
>ballroom.
I understand your later point -- it is the same point that
Peter made. I don't understand your former point, though. Why
would such characters be impossible? It is trivial to give the
engineer an aptitude for dance but not buy up the skill. OTOH,
I *don't* think that this should be something mandated by the
system (i.e. all math/science types are better at ballroom dance),
because the correlation is just a statistical thing. I know plenty
of good physicists who are not good dancers, for example. I was in
the folk dance group at Fermilab as a graduate student... These
were people who really tried to dance but didn't actually show any
aptitude.
As for your latter point, like Peter, you seem to want
*players* to serendipitously realize during play that their PC
is good at something which they didn't realize (or presumably, is
poor at something they didn't realize). Essentially, you want the
number-crunching to be obscure enough that the player doesn't realize
some important ramifications for her PC.
This seems tricky to me. A math-happy player (like myself,
say) won't be fooled for an instant. I can see what the system does
and am highly unlikely to be surprised by such math tricks. A
math-avoiding player may get a character which is very far from
what he wanted -- too far for him to enjoy it.
-*-*-*-*-*-*-
>
>I think it's more a matter of approach. It seems that you - and Mary
>Kuhner, in another thread - have a complete character concept in mind
>at the start that you want to play. I, and it seems Peter, would prefer
>to start with a less completely defined concept, and see how the
>character develops in play.
I'm not clear on this. Do you mean "develop" in a meta-game
sense (i.e. the character was fuzzily-pictured at first and then
resolves) or "develop" in an in-game sense (i.e. the character didn't
know dance before, but know learns it)?
As for meta-game development, yes, I am mostly Design-At-Start.
On the other hand, I don't see offhand how the abstract-attributes
vs skill-groups helps here. My preferred solution to Develop-In-Play
is to let the player shuffle points around between sessions.
As for in-game development, the main difference seems to be
in the obscuration -- i.e. the point I mentioned before, where some
players are surprised at what their character turns out to be good at.
Personally, I am almost never surprised by math tricks in my character,
because I tend to reverse-engineer the system during character creation
anyhow. I have a hard time picturing players really wanting to be
confused by the math of a system, though. Can you describe more about
this?
Mary K. Kuhner
>On Tue, 06 Mar 2001 18:07:37 +0100, Peter Knutsen <pe...@knutsen.dk>
>wrote:
>:There was a time when I didn't know that I was good at science.
>:The school I attended didn't take science seriously, to state
>:it mildly. Then I moved to a new school, and suddenly I
>:discovered that this particular subject was easy for me. That
>:was a lot of fun. It certainly affected my personality and
>:my general outlook on things.
> So maybe character skill aptitudes should be
>rolled for randomly the first time they come up in play?
My PC Markus turned out to be unexpectedly good at something.
The way this manifested was that I noticed, in the first
scenes where he made fumbling attempts to do the new thing,
that it really "clicked" for him. And, to add a random
factor, he rolled really well. So that's where I put the
development points. I didn't get any bonuses, but still,
the subjective effect was that the character was talented,
surprising both me and him. (And definitely his wife, who
hadn't expected to be married to a psychic vampire.)
For me personally, this is better than surprises driven by
the mechanics. It's one thing not to know your character is
good (or bad) at X. But with mechanical surprises, it's
equally or more likely that you *did* know, put in a million
little details conforming to your intuitive knowledge, and
now you're just wrong and the character portrayal is
inconsistent.
I didn't know I was going to be a biologist until fairly
late in my school career, but in retrospect there were lots of
details consistent with this. I just hadn't realized that
fondness for bugs had anything to do with a career path.
If I'd played a character for some years, and all the little
details were in place--the fondness for bugs, the analytic
puzzle-solving skills, the curiosity, the quick grasp of
logical relationships--but when the character finally took
up the skill she was actively bad at it because of my
lack of understanding of the mechanics system, well, I
don't think I would find that a helpful or entertaining
form of surprise.
But if the game's on a different level of detail maybe this
wouldn't happen, and so the problem wouldn't arise. We tend
to play very high-detail and very slowly.
Mary Kuhner mkku...@eskimo.com
Leszek Karlik
disseminated foul capitalist propaganda:
[...]
>> Well, yes, but that's because GURPS point system is broken.
>Why is it broken?
Well, there's the fact that you can have two characters with the same
set of skills and abilities and one will cost you more and be less
powerful than the second one (the infamous DX 15, lots of 1/2 point skills
character). And then there are various ill-priced advantages from
Supers...
[...]
>GURPS has two faults. It has way too few attributes, and it
>uses a system wherein:
>Skill = (Attribute + Training)
>FFRE will not use that kind of system, instead it uses a
>system similar to Quest FRP, where the attribute determines
>how much experience each level of a skill costs.
But the principle is the same - if you raise your attribute
(for X points), you pay Y points less to raise your skill to
a set competence level.
If X < Y then there is a problem. In GURPS, this condition
is easy to meet.
But no system that has attributes influencing skill price and
a common pool of points to buy both attributes and skills is
completely free from that problem, I'm afraid.
[...]
>> OK. You do have a point there. In Idee Fixe, there's a broad optimization for
>> either physical (dexterity) or intellectual (intelligence) skills, and a
>> narrow optimization done by Ads/Disads.
>I don't like that. It becomes difficult to have a party of more than
>2 characters, since you can't make them different on the most basic
>and primal level (general attributes).
>You have to resort to the secondary (advantages) or tertiary (actual
>skills) to differentiate them, and that's ungood.
Well, that depends on what one considers to be primary level of differentiating
between characters. For some, this will be attributes (he's the guy with
great coordination, and she's the gal with high dexterity), for some
it will be skills (he's great at archery and hunting skills, she's a master
swordfighter), and for some, advantages (my character's the one with combat
reflexes and night vision, her - the one with empathy and high Status).
[...]
>Do you use the Skill = (Attribute + Training) method, or something
>else?
Skill is attribute modified by the skill difficulty plus number of skill
levels (easy skills start at attribute -3, moderate at Attribute -6,
and hard at Attribute -10).
>Anyway your solution doesn't solve the problem, as it then
>becomes flat out impossible to play a character with an
>exceptional attribute - for instance to play one of the ten
>most dextrous men on the planet.
True. This is not extremely cinematic, but then, this mechanic
is not supposed to be for playing such man. Still, I am now
considering changing it, as a matter of fact.
[...]
>> Also, the second part of GURPS syndrome is the use of 3d6 for skill
>> checks. Thus, the difference between a skill of 12 and a skill of 16
>> is a very big one, between 16 and 20 is lower, and 20 and 24 is even
>> smaller yet. This is prone to abuse (as GURPS tends to be).
>How is this part abused?
>It's very useful to have a skill of 18 or 20, or higher, since
>you can still perform well even under adverse conditions.
[...]
>> The testing system in IF is similar to Interlock, only without the
>> Murphy World rule of crit fail on 1 and auto success on 10 (the d10 roll
>> that is added to skill or attribute is open-ended both ways, and only
>> on negative final roll result does one get a critical failure).
>But doesn't this produce a flat critical chance for everyone?
No - critical success is rolling 20 points over the difficulty. Somebody
with a higher skill has a larger chance of scoring a critcal (and,
as a matter of fact, my do so quite reliably for easy tasks, if he
chooses to roll, that is, and not take an automatic success...).
>FFRE is my attempt to get 100% away from the flat probabilities
>of criticals. For each increase in skill, your chance of a
>critical success should go up, and your chance of a critical
>failure should go down. And it should continue and continue and
>continue to go up (down), rather than hit a plateau beyond a
>certain skill rating.
True.
>> So, it is possible to avoid the pitfalls of gurpsification ;) when
>> creating a system with less attributes.
>I don't see how that avoids GURPS syndrome.
>All it does is to prevent players from making characters who
>are exceptional attribute-wise.
Yeah, you're right. I am now thinking about moving to more attributes,
separating attribute and skill points, changing the relationship
between attributes and skills (though using this solution alone still
does not solve the problem) and various other solutions to this problem.
(Of course, I should really be studying for my American History exam which
I take on Friday. Sigh. UseNet addiction.)
>A couple of months ago I tried to model myself in GURPS, but I
>gave up again. Even with an IQ modestly pegged at 14 I got
>ridiculously high levels in skills that I had invested microscopic
>(default) or minimal effort(0.5 CP) in.
>It could be an interesting experiment, once FFRE is completed,
>to actually finish my GURPS writeup and then do a similar writeup
>in FFRE, in order to compare the two systems and show (prove?)
>that FFRE is lightyears ahead.
Well, GURPS is an, ahem, rather old system. :)
[...]
>> >FFRE, on the other hand, can not be abused.
>> I think this statement is rather optimistic. :)
>Only because you're infected with the meme that all systems
>are equally good and equally easy to abuse.
Now now, I never said that, did I? :>
Not all systems are equally easy to abuse. All systems _CAN_ be
abused. The degree of possible abuse is variable.
>The reality is that some systems are better than others, and
>that some systems may very well be abuse-proof.
I guess that's what
[...]
>And please note that I do not consider "optimization for a
>specified area of endavour" to be abuse. It is not abusing the
>system to make character creation choices such that your
>character is optimized for meelee combat, or archery, or
>sneaking, or wizard magic.
OK, how is about 'superiority of one concept over another'?
If 'talented character with moderate training' has the same effectiveness
as 'grizzled veteran with lots of points', and is cheaper (because the
character spent more points on attributes and thus could get to the
same level of competence as the veteran for an overall price that is
lower), that constitutes abuse, IMO.
It's rather hard to make a system that:
a) has a connection between attributes and skill price (learning speed, IOW)
b) allows one to customize character by assigning variable amount of
points to either attributes or skills
and
c) does not have the abovementioned problem.
One can just make abusing the system _harder_ to do.
>It is only abuse if you get "edge" too cheaply, which usually means
>that you can gain much optimization for a broad area of endavour
>(skills period) for a low cost (such as in GURPS).
>Other people may have a broader definition of the term "abuse".
Yes. I think that if a player who knows the system better can make
a better character then somebody who knows it less, that constitues
abuse.
[...]
>I haven't thought much about it, though, but it seems to me that
>many of the GURPS advantages are only necessary because of the
>low number of attributes.
>
>So the more of the character you can express on the primary level,
>the attribute level, the better.
Yeah, for example, splitting each of the four attributes into two
sub-attributes was done just to eliminate many of the advantages
and disadvantages.
The problem with IF mechanics is that it was created before I
joined the team, and I couldn't just throw it out through the
window and begin anew. You did give me some good arguments, though,
so I'll try and convince the others to move to more attributes
and perhaps a split AP/SP system.
[...]
>Characters with advantages will still exist, be legal and be
>necessary. Anybody with above-average quality of appearance
>(Attractive, Very Handsome, Mindbogglingly Beautiful) needs
>an Advantage.
This could be an attribute, you know. :)
[degree of precision of simulation]
>> Feel. You see, in cyberpunk (or post-cyberpunk, as the case may be ;))
>> there's this obsession with 'the Edge'. You need to get every possible
>> edge to survive. In case of very wide steps, all those tiny actions
>> you do to gain a slight advantage are lost in the big, blurry areas
>> between steps.
>
>I can get into the mindset.
>I also saw something like that in Brian Gleicman's "Age of
>Heroes" system. All those tiny little things you could do
>that would give you a +1% or +5% bonus.
>
>But from a gaming/simulation perspective it seemed silly to me.
>
>I know, intellectually, that in the long run a +5% or even a
>microscopic +1% benefit, *will* benefit me. But emotionally,
>intuitively and irrationally, I can't wrap my head around it.
>I keep "going why bother".
Well, from a simulation perspective neglecting all those small
benefits because they won't yield a modifier is... welll.. an artifact
of the gaming system?
Oh well, different strokes I guess.
[...]
>Perhaps one reason I like FFRE is that the system is not
>statistically transparent.
It's just like Shadowrun. It's not statistically transparent
without an appropriate table. Somebody did crank such a table
for Shadowrun a long time ago, when I was subscribed to the
ShadowRN list, using a spreadsheet.
>FFRE can't be translated that way, because of the non-linear curves in the
>multiple-d10 system.
That might be a problem for the Game Master, you know.
[...]
>But from a character creation perspective, I fear that it
>creates too much fuss. That the player spends billions of
>brain cycles on deciding whether to give his character
>DX 12, 13 or 14.
Well, yes, that is a problem. But that is a one we chose to
accept because of specific reasons (ie. the consistency with
the setting).
>Also from an IC-point of view, it creates problems. My
>character has DX 11, and your character has DX 12. I, as a
>player, knows that your character is more dextrous than mine.
>But does my character know this?
>In FFRE, if my character has Dex 3 and your character has Dex 4
>then it should become obvious after some days or weeks of
>physican acquantance. Beyond doubt.
>Your system leaves room for doubt and uncertainty.
Yeah. That's the point. My brother and I, we practiced martial arts
together. I don't know who was more dextrous, me or him. We were kinda
even - sometimes I won, sometimes he did. There was doubt and
uncertainity.
And it fits in with the style of the setting.
(Now, pulp IF, with it's compressed scale, leaves no room for
uncertainity - the scale is even tighter than in FFRE, with 'normal'
humans compressed to 0-5 scale. It also fits with the style of
the setting. :)))
[...]
>d100 is nanoscale, 1-20 is microscale.
OK, a matter of nomenclature I guess.
[...]
>> Well, you do have a problem of impossible difficulties - TD 11 in
>> FFRE is just Plain Impossible (TD 13 for somebody with a Major Talent).
>
>Actually it's TD 10 and 12, since I use the nonstandard d10-
>reading 0=0 convention instead of the standard 0=10.
Oh, yeah.
>And apart from that, I don't see a problem.
>
>In other RPG systems, you simply never roll for TD 13+ tasks.
>Such as trying to hit a jet aircraft with a thrown stone.
>FFRE works the same way, so what's the problem?
Well, but I don't feel that TD 12, for example, is like trying to hit
a jet aircraft with a thrown stone. It's more like dodging bullets, to
use a cinematic example. :>
OK, let me elaborate.
We have Robin of Locksley. He is shooting at a distant target, in a bad
lighting. TD of 11, but with his Major Talent it goes down to 9.
He rolls his Archery skill of, IIRC, 12. 011223455689. A hit.
Now, he also is shooting at a distant target, this time in plain
daylight, but with a crosswind. Again, TD of 11, and this time he
rolls 00223577899. A hit, with 2S. So, confidently, he reaches for
another arrow and... well.. the sun has just set. Distant target, TD 10,
crosswind, TD 11, poor lighting, TD 12. Ooops. Even Locksley can't do that.
There's a _sharp_ drop in hit probability, to a nice round 0.
>> In Pulp IF, there's no limit to how high can one go. It's something
>But that's because you apply task difficulty as a modifier to
>the skill roll. Which results in Talented characters being
>only quantatively better than normals.
No, task difficulty is the target number (OK, not directly, TN is based
on how much the the task difficulty is higher than the skill). This is
because there is no upper limit to skill, and the difficulty of task
one can accomplish is dependant on how high is the skill.
>In FFRE, a character such as Robin of Locksley would be
>qualitatively better than a normal.
Than make the open-ededness a result of Talent.
Now the poor Robin is mainly quantatively better than normals
(not because of his skill, but because of his Talent - the sharp, two
point drop in TD does wonders for the number of successes at medium
and low difficulties - crunch the numbers and you'll see), and
his qualitative advantage only allows him to barely succeed at a
slighly more difficult tasks than the normals are capable of.
If you say that, for example, Talent allows one to re-roll zeroes
(or nines, in your case) when rolling against TD of 10+ _once_, and
Major Talent allows one to re-roll the nines _again_, then Robin of
Locksley will not be 80% quantatively and 20% qualitatively better
than normal humans (his main advantage being the much higher number
of successes he gets due to lower TDs), but he will be 100% qualitatively
better (he shoots like a normal archer, but can do things normal
archers cannot do - like reach TD 15).
And if you use both options (Major Talent being both a minus to TD and
the ability to re-roll for TDs that are still higher than 9) then
Robin will be not 'slightly qualitatively better', ie. able to compensate
for 2 more points of bad condition modifiers, he will be considerably
qualitatively better.
[...]
>> Make the capability to make open-ended rolls an Advantage that is available
>> only in cinematic campaigns, and voila. The difficulties above 10 will still
>> drop the number of successess sharply, but FFRE will now scale upwards
>> without limits (theoretically ;))).
>
>But if I allow that then Talented characters looses their
>ability to be qualitatively better than unTalenteds. Then they
>can only be quantatively better. Then any random person could,
>with persistence and decades of training, become as good as
>a Born archer such as Robin of Locksley.
No, why? This random persons shoots as well as Robin in normal conditions
(well, one can't get closer to the bullseye than the exact centre of it),
but when the J. Random Bowyer shoots at a flying dragon in the midst of a
raging storm, using only the light from the dragon's fiery breath, he'll
have absolutely no chance at all of succeeding. And Robin of Locksley will.
Using the system of -2 to TD, Robin will shoot better than J. Random Bowyer
in normal conditions, (OK, with TDs above 3), but will be able to do only
slighly better when it comes to shootin in adverse conditions.
Neelakantan Krishnaswami
>
> As for your latter point, like Peter, you seem to want *players* to
> serendipitously realize during play that their PC is good at
> something which they didn't realize (or presumably, is poor at
> something they didn't realize). Essentially, you want the
> number-crunching to be obscure enough that the player doesn't
> realize some important ramifications for her PC.
>
> This seems tricky to me. A math-happy player (like myself, say)
> won't be fooled for an instant. I can see what the system does and
> am highly unlikely to be surprised by such math tricks. A
> math-avoiding player may get a character which is very far from what
> he wanted -- too far for him to enjoy it.
It's possible to keep some of the mechanics secret, so that the
players and characters have to learn about the interrelationships
through trial and error. I'm going to do this for my Kult/Continuum
crossover. The PCs will begin ignorant of the magical structure of the
universe, and I think that letting the players discover the inter-
relationships as the PCs do will potentially be more fun. (If I'm
wrong I can always infodump, of course.)
Neel
Warren J. Dew
determine attributes when they aren't specifically defined in terms of skill
groups:
True, but it's a lot easier to reverse-engineer 3 attributes
than to reverse-engineer 12, I guess. :-)
That's a good point. I personally don't have an objection to people reverse
engineering a point based system to get the character they want. I'm curious
about Peter's opinion, here, though. Peter, do you specifically want to
prevent reverse engineering the system so that there are no surprises in
character definition?
Regarding having attributes based on general skill categories making people I
know impossible to generate:
I don't understand your former point, though. Why
would such characters be impossible? It is trivial to give the
engineer an aptitude for dance but not buy up the skill.
Because that "general skill category" would include a lot more than just
ballroom dance, and would include things that the people I'm thinking of
wouldn't have an aptitude for. For example, if it were 'performing arts', the
the system would say that these people would also have a knack for acting and
singing, but in fact many of them don't. Even if the category were just
'dance', some are bad at rote memorization, and would thus be bad at line
dancing. What makes them better at ballroom dance than at these other skills
is their mental faculties - the same spatial reasoning abilities that make them
good engineers.
OTOH, I *don't* think that this should be something mandated by the
system (i.e. all math/science types are better at ballroom dance),
because the correlation is just a statistical thing.
Speaking as a coach of the top University ballroom team in New England, I
assure you it is not "just a statistical thing". Smarter people learn ballroom
faster than less smart people, other things being equal. That's why MIT and
Harvard are the #1 and #2 teams in the area. If it were "just a statistical
thing", these schools would be handicapped by not having as large a student
body to draw from.
Other things are, of course, not generally equal; engineers with weak feet make
poor competitive ballroom dancers, because physical capabilities limit one's
performance. That's why MIT is #1 and Harvard only #2: MIT's training regimen
includes foot strength exercises that improve balance, so those who don't start
with the agility and balance to become competent competitors can get those
attributes by coming to class regularly.
I know plenty of good physicists who are not good dancers, for
example. I was in the folk dance group at Fermilab as a graduate
student... These were people who really tried to dance but didn't
actually show any aptitude.
Which just goes to show that ballroom dance is different from these other forms
of dance, and is an illustration of why the different dance forms should not be
part of the same general skill group. For what it's worth, my admittedly
uninformed impression of folk dances is that they are intended to be easy to
pick up, and thus don't require more than minimal spatial reasoning to
understand the basic movements. This is very different from ballroom,
particularly competitive ballroom, where the movements and technique are more
complex, because peak performance is more important than broad participation.
As for your latter point, like Peter, you seem to want
*players* to serendipitously realize during play that their PC
is good at something which they didn't realize
Yes.
(or presumably, is poor at something they didn't realize).
I'm not sure what you mean by this. Can you clarify?
Essentially, you want the number-crunching to be obscure enough that
the player doesn't realize some important ramifications for her PC.
No. I don't have a problem with players realizing ramifications the player
considers important; what I'm interested in are the ramifications the player
considers unimportant.
For example, the player just wants to run an MIT freshman. This requires the
character to have a very good high school math and science background, and
implies an aptitude for these things - high IQ , say, and maybe a point or two
in a couple skills. The player randomly places the leftover points into DX.
The player hasn't even thought about dancing - or if she has, she looked up the
rules on ballet and concluded the character would make a mediocre ballerino,
since he'd need a much higher DX (though not nearly as high an IQ) to succeed
in ballet.
Now, after getting to MIT, the character happens to try dancing with the
ballroom dance club. Both the player and character find that, serendipitously,
this particular combination - extremely high IQ and reasonable DX - happens to
be perfect to generate a very good beginner ballroom dancer. The character,
finding that he really enjoys this activity - after all, he's a lot better than
expected with a minimal initial investment - pours all his effort into the
skill. A few years later, he finds himself, with his partner, the U.S. amateur
champion; he turns pro, and eventually makes it into the top professional ranks
in the U.S.
This all happens despite the fact that neither the player nor the character
ever considered the character as a dancer until well after the character was
designed - they both expected the character to end up as an engineer or a
physicist. (And yes, this example does describe someone I know.)
This seems tricky to me. A math-happy player (like myself, say)
won't be fooled for an instant. I can see what the system does
and am highly unlikely to be surprised by such math tricks. A
math-avoiding player may get a character which is very far from
what he wanted -- too far for him to enjoy it.
I don't have a problem with your being able to design the character to your
liking. If you were the player in the above example, you could also take the
character and make it into your initial conception, perhaps eventually turning
him into a Nobel laureate physicist, for example. I'm even in favor of making
the system sufficiently easy to reverse engineer that the math avoiding player
can still have his Nobel laureate, and avoid turning the character into a
ballroom dancer instead.
My problem is that if you define attributes in terms of skill groups, then
you'd know in advance the types of skills your character would be good at. It
wouldn't be possible to get the above character, because with high Scientific
Aptitude (instead of IQ) and only a reasonable Indoor Athletics Aptitude
(instead of DX), the character would never be more than an adequately competent
ballroom dancer, because the points in Scientific Aptitude would be wasted in
things not obviously related to science.
As for in-game development, the main difference seems to be
in the obscuration -- i.e. the point I mentioned before, where
some players are surprised at what their character turns out
to be good at.
Hopefully the above will explain how what I want differs from obscuration. If
not, I can try to explain further.
Steve Mading
: But no system that has attributes influencing skill price and
: a common pool of points to buy both attributes and skills is
: completely free from that problem, I'm afraid.
Yes. I wonder how to fix this problem. Simply enforcing two
seperate pools is unacceptable, because it makes *sense* to have
some characters be ill-trained but have high natural talent, while
others have lots of training, but little natural talent.
Hmm... here's an idea: Assign each skill a "natural factor" and
a "learned factor". When you add natural to learned for
a skill you must get N (I'll set N=10 for this off-the-top-of
my-head discussion, but it might be better to pick something
else, I don't know). So, while one skill might have a natural
factor of 2 and a learned factor of 8, another might have a natural
factor of 8 and a learned factor of 2, while another might have them
balanced at 5 and 5, and so on. To determine what the skill's value is,
you multiply your attribute by the natural factor, then multiply
your training ranks by the "learned factor", then add the two together.
Thus, when skills have a higher natural factor, they don't benefit
much from learning the skill, either you 'got it', or you don't.
For example, no matter how many points you dump into the skill,
if you have a bad agility you'll never get an excellent jump skill,
and conversely, if you never spend any points on the skill, you'll
never be able to pick up hyperspace astrogation by just 'winging
it', even if you are super-smart.
Here's an example, let's say it's a system with percentile skills,
where attrributes typically range from 0 to 9 and skill ranks typically
range from 0 to 9 (but could get more at really high levels). Let's
say you have a strength of 3, and an intelligence of 8 (and some other
attributes I'm not bothering to go into). Let's say you want to
learn 'mace'. Lets say that "mace" has a natural factor of 8 and a
learned factor of 2, indicating that it depends more on natural
ability (strength) than learning. So then you put 9 ranks of
training into mace. Then your mace skill ends up being:
str * natural_factor + rank * learning_factor ==
3 * 8 + 9 * 2 = 24 + 18 = 42 %
Thus, even though you spent 9 ranks on mace, you still only have a
skill of 42%, because your strength is so low, and mace is a skill
that depends more on natural ability than time spent learning.
Now let's take an opposite example, where learning matters more.
Let's say you want to know hyperspace astrogation, and you
have the same stats as above (int = 8). But this time, hyperspace
astrogation is a skill with a natural factor of 2 and a learning
factor of 8. Let's say you put only 1 rank into it:
int * 2 + rank * 8 = 8*2 + 1*8 = 24 %
This is a skill where you just can't get a good percentage by intuition.
you *have* to have the ranks in training or your high int doesn't help.
Now, my examples above assuemd a percentile system. But the same
type of idea could be done on some other sort of scale. The key
thing is to make sure that the typical ranks in a skill and the
typical attribute ranges are on approximately the same scale.
(So if a X is an average attribute score, then X ranks is an average
amount to put in a skill), and then to make sure that the scale
in which a skill check is made is roughly N times as large as
the scale in which the attributes and ranks fit. (Where N is the
total that all natural and learned factors have to add to in your
skills.)
Neelakantan Krishnaswami
No, because most of the campaign is still in the planning stages.
I'll probably ask them in a couple of months.
> I think that doing a campaign wherein the players discover how magic
> works could be fun, given the right group of players. But normally
> I would not do such a thing. I'm 100% pro rules visibility.
Usually, so am I. The players are good about finding holes in the
rules because they tend to have to operate within stricter limits
than the GM.
But for this game, the rules reflect the game world too closely for
full transparency to be feasible. For example, the different varieties
of magic all vary according to the personalities of the immortals of
Metropolis who empower them. There's just too much world-detail embodied
in the rules.
Neel
Peter Knutsen
John Kim wrote:
>
> Peter Knutsen <pe...@knutsen.dk> wrote:
> >John Kim wrote:
> >> The problem I have seen with nearly all "many-attributes" schemes is
> >> that the meaning of the attributes becomes increasingly obscure.
> >> i.e. Rather than making your character good in categories like "ranged
> >> combat skills" or "thief skills", instead you end up with characters
> >> who are "very good Perception, fair Reasoning, etc.
> >
> >Why should you be able to define your character according to skill
> >performance? That is, why should you be able to define the direct
> >competence of your character for various skill groups?
>
> I guess it depends on your approach to character design. In
> general, I have a defined character concept which I picture and then
> I try to build that character in the rules system. Sometimes I will
> base my character design on a real-world person or fictional character.
> In this approach, I want to be able to directliy set my character's
> competences in a way that fits her conception.
George says something similar, so I suppose that this attitude
is not as uncommon as I've previously assumed. Until now, I've
taken it for granted that roleplayers are very willing to think
in terms of attributes (since attributes are such widespread
concept in roleplaying gaming systems).
> You're right that I hadn't considered the opposite approach,
> which is letting the rule system dictate what your character is like
> and playing the character that you get. Certainly if one isn't trying
> to build a particular conception, then no, my objection doesn't matter.
I don't like random character creation. The fact that you play
a character that you've yourself created is an important aspect
of roleplaying gaming, as far as I'm concerned.
But you can still design your character by setting attribute
values from which aptitudes are derived. It requires that you
either try to think about how attributes may be linked to
skill aptitudes (and involves the danger that you might dis-
agree with the RPG system designer), but I don't think that
it should be very difficult to analyze your way to the same
conclusions I've reached - perhaps the most difficult step is
to accept and embrace the notion of attributes (also, the
danger of assumption clash is, here, not too high).
Alternatively, you can read the rule book. It requires minimal
effort to open the rules document and read the aptitude formula
for any skill.
Or ask the GM. He's there to help you.
> -*-*-*-*-*-*-*-*-*-
> >
> >Secondly, if you define your skill category competences directly
> >then you never get any surprises. This means, most likely, that
> >your character never gets any surprises either.
> [...example snipped...]
> >Maybe the player doesn't know that his character happens to
> >have an aptitude for these skills? The character almost
> >certainly doesn't know. At gamestart. But maybe he finds out.
> [...]
> >Some years ago, I decided that I like a little bit of obscuring.
> >Players doesn't need to, and shouldn't have to, know the exact
> >Aptitude Formulae for skills.
>
> Yes, that is exactly what I am talking about. As a player
> and *especially* as a GM, I don't want to be surprised at what a
> character I design is like. That is a flaw in the system, IMO.
> I will state outright that I am biased against systems which are
> confusing and obscure to the user.
We're talking about very small amounts of obscurity here.
GURPS is totally non-obscure. It requires no effort to find
out the linkage between attributes and any given skill.
Quest FRP and FFRE are only very slightly obscure. You have
to spend a few milliseconds thinking and understanding and
reading a line of rules.
But since they are also (both Quest FRP and FFRE) reasonable
systems, you don't need to penetrate the incredibly thin veil
of obscurity. You just need to read the short document that
tells you the meaning of each attribute, and then you need
to accept and embrace the notion that the value of your
attributes impacts the speed with which you learn skills.
Once you've done that you can intuitively define attriute
values that will yield fairly good skill aptitudes.
You won't get perfect optimization, only good optimization,
but in FFRE the difference between perfect and good optimization
is that you leanr 10% (or in grave cases 20%) faster, which is
almost a trivial difference. Unlike GURPS wherein the difference
is that a perfectly optimized character can learn 500% or 2000%
faster than a merely "good optimized" character.
And as I wrote to George, there are no surprises in the areas
that you are concerned about and intersted in. If you want to
make an archer/woodsman character, then you can and will
tailor the aptitudes for the relevant skills, by defining
attribute values.
Only those areas which you spend no time on might surprise you.
> I have frequently seen players who suddenly found out that
> their character's competence was different than they imagined
Note the word "imagined". If you imagine that your character is
good at shooting a bow, hiding in bushes, sneaking around in
the forst, finding and followin tracks, then you can make him
good at this with a trivial amount of effort.
But if I am making a magician who specializes in element magic
and body-enhancement magic, I have probably not been inte-
rested in his aptitude for Dodging or Archery or Camouflage.
This means that I and/or the character might some day discover
what his aptitude is. Or I may have read the aptitude formula
for those skills and thus know in advance what his aptitude is.
Or he may have been a boy scout in his preteen years and thus
know what his aptitude for such activities are.
But in no way is it random. The aptitude is derived from the
attributes. So it is knowable. If you want to know it. You
might not. Either due to laziness or a mild phobia of rules
(don't want to read more rules than you have to) or some desire
to make it easier for yourself to do firewalling (my character
Dennis doesn't know what his aptitude for Sneaking is, because
he has never tried it, so I don't want to know it either).
I think you overlooked the words "little bit" when I wrote that
I like a little bit of obscurity.
I may also have written an insufficient description. English is
my second language and I occasionally make mistakes by using
words that are stronger or weaker than what I actually mean.
A second try: I find completely transparent systems annoying.
They make optimization easy. Also complete transparency is
(often or always?) caused by simplicity (GURPS is completely
transparent because the linkage between skill and attribute
is extremely simplistic) and I prefer more complex and detailed
and realistic systems. This makes the system not completely
transparent (you have to perform work, although not much, to
see throug it) and it happens to be a side effect that I like.
> because of obscure influences within the system (like, say,
> _Rolemaster_). I have rarely, if ever, seen a player who was
I'm not familiar with Rolemaster, can you give an example?
> pleased at this. More commonly, the players grumble and ask for
> help understanding the system better so they won't make that
> mistake next time.
There should not be any surprises in the areas of competence
that you care about. Only in those you don't care about.
Only in those that are not central to your character concept.
Also keep in mind that surprises are often nothing more than
a 10% or 50% higher or lower Learning Speed.
The surprise might be that you discover that your archer/
woodsman can learn some new skill 50% faster than a normal
person. Or 20% slower.
Seriously, that is no big deal.
Your character can learn his key skills 2 to 5 times faster
than the average person. 100% to 400% faster (the later is
only the case if you give him really high values in the
relevant attributes, though).
> I think the default should be that the player and GM have
> accurate knowledge of character's capabilities. If you want the
The player can get accurate and perfect knowledge. He can read
the rules. But he may not want to. Chances are he will care
about some competences and not care about others - he's
disinterested. His Learning Speed for Computer Operation may
be 0.2 or 3.5, he doesn't care. Computer Operation is not a
part of his character concept. (He may need to learn the skill
later, though).
I'm not pleased at the possibility of a GM, who likes to
hide the rules from the player, using FFRE, but it might
happen.
> character surprised, then either (1) the player can choose a
> personality open to that, or (2) the GM can explicitly tell the
> player that their PC may have differences from what is written.
I'd never do (2). I'm perfectly okay with players doing (1),
however.
> -*-*-*-*-*-*-*-*-*-
> >
> >What you advocate is a kind of top-down design system. You
> >directly design your skill aptitudes. My approach is bottom-up.
> >You decide your attributes, and from those aptitudes are derived.
>
> No, I don't think the analogy fits. My approach does start
> at the competences -- but it isn't top-down in that there is no need
> to go "down" from the top. Obscure and abstract categories like
> "Reasoning" or "Intelligence" aren't (or shouldn't be) neccessary
> for play.
But they aren't obscure, they're real.
Some people are more intelligent than others. Some people ar
more agile than others. Some people are fitter than others.
Some people are more charismatic than others.
> They might in theory be useful for enforcing realism -- but
> the problem is that the "science" of intelligence and aptitude testing
> is extremely unreliable. Thus, I am extremely skeptical of any claims
> to realism on this front.
The problem is that many of the people who do research in
the area of intelligence are not scientific but political.
Howard Garder, for instance, is motivated by left-oriented
beliefs that everybody are equal. He is greatly disturbed
by the evidence that one general intelligence has a huge
effect on the socioecomic status of people (since this can
be used to justify a society in which some people are richer
than others) and thus he invents a hyppthesis of multiple
"intelligences" out of thin air, and it becomes very popular
among laypeople because it is politically correct.
The Bell Curve guys (Hernstein and something) may also be
politically motivated (wanting to justify a society in
which the difference betwee rich and poor is large, and
in which the rich gives very little aid to the poor).
The issue is confused because there is also a large subset of
the population (primarily in the USA - the notion is very
uncommon in Denmark) that would welcome such results without
skepticism. Also many so-called IQ tests (but not all) are
very bad, measuring aquired skills (language, math), rather
than innate skills (pattern recognition). The Bell Curve
people may have based their research on people measured by
such bad tests.
I'm not sure, however. Some discussions I've had with a
Norwegian psychologist, and a lecture by a top-class Danish
psychometricist (Helmut Nyborg) has lead me to believe that
there is in fact a very strong g-factor (general intelligence,
as opposed to specific and uncorrelated intelligences) at
play.
A system such as GURPS is thus eminently realistic, with IQ
being a great simulation of the g-factor. But it makes for
bad play. IQ is cheap compared to what it gets you, and it's
an enormous advantage to have a high IQ stat.
You get a better game, plus you are better able to simulate
genre conventions (the brilliant scientist who doesn't notice
what goes on around him) if you have many mental attributes.
> -*-*-*-*-*-*-
> >
> >Then, one day, the system for Talents struck me, and I realized
> >that this was a very cool thing.
> [...]
> >Someone raised among woodsmen or hillsmen could have a Talent
> >with the Tracking skill, or with Wilderness skills in general.
> >
> >> Why not have stats based around the general skill categories, like
> >> Peter's example of thief skills vs knowledge skills vs melee
> >> skills, etc.?
> >
> >But then you open up for miss-correlated categories, which means
> >that the GM is required to check and approve each character sheet
> >(rejecting those who are abusive or unrealistic).
>
> I don't think that you can enforce realistic characters,
> for several reasons. First of all, your claim to realism is based
> on the unreliable and problematic field of intelligence and aptitude
> testing. Even if you take the "best" of this field, I am highly
> skeptical of any solid claims.
I'm not basing FFRE on psychometric research. If I did that I'd
end up with something like GURPS. (Besides, I'm a bit lazy. Chances
are I'll read eventually Arthur Jensen's "The g-factor" and then
maybe some Howard Garder just out of curiosity, but mostly I'll
stick to psychometricists and experimental psychologists (such
as Eyseneck) and avoid the idle thinkers or at least not take
them seriously. And since I don't need to read those books
at all I'll probably not do it this year).
Rather I'm basing it on the ability to simulate genre conventions
without those genre conventions being forced. For instance it is
possible within the rules-as-written to make a character who
has Intelligence 8 and Perception 1. But individual GMs can
change those rules if they want to avoid such unrealistic
characters. The blunt solution is for the GM to declare that
players must always take the same value in PER as in INT.
A subtler solution is to declare that players who have a
non-correlation must purchase an "Unusual Character" advantage
that gets more expensive the larger the non-correlation is.
So the system starts out unrealistic but it's quite easy to
enforce realism (at least in this particular area).
Also it seems to me that Perception is easier to train than
Intelligence.
>From a realism point of view, I'd say that you'd need hard
and intensive dedicated (not incidental) training for
many years to raise INT by +1 point, and that a further
increase than that is flat out impossible.
But I'd say that increasing Perception is easier. If your
city-dwelling character suffers a plane crash and have to
live among native indians, and learn their skills, then he
may very well end up improving his PER, maybe from 2 to 5
or from 3 to 5, after having lived as a hunter (in a very
dangerous enviroment) for a couple of years.
So I leave this open to the GM. There's one set of rules,
very defined (as opposed to vague) and then a list of
optional tweaks that the GM can consider, and implicit
permission for the GM to introduce tweaks of his own.
So I'm basing the skill aptitudes on common sense, rather than on
scientific research.
> Second, you admit outright that you allow a character's
> competence to vary from the Attribute base by using Talents. Doesn't
> this introduce exactly the sort of mis-correlation that you are
> talking about?
Yes it does.
But individual GMs can choose to disallow talents.
Either disallow them completely, or else limit them a lot.
One possibility is that the average condition is that all Humans
are Incompetent at Wizard magic. This means that they must
roll against a higher target, which means a lower chance of
being successfull and a higher chance of fumbling and a higher
chance of the really bad fumbles. Which boils down to Humans
not really wanting to learn Wizard magic. But perhaps some
Humans have a "gift" of not being Incompetent with Wizard magic.
They are not Talented (except relative to the Human norm) but
they are not handicapped either. These may be titled (in a
societal sense) "Mages" or something. Or perhaps all Humans
(without exception) are Incompetent at Wizard magic but
Elves are not. It is possible to use Incompetences without
using Talents at all.
Another possibility is that Incompetences can simulate vows.
To simulate an AD&D-style Cleric, you could give him the
disadvantage "Vow: never use sharp weapons". But you could
use the more mechanical solution (the player is not required
to "roleplay the disadvantage") of giving him an Incompetence
with sharp weapons. That way, the character can pick up and
use any weapon, but if it's a sharp one then he will be so
nervous that his entire body is shaking and he can't use it
very well.
Intensive training can also be simulated by Talents. The GM
may rule that no character can have an inborn Talent, but if
the character had intensive childhood training in something
he can have a Talent. Thus members of a Ninja clan may have
any or all Talents in combat- and thief-skills, while non-
members simply cannot have those talents.
Or the GM can decide not to restrict things, and go with the
default rules which is that characters may take Talents and
Incompetences without restriction.
Anyway, a Talent overrides neither your skill nor your
attribute-derived aptitude. It merely enhances them.
A character with a Major Talent in archery and a skill of 3
will rarely miss, but he won't get in the lethal hits often.
He will, however, get very few fumbles.
A character with a skill of 6 and no Talent is better off,
at that snapshot in time. Not, obviously, if the Talented
character keeps training. But if his aptitude is only normal
(he may have Dexterity 3, Strength 3 and Perception 3) then
it will take him a while to improve his skill. The skill 6
character may have a higher aptitude.
So it's several ways for a character to be good at something.
One realistic (high aptitude) and one unrealistic. And they're
synergetic, so for optimal result you should take both.
> Third, I can create an unrealistic character anyway, by
> taking a mix of skills or other perks that don't make sense.
That's a subjective question.
Lots of people will say that my skill list doesn't make
sense. But that's because they're career-oriented or
think in terms of character classes. My list of skills
makes perfect sense to me.
What I mean is that skill choice is a simulation of child-
hood and youth (and adulthood for older characters) choices,
and that if the aptitudes are derived in a proper and realistic
fashion then the GM is not needed to opiniate on those choices.
The fact that it is hard for your character to learn Bicycling
will almost ensure that he doesn't learn it. Or maybe he did.
Maybe he decided to spend years learning what the neighbour
kids could learn in a few months. No problem. He paid dearly
for the skill, so let him have it.
> I don't think there is any way to get around this. Abusive
> character can in theory be ruled out by the system, but that is
> just as possible in a skill-category system as in an abstract-
> attribute system.
No, there's something you have overlooked.
A direct aptitude system has a direct link.
Any given skill links back to only one aptitude.
This means that you can have the Athletics-1 Combat-10 character.
In a derived aptitude system, there is instead a web of links.
Sword links back to Dexterity (much), Strength (much) and Agility
(some).
Bow links back to Dexterity (much), Strength (some) and Perception
(some).
This web makes it difficult, perhaps impossible, to make
unrealistic characters.
For instance you cannot make a character who turns out to
have a very high aptitude for combat but a very low aptitude
for athletic sports (unless you employ Talents and Incompetenes
but here I'm talking solely about attributes).
Peter Knutsen
John Kim wrote:
>
> Warren J. Dew <psych...@aol.com> wrote:
> >John Kim posts, in part:
> >> In general, when I make a character I think of what he is good at:
> >> i.e. "he is good at science skills" rather than the abstract
> >> categories like "Perception" as opposed to "Reasoning". In practice
> >> I am forced to reverse-engineer the system.
> >
> >A bit of a tangent, but how is that related to the number of attributes?
> >It seems to me that it's more an issue of attributes being not directly
> >related to skills. Even with only the three Fantasy Trip attributes
> >(ST, DX, IQ), if you want a good archer, you still have to reverse
> >engineer the system
>
> True, but it's a lot easier to reverse-engineer 3 attributes
> than to reverse-engineer 12, I guess. :-)
It is. No doubt.
> -*-*-*-*-*-*-*-*-*-
> >
> >> Why not have stats based around the general skill categories,
> >> like Peter's example of thief skills vs knowledge skills vs
> >> melee skills, etc.? By having categories like outdoors skills
> >> and so forth (as many logical skill groupings do), you can
> >> represent a lot of characters much more simply.
> >
> >I don't because a lot of other characters, including most of those I
> >know personally, become impossible.
> [...]
> >I can't see putting ballroom dance in the same general skill category
> >as math and physics, especially if you want that category to exclude
> >literature and humanities. But dropping the correlation would basically
> >prevent players and their characters from serendipitously discovering,
> >during play, that their engineers also happen to have an aptitude for
> >ballroom.
>
> I understand your later point -- it is the same point that
> Peter made. I don't understand your former point, though. Why
> would such characters be impossible? It is trivial to give the
> engineer an aptitude for dance but not buy up the skill. OTOH,
> I *don't* think that this should be something mandated by the
> system (i.e. all math/science types are better at ballroom dance),
> because the correlation is just a statistical thing. I know plenty
But "just a statistical thing" means that 90% of all math/
science types are better at ballroom dancing than the average
person. You can't dismiss that. It's a significant difference,
not something you can dismiss.
The math/science type who is inept at learning how to dance
(note: at learning how to dance, not: at dancing) represents
the unsual, the exception from the norm, and thus it is both
reasonable and most economic to require him to take a Trait
(in this case a Disadvantage) rather than to require the
vast majority of math/science types fo take a Trait (in
this case an Advantage).
> of good physicists who are not good dancers, for example. I was in
> the folk dance group at Fermilab as a graduate student... These
> were people who really tried to dance but didn't actually show any
> aptitude.
I'll leave Warren to reply to that, since I don't know a
lot about dancing.
> As for your latter point, like Peter, you seem to want
> *players* to serendipitously realize during play that their PC
> is good at something which they didn't realize (or presumably, is
> poor at something they didn't realize). Essentially, you want the
> number-crunching to be obscure enough that the player doesn't realize
> some important ramifications for her PC.
I'm not 100% sure if Warren agrees with me here, but I'm only
interested in the player serendipitously discovering *un* -
important ramifications.
I assume that the player has been busy with the *important*
ramifications during character creation. Or rather those that
he feels are important. (External circumstances may later force
the feeble old priest to learn to fight with a quarterstaff,
and other such happenings that the player should not be
expected to foresee).
I also assume that the GM will step in and give advice if a
player makes a somewhat reckless choice, such as giving his
character below-average values (all 2's) in Size, Fitness
and Hardiness. Such a character would be very easy to inca-
pacitate and kill, compared to a normal person (3/3/3) and
the player should be warned about this and given a chance
to raise at least a few of the attributes (I'd raise
Hardiness first).
> This seems tricky to me. A math-happy player (like myself,
> say) won't be fooled for an instant. I can see what the system does
> and am highly unlikely to be surprised by such math tricks. A
Fine with me. You're welcome to read all the rule books.
If there's something you don't understand, I'm willing to
help you to understand it, by explaining it in a different way
(if I can) and giving examples.
The rules are your friend, not your enemy.
> math-avoiding player may get a character which is very far from
> what he wanted -- too far for him to enjoy it.
A math-avoiding player is welcome to ask the GM for
assistance, either in the form of advice, or else in the form
of the GM being present while the player makes the character,
or else in the form of the player writing down a description
of the character and then the GM tries to create that character
within the constraints of the point budget.
> -*-*-*-*-*-*-
> >
> >I think it's more a matter of approach. It seems that you - and Mary
> >Kuhner, in another thread - have a complete character concept in mind
> >at the start that you want to play. I, and it seems Peter, would prefer
> >to start with a less completely defined concept, and see how the
> >character develops in play.
>
> I'm not clear on this. Do you mean "develop" in a meta-game
> sense (i.e. the character was fuzzily-pictured at first and then
> resolves) or "develop" in an in-game sense (i.e. the character didn't
> know dance before, but know learns it)?
Good question.
> As for meta-game development, yes, I am mostly Design-At-Start.
> On the other hand, I don't see offhand how the abstract-attributes
> vs skill-groups helps here. My preferred solution to Develop-In-Play
> is to let the player shuffle points around between sessions.
Yes, that's a sensible DIP solution. Particularly if it can
be done without having to retcon.
> As for in-game development, the main difference seems to be
> in the obscuration -- i.e. the point I mentioned before, where some
> players are surprised at what their character turns out to be good at.
Thin, mist-like obscuration. Penetrable if you make just a
minimal effort.
> Personally, I am almost never surprised by math tricks in my character,
> because I tend to reverse-engineer the system during character creation
> anyhow. I have a hard time picturing players really wanting to be
> confused by the math of a system, though. Can you describe more about
> this?
I don't want to confused, but I want the system to be complex
(as opposed to simple) because simple is always unrealistic
and makes character differentiation hard. The microscopic
amount of obscuration that this creates is a beneficial side
effect, to me.
Peter Knutsen
"Warren J. Dew" wrote:
>
> John Kim posts, in part:
>
> The problem I have seen with nearly all "many-attributes" schemes
> is that the meaning of the attributes becomes increasingly obscure.
>
> Agreed ...
Yes, that's true. It's easier to understand GURPS because it has
only 4 attributes, whereas FFRE has 12.
But is it easier to understand Quest FRP which has 9 attribute
than FFRE which has 12? I'm not sure here.
Also, there's gross oversimplification in the systems with
few attributes. From a realism point of view, GURPS IQ doesn't
bother me (but from a game blance point of view it bothers me
so much that I'll never GM GURPS), but the unified Dexterity
is very problematic. Both in GURPS and AD&D/D&D. Why can't I
choose the narrower ability of having manual dexterity while
bodily agility is lowered? GURPS won't let me do that (I can
be more dextrous but it doesn't affect combat skills, not
even ranged weapons) and it will not let me reduce my agility.
AD&D/D&D is even worse. (I did mention my PBEM Rogue earlier,
didn't I? The one who had 16 DEX but who whould have been much
better served by a split of AGI 8/DEX 18).
> In general, when I make a character I think of what he is good at:
> i.e. "he is good at science skills" rather than the abstract
> categories like "Perception" as opposed to "Reasoning". In practice
> I am forced to reverse-engineer the system.
>
> A bit of a tangent, but how is that related to the number of attributes? It
> seems to me that it's more an issue of attributes being not directly related to
> skills. Even with only the three Fantasy Trip attributes (ST, DX, IQ), if you
> want a good archer, you still have to reverse engineer the system to figure out
> whether you need high ST, high DX, or both (or for that matter, high IQ, I
> suppose).
True.
But Johh Kim has a point in that it's very easy (so easy that
I won't hesitate to call it effortless) to reverse engineer
GURPS, whereas it is not effortless to reverse engineer
Quest FRP or FFRE. (Note, however, that I'm not saying it's
difficult or taxing to reverse engineeer FFRE. Merely that it
requires a little bit of effort as opposed to no effort).
> I think it's more a matter of approach. It seems that you - and Mary Kuhner,
> in another thread - have a complete character concept in mind at the start that
> you want to play. I, and it seems Peter, would prefer to start with a less
> completely defined concept, and see how the character develops in play.
Not really. I like to define my character prior to gamestart.
At least the mechanical parts. Personality starts out loosely
defined and becomes sharper after I've played some sessions.
But there may be ares of my character that I am not intersted in.
Areas where I don't mind how it turns out. "I'm creating a wizard,
why should I care about his ability to learn Brawling?" for
instance.
I would not mind discovering that his aptitude for learning
Brawling was much lower (or much higher) than that of the
average person.
But I'd want to be ensured that his aptitude for Wizardly
things is quite high.
> I suspect this may in part be a difference in preference regarding how much
> game time passes in a campaign. I like many years to pass - so I'm expecting
> my characters to change a lot over the course of play. I'm guessing that the
I think it would be fruitfull to differentaite between game
mechanical change (or what one might label performanmce change,
almost always performance increase) and psychological change.
> 'complete character concept' approach is better when the characters are
> expected to remain relatively static over the course of the campaign. On the
> other hand, Mary at least is explicitly trying a long campaign, so I could be
> wrong here.
I don't know much about Mary's campigns, but it might be the
case that her characters tends to be very static when it
comes to game mechanics/capability/performance. Maybe they
only change on the psychological level?
As GM, I like to start out with "beginner" PCs because it is
problematic for me that a very competent character has a very
brief and shallow background history. For instance a 150 CP
GURPS character with a 3 page backstory. It badly hurts my
suspension of disbelief that this *formidable* individual
has not already defeated many enemies and done many heroic
and/or difficult deeds and has travelled over most of a
continent.
As player, I tend to write long backstories (what limits me is
that I don't know much about the world, and often that the GM
is unwilling to work with me to enable me to write a longer
backstory) so I don't see it as a problem if I am to play a
very competent character.
But starting from scratch can be fun too, provided the character
has an adeqate survival capability.
> To me, though, this looks like you'd be better off with a skill only system
> with no attributes at all.
John Kim?
You might be right.
[...]
> Warren J. Dew
--
Peter Knutsen
Warren J. Dew
A math-avoiding player is welcome to ask the GM for
assistance, either in the form of advice, or else in the form
of the GM being present while the player makes the character,
or else in the form of the player writing down a description
of the character and then the GM tries to create that character
within the constraints of the point budget.
What if the gamesmaster is math-avoiding as well?
Peter Knutsen
Steve Mading wrote:
>
> Leszek Karlik <les...@ideefixe.nom.pl> wrote:
>
> : But no system that has attributes influencing skill price and
> : a common pool of points to buy both attributes and skills is
> : completely free from that problem, I'm afraid.
>
> Yes. I wonder how to fix this problem. Simply enforcing two
> seperate pools is unacceptable, because it makes *sense* to have
> some characters be ill-trained but have high natural talent, while
> others have lots of training, but little natural talent.
You can implement a system for transferring points from one pool
to another. Just without a 1:1 transfer rate.
I tried to do something like that in Quest FRP (one of many
house rules that I've created) and it seems to work okay,
although the character creation system really isn't geared to
anything but "newbie adventurers" (who start out highly
talented and with few skills).
The FFRE solution might be labeled a very elaborate system
for transfering points between pools.
> Hmm... here's an idea: Assign each skill a "natural factor" and
I'm sorry but it's not a *good* idea.
> a "learned factor". When you add natural to learned for
> a skill you must get N (I'll set N=10 for this off-the-top-of
> my-head discussion, but it might be better to pick something
> else, I don't know). So, while one skill might have a natural
Assuming your idea is good, N=10 is a good choice.
You're saying that an untrained person, even if highly intelligent,
cannot perform well here.
So far so good.
The problem is that people, under your system, still improve at
the same speed.
Beatrice has an Intelligence of 10 and Chris an Intelligence of 1.
This means that their default skill ratings are 20% and 2%, which
boils own to (both!) being completely useless. If they try to
astrogate they'll end up in a black hole. No problem so far.
But then they enter Astrogator school and spend 10 years there,
each spending the same amount of time in classes and on homework.
This gives them both 9 levels of training.
When they emerge Beatrice has a skill of 92% and Chris has a skill
of 74%.
And my suspension of disbelief evaporates instantly.
I need people with high aptitude to learn faster than people
with normal aptitude. I need people with low aptitude to
learn slower than normal people. Chris is severely retarded.
He needs constant care to ensure that he's kept healthy and
well fed. Beatrice is a genius, probably smarter than all of
us rgfa'ers added together.
Beatrice should emerge, after having spent the same amount of
effort as Chris, with a skill at least 10 times higher.
Preferably 30 or 100 times higher.
If you don't believe me then it's because you've never
met and interacted with a person who is severely retarded.
> you *have* to have the ranks in training or your high int doesn't help.
Yes but it's wrong that getting the ranks requires the same
effort for everyone.
Also, your Mace example is problematic in that only strength
plays a role. What about hand-eye coordination? The Mace
skill is presumably about hitting something.
> Now, my examples above assuemd a percentile system. But the same
> type of idea could be done on some other sort of scale. The key
> thing is to make sure that the typical ranks in a skill and the
> typical attribute ranges are on approximately the same scale.
> (So if a X is an average attribute score, then X ranks is an average
> amount to put in a skill), and then to make sure that the scale
> in which a skill check is made is roughly N times as large as
> the scale in which the attributes and ranks fit. (Where N is the
> total that all natural and learned factors have to add to in your
> skills.)
I'm not playing if Chris can learn at the same speed as
Beatrice. It's beyond absurdity.
--
Peter Knutsen
Warren J. Dew
You should, of course, be able to choose. Because these ones
are important for your character and for your character
concept. You are intersted in those aspects of your character.
You care about them (these aspects). Randomizing them would
be criminal, it might mean that your character becomes
unplayable.
Bad reasoning.
You are proposing a system where people who have a skill based concept have to
do extra work, and might not get exactly what they want, just in order to get
more varied, and thus more believable, skill landscape. For a few people, this
might make their characters unplayable, because the system's rules might not
allow them to create exactly the character they want - for example, the great
swordsman who is terrible at athletics. It's a tradeoff.
However, depending on the skills that are important in your campaign, there
will be particular attribute landscapes that are favored. Certain attributes
will tend to correlate with other attributes, because they happen to affect the
same skills. For some, more picky than you, this might be enough to negatively
affect suspension of disbelief.
Randomized attributes are a solution to this; they can be designed to produce
exactly the statistical distribution that the game designer thinks is correct.
However, it makes it even more likely that the player won't get to play exactly
the character that they want. It just pushes the tradeoff further in that
direction.
I actually think there are advantages, as well as disadvantages, to preventing
players from playing exactly the character concepts they want, forcing them to
play characters that aren't something that they would initially have thought
of. Some of my favorite characters are ones that initially didn't fit what I
wanted to play. And it does tend to extend the range of characters that one
can enjoyably roleplay.
From another post:
But since they are also (both Quest FRP and FFRE) reasonable
systems, you don't need to penetrate the incredibly thin veil
of obscurity. You just need to read the short document that
tells you the meaning of each attribute, and then you need
to accept and embrace the notion that the value of your
attributes impacts the speed with which you learn skills.
I think this is a little naive. It works for you with FFRE because you
designed the system; it works for you with Quest perhaps because you share
assumptions with the system designer. Elsewhere in your post, you admit that
the FFRE system is less realistic than GURPS because it aims to satisfy genre
conventions rather than just player world reality; since genre convention
knowledge is not shared among all possible players, I think you'll get quite a
few players for whom reading the attribute descriptions will result in quite a
different notion of what skills are affected than is actually written into the
game rules.
Warren J. Dew
I also tend to define characters in terms of skill sets rather
than attributes. I don't think of real people in terms of
attributes, so why would I do that for a character? ... I'm
unlikely to tell you "Susan has a very high IQ and willpower,
average strength and slightly below-average agility" ... These
are certainly things that affect who they are, but they don't
define who they are.
I'm curious: do you think of real people in terms of attributes,
or only characters?
I'm not who you asked, but I definitely think of real people in terms of
attributes and not in terms of skills. When I think of my friends and
acquaintances, the most important differentiators are attributes like
intelligence or agility/balance. The next most important differentiators are
personality, gender, and physical dimensions. Skills are a very distant third.
Partly this is because I've found that skills have been a very poor
differentiator in my circle of acquaintances. None of my friends has a job
related to the subject of their college degree. All have picked up new skills
as necessary to perform well at new jobs. Only one or two have ever actually
had to work at picking up a new skill; in the vast majority of cases, going
from total lack of knowledge to competency at desired new skills has been an
effortless process barely impinging on the consciousness.
In contrast, attributes remain fairly constant. If I were to order the
intelligence or strength of the dozen or so people I've known for 20 years or
more, I'd say there have been only two or three changes, each of which was a
minor swap in position between people who were close in the first place. They
are more constant than personality; while the basic personality seems pretty
constant, the current personality seems to vary a fair amount as a function of
current job satisfaction.
Mary K. Kuhner
stats and then working through the consequences of those stats
doesn't work well for me.
I think there are two basic reasons.
One is that my own working model of the world doesn't go like this,
so I encounter assumption clash problems. I note that my bright,
self-disciplined students who do well in Classical Genetics and
Molecular Genetics sometimes bomb in Population Genetics. These
are closely related topics, and yet there are substantial differences
among my students in whether they "get it". I saw the same with
advanced algebra, geometry, trig and calculus; I found algebra very
easy and calculus quite tractable, but had a bad time with
geometry. About 1/2 of my classmates were the reverse.
So going from "Intelligence 4, Willpower 5" to "she is a population
geneticist" doesn't really happen for me. I agree that general
intelligence and willpower are important, but the quirky skill-
specific stuff has a lot more to do with the way I personally experience
the world.
I don't think I'm a population geneticist because I'm smart and
(sort of) self-disciplined; I think I have some specific fit with that
topic, and the smarts are just enabling. So my own perception of
myself has much more to do with specific skill categories than
general attributes. This is also the way I tend to see other people.
I use categories like "knows his history really well" and "fluent
public speaker" a lot more in my thoughts than "smart" or "willful."
The other problem is that attributes (even 12 of them) seem like
awfully blunt tools. They emphasize the commonality of characters.
All seven of my current PCs are bright people, though the ten-year-
old has some definite intellectual limitations due to his age.
I'm not sure I'd stat any of the six adults differently at all for
Intelligence. They'd vary in Willpower, but four of the six would
still be the same.
I happen to know that those six adults are very different people,
with different intellectual competancies, different aptitudes,
different ways of approaching problems, different areas of interest.
If I start with their skill sets (which vary widely) I see them
as different first, and can then note that in some underlying
respect they are similar (all bright). If I start with their stats,
I am forced to start out thinking of them as similar. I find
this drastically unhelpful if I have to play all six at once.
I don't want to focus on commonalities, because that will lead me
to have either characters too poorly differentiated to allow multi-
PC play, or freaks with wildly variable stats (to get some
differentiation). It is better for me to focus on skills, where
wide dissimilarity is not freakish.
This is probably exacerbated because the campaign provides little
use for the physical stats, so the visible differences among the
PCs would boil down to mainly just the mental ones.
Peter asks why I think very slow paced campaigns emphasize this
problem. I think surprises become more problematic if you have
a very detailed, day-to-day character depiction ("how come we
didn't notice she was quick/slow at picking up X? She picked
up some of the subskills onstage very badly/well.") Skill learning
is less likely to be a matter of "I take 4 years offstage to
learn biology" and more likely to be a matter of "Having been
studying shielding on- and off-stage for six months, I take a
skill increase in it." In other words, there are more detailed
non-mechanical ways to depict learning, so the chance that they
will clash with th mechanical ones is greater.
I think I would fight with Peter over the details of his system,
and this would be distracting for both of us, so I personally
prefer not to try to systematize so much. For one thing, in his
given example I am deeply unconvinced that two people, of very
different intelligence, with degrees in Biology have the same
skills. I am also deeply unconvinced that the dumber person's
problem is uniform across all sub-topics and all years of
study. I don't think I could make a better system; I just don't
think that a flawed system is superior, in this area, to
player judgement.
But you don't have to please everybody to make a good rules
system, and shouldn't even try. I am unlikely to enjoy a highly
mechanical system no matter how you make it, so I'm not in the
target audience for one anyway.
Mary Kuhner mkku...@genetics.washington.edu
Peter Knutsen
Leszek Karlik wrote:
>
> On Tue, 06 Mar 2001 17:11:51 +0100, Peter Knutsen <pe...@knutsen.dk>
> disseminated foul capitalist propaganda:
>
> [...]
> >> Well, yes, but that's because GURPS point system is broken.
> >Why is it broken?
> Well, there's the fact that you can have two characters with the same
> set of skills and abilities and one will cost you more and be less
> powerful than the second one (the infamous DX 15, lots of 1/2 point skills
> character). And then there are various ill-priced advantages from
> Supers...
>
> [...]
> >GURPS has two faults. It has way too few attributes, and it
> >uses a system wherein:
> >Skill = (Attribute + Training)
> >FFRE will not use that kind of system, instead it uses a
> >system similar to Quest FRP, where the attribute determines
> >how much experience each level of a skill costs.
>
> But the principle is the same - if you raise your attribute
> (for X points), you pay Y points less to raise your skill to
> a set competence level.
The effect is still there, yes. If your attribute is higher
than it becomes cheaper to raise the skill. Yes.
But in FFRE the skill starts from zero. In GURPS your attribute
gives you a head-start.
Also in GURPS the "cheapening" effect is absurdly large. In
FFRE it's small.
Finally, in GURPS extreme optimization requires no effort.
In FFRE you have to spend two minutes reading some rules.
Plus you have to optimize in a more focused fashion, for
specific skills rather than for just skills in general.
> If X < Y then there is a problem. In GURPS, this condition
> is easy to meet.
In FFRE it's harder to meet. But it's there, since it would be
extremely unrealistic for characters to all learn at the same
speed.
> But no system that has attributes influencing skill price and
> a common pool of points to buy both attributes and skills is
> completely free from that problem, I'm afraid.
True but what's the alternative?
I try to make optimization expensive (in terms of time it
requires, and in terms of cost-benefit ratio), hoping to
raise the treshhold so that fewer players will bother.
> [...]
> >> OK. You do have a point there. In Idee Fixe, there's a broad optimization for
> >> either physical (dexterity) or intellectual (intelligence) skills, and a
> >> narrow optimization done by Ads/Disads.
> >I don't like that. It becomes difficult to have a party of more than
> >2 characters, since you can't make them different on the most basic
> >and primal level (general attributes).
> >You have to resort to the secondary (advantages) or tertiary (actual
> >skills) to differentiate them, and that's ungood.
>
> Well, that depends on what one considers to be primary level of differentiating
> between characters. For some, this will be attributes (he's the guy with
> great coordination, and she's the gal with high dexterity), for some
> it will be skills (he's great at archery and hunting skills, she's a master
> swordfighter), and for some, advantages (my character's the one with combat
> reflexes and night vision, her - the one with empathy and high Status).
I think it's obvious that attributes are more primary than
skills because they change much less often, because they're
inborn.
A character may leave a friend, and meet him again some years
later, and the friend can then have gone from having no (or
no remarkable) archery skill to being a master archer.
Actually, advantages change even less often an attributes,
since attributes can all be trained (except Size), while
most advantages are inborn.
But advantages are also exceptions. Every character has a
Strength score and a Willpower score and a Charisma score.
So it makes sense to define attributes as the basic level
of character differentiation.
> [...]
> >Do you use the Skill = (Attribute + Training) method, or something
> >else?
>
> Skill is attribute modified by the skill difficulty plus number of skill
> levels (easy skills start at attribute -3, moderate at Attribute -6,
> and hard at Attribute -10).
So your answer is yes.
> >Anyway your solution doesn't solve the problem, as it then
> >becomes flat out impossible to play a character with an
> >exceptional attribute - for instance to play one of the ten
> >most dextrous men on the planet.
>
> True. This is not extremely cinematic, but then, this mechanic
> is not supposed to be for playing such man. Still, I am now
> considering changing it, as a matter of fact.
I'd like the potential "character space" of FFRE to be as
large as possible.
But that's a specific design goal, and it's not necessary
that you strive for the same design goal of your system.
Because it may be possible that there's a price of designing
a system in which a huge space of different characters are
possible, that this involves from disadvantage which is
not (necessarily) found in a system that allows a smaller
space of potential characters.
Apart from that, I'd like to point out that I'm not talking
about cinematics, I'm talking about realism here.
The experience of playing a character that is as smart as a human
can be, in GURPS, is seriously unrealistic.
I'd like to give players a realistic experience of playing such
a super-bright character in FFRE. Yes, being extremely smart
makes some things easier, but they don't make them easier by
many orders of magnitude such as will have it.
If you say that it's not realistic to play a character who is
as smart as it's possible for a human to be (I guess the best
example would be Leonardo daVinci) or as agile as it's
possible for a human to be (Bruce Lee) then you're using the
word "realism" very differently from the way I'm using it.
Leonardo daVinci existed. Bruce Lee existed. Heck, I can do
some pretty amazing things and yet I exist. I am *realistic*.
If a system claims to be realistic them it should be able
to simulate Leo, Bruce and I.
There's a perverted branch of literary criticism that labels
any fiction, which portrays exceptional characters, as "unrealistic"
or "escapism". I do not hesitate to call this branch of criticism
perverted because it is obviously sick, since it refuses to
acknowledge the existence of *real* people who can do exceptional
things.
> >> So, it is possible to avoid the pitfalls of gurpsification ;) when
> >> creating a system with less attributes.
> >I don't see how that avoids GURPS syndrome.
> >All it does is to prevent players from making characters who
> >are exceptional attribute-wise.
>
> Yeah, you're right. I am now thinking about moving to more attributes,
> separating attribute and skill points, changing the relationship
> between attributes and skills (though using this solution alone still
> does not solve the problem) and various other solutions to this problem.
I guess I've made you reconsider your design choices. I should
probably be proud of that :-)
> >A couple of months ago I tried to model myself in GURPS, but I
> >gave up again. Even with an IQ modestly pegged at 14 I got
> >ridiculously high levels in skills that I had invested microscopic
> >(default) or minimal effort(0.5 CP) in.
>
> >It could be an interesting experiment, once FFRE is completed,
> >to actually finish my GURPS writeup and then do a similar writeup
> >in FFRE, in order to compare the two systems and show (prove?)
> >that FFRE is lightyears ahead.
>
> Well, GURPS is an, ahem, rather old system. :)
"3rd Edition revised" is only a couple of years old :-)
> [...]
> >> >FFRE, on the other hand, can not be abused.
> >> I think this statement is rather optimistic. :)
> >Only because you're infected with the meme that all systems
> >are equally good and equally easy to abuse.
>
> Now now, I never said that, did I? :>
True. I'm making offensive guesses :-)
> Not all systems are equally easy to abuse. All systems _CAN_ be
> abused. The degree of possible abuse is variable.
I'm not willing to acknowledge that, because I've never seen
anybody abuse Quest FRP. I've heard of a case in which the
system was abused but that was because the GM was incompetent
(a player had given his character, with 12 being the human
average), 3 in Strength, 3 in Agility, 3 in Dexterity, 3 in
Constitution, 3 in Leadership, in order to be able to have
high values in the other attributes) and failed to simulate
the consequences of the character (a character with such
low physical attributes would function very differently from
a normal human, in terms of such simple tasks as movement
and fighting off the ever present bacteria and vira - and
parasites if the adventuring is in a cave or wilderness).
> >The reality is that some systems are better than others, and
> >that some systems may very well be abuse-proof.
> I guess that's what
Incomplete sentence?
> [...]
> >And please note that I do not consider "optimization for a
> >specified area of endavour" to be abuse. It is not abusing the
> >system to make character creation choices such that your
> >character is optimized for meelee combat, or archery, or
> >sneaking, or wizard magic.
>
> OK, how is about 'superiority of one concept over another'?
Good point, I find that as problematic as you do.
> If 'talented character with moderate training' has the same effectiveness
> as 'grizzled veteran with lots of points', and is cheaper (because the
> character spent more points on attributes and thus could get to the
> same level of competence as the veteran for an overall price that is
> lower), that constitutes abuse, IMO.
I agree with that.
I doubt if I can remove the difference (in cheapness and
effectiveness) completely in FFRE, but I hope to reduce it
vastly vompared to GURPS.
Hopefully reduce it so much that it becomes tolerable for
everyone.
> It's rather hard to make a system that:
> a) has a connection between attributes and skill price (learning speed, IOW)
> b) allows one to customize character by assigning variable amount of
> points to either attributes or skills
> and
> c) does not have the abovementioned problem.
True.
> One can just make abusing the system _harder_ to do.
Maybe you're right.
I intend to make abuse hard, in FFRE.
If I succeed really, really well (as in 5S) then I might
end up making it impossible :-)
> >It is only abuse if you get "edge" too cheaply, which usually means
> >that you can gain much optimization for a broad area of endavour
> >(skills period) for a low cost (such as in GURPS).
> >Other people may have a broader definition of the term "abuse".
>
> Yes. I think that if a player who knows the system better can make
> a better character then somebody who knows it less, that constitues
> abuse.
Depends on how much better "better" is.
If a complete newbie to the system, who has understood the
attribte list (=he knows the difference between Agility and
Dexterity, for instance) and embraced the notion that
attributes helps you learn skills faster/easier, makes a character
that is 10% or 15% more expensive than a character made by a
system expert who has gone to great lengths to optimize, then
I don't think it's a problem. Because it's only a 10% or 15%
difference.
In GURPS, the difference is often one of many hundred percent.
This is where the problem is.
> [...]
> >I haven't thought much about it, though, but it seems to me that
> >many of the GURPS advantages are only necessary because of the
> >low number of attributes.
> >
> >So the more of the character you can express on the primary level,
> >the attribute level, the better.
>
> Yeah, for example, splitting each of the four attributes into two
> sub-attributes was done just to eliminate many of the advantages
> and disadvantages.
Good move!
> The problem with IF mechanics is that it was created before I
> joined the team, and I couldn't just throw it out through the
> window and begin anew. You did give me some good arguments, though,
> so I'll try and convince the others to move to more attributes
> and perhaps a split AP/SP system.
Cool :-)
> [...]
> >Characters with advantages will still exist, be legal and be
> >necessary. Anybody with above-average quality of appearance
> >(Attractive, Very Handsome, Mindbogglingly Beautiful) needs
> >an Advantage.
>
> This could be an attribute, you know. :)
It probably will be an extrinsic ("fifth category") attribute.
It will be extrinsic because it is trivially easy to alter it,
using Illusion magic, Shapechange magic or plastic surgery.
I could probably do something such as
3 being Normal
4 being Attractive (similar to the GURPS advantage)
5 being Beautiful/Handsome (again similar to GURPS in spirit)
6 being Very B/H
7 being Incredibly B/H
8 being Mindbogglingly B/H
with lower-than-3 representing reduced attractiveness.
A similar trait scale could exist for Social Status, Legal
Status, Wealth and maybe others as well.
> >I know, intellectually, that in the long run a +5% or even a
> >microscopic +1% benefit, *will* benefit me. But emotionally,
> >intuitively and irrationally, I can't wrap my head around it.
> >I keep "going why bother".
>
> Well, from a simulation perspective neglecting all those small
> benefits because they won't yield a modifier is... welll.. an artifact
> of the gaming system?
I agree that it's bad simulation, in a way.
But I can't get myself motivated to bother. Sorry :-)
> Oh well, different strokes I guess.
In part, it's a question of practicability. I dislike
numbers that doesn't affect things much.
Once, when I was playing a friend's homebrew system my character
(an Elf warrior) had an IQ attribute on the character sheet byt
my friend was unable to explain what the IQ attribute meant.
I was tempted to ask for his permission to take an eraser and
erase the number. Because while I like to have a character
sheet that is filled with numeric information (FFRE will be
*stuffed* with numbers such as General Reflexes, Body Reflexes,
Mental Reflexes, Blood, Sanity, Wound Treshhold.............)
I dislike having my character sheet cluttered with *useless*
information.
> [...]
> >FFRE can't be translated that way, because of the non-linear curves in the
> >multiple-d10 system.
>
> That might be a problem for the Game Master, you know.
Only if you're of the opinon that the GM should control things.
I'm not. When I GM, I don't feel a need to have much control
over a lot of things. As long as I can keep the cultural level
of the game world on track (=simulating the fact that if the PCs
tries to introduce foreign memes then the world culture will
resist this to the best of its ability) and ensure realism
I'm happy. And the more decisions I can delegate to mechanics,
the less work I have to do so the easier my duty becomes.
Some GMs may feel, for odd reasons, a need to be able to set things
up so that PC Gabriel has exactly a 55% chance of hitting the
pear with his hand crossbow. But I don't understand that mentality.
Why do you need to know the exact probability? It should be
enough to know the difficulty of the task (whether it's TD 4,
TD 5, TD 6, TD 7, TD 8 or what) and then tell that to the player.
Of course my style of GMing doesn't care what happens in any
isolated case. I don't care if the PCs all die. Or if they
survive something really dangerous that they shouldn't
have survived. I only get upset if the unrealism persists. If
the players keep making new PCs and they keep dying. Or if
they keep going into adventures they aren't prepared for and
still survive. Then something is wrong and the system needs
to be examined and repaired.
Other GMs, presumably all Dramatists, have a need to determine
or at least effect the outcome of individual events.
> [...]
> >But from a character creation perspective, I fear that it
> >creates too much fuss. That the player spends billions of
> >brain cycles on deciding whether to give his character
> >DX 12, 13 or 14.
>
> Well, yes, that is a problem. But that is a one we chose to
> accept because of specific reasons (ie. the consistency with
> the setting).
Probably a good idea.
Although I honestly still can't bring myself to care :-)
I would not object to such a system. Playing under it is okay,
and I don't mind GMing under one either (Quest FRP has a
similar attribute scale of 3-12-20). It's just that while
I'm designing a system from scratch, I might as well design
it so as to avoid the problem.
> >Also from an IC-point of view, it creates problems. My
> >character has DX 11, and your character has DX 12. I, as a
> >player, knows that your character is more dextrous than mine.
> >But does my character know this?
> >In FFRE, if my character has Dex 3 and your character has Dex 4
> >then it should become obvious after some days or weeks of
> >physican acquantance. Beyond doubt.
> >Your system leaves room for doubt and uncertainty.
>
> Yeah. That's the point. My brother and I, we practiced martial arts
> together. I don't know who was more dextrous, me or him. We were kinda
> even - sometimes I won, sometimes he did. There was doubt and
> uncertainity.
Maybe you were equally dextrous?
Such situations are of course more likely to occur in FFRE
than in IF or GURPS, since the variation in attributes is
smaller - almost everyone in FFRE is 3. For GURPs the
range of normality is 9-11 or 8-12 depending on what
table you look at.
> And it fits in with the style of the setting.
True.
> (Now, pulp IF, with it's compressed scale, leaves no room for
> uncertainity - the scale is even tighter than in FFRE, with 'normal'
> humans compressed to 0-5 scale. It also fits with the style of
> the setting. :)))
What does 0 represent on this compressed scale?
> [...]
> >d100 is nanoscale, 1-20 is microscale.
>
> OK, a matter of nomenclature I guess.
Or sophistry on my behalf :-)
> [...]
> >> Well, you do have a problem of impossible difficulties - TD 11 in
> >> FFRE is just Plain Impossible (TD 13 for somebody with a Major Talent).
> >
> >Actually it's TD 10 and 12, since I use the nonstandard d10-
> >reading 0=0 convention instead of the standard 0=10.
>
> Oh, yeah.
I know this is confusing and I wish I could change it but
it would make computerized tools much less useful if I
changed to the 0=0 convention. And I'm just in the
process of writing a great and very fucntional QBASIC
program that rolls dice for you in all sorts of situation
(single rolls, Long Task rolls, Spellcasting rolls,
Barter rolls, Gambling rolls, Damage rolls).
> >And apart from that, I don't see a problem.
> >
> >In other RPG systems, you simply never roll for TD 13+ tasks.
> >Such as trying to hit a jet aircraft with a thrown stone.
> >FFRE works the same way, so what's the problem?
>
> Well, but I don't feel that TD 12, for example, is like trying to hit
> a jet aircraft with a thrown stone. It's more like dodging bullets, to
> use a cinematic example. :>
OK.
> OK, let me elaborate.
>
> We have Robin of Locksley. He is shooting at a distant target, in a bad
> lighting. TD of 11, but with his Major Talent it goes down to 9.
> He rolls his Archery skill of, IIRC, 12. 011223455689. A hit.
> Now, he also is shooting at a distant target, this time in plain
> daylight, but with a crosswind. Again, TD of 11, and this time he
> rolls 00223577899. A hit, with 2S. So, confidently, he reaches for
> another arrow and... well.. the sun has just set. Distant target, TD 10,
> crosswind, TD 11, poor lighting, TD 12. Ooops. Even Locksley can't do that.
> There's a _sharp_ drop in hit probability, to a nice round 0.
True if you see it from a success-centric point of view.
If you take a step back, you will see that the most likely
outcome only changes from 1S (barely a hit) to f-1 (barely
a miss).
Not much difference. It's true that f-1 is not a hit and that
it never does any damage. But Robin can still keep shooting,
confident that he will get no Fumles (at least not any of
the bad ones, like F-4 and F-5).
Also, there may be Aiming rules that lowerst he TD by 1
if you spend two rounds (a round is 10 seconds) to aim.
Of course this does not solve the problem, only postpones it
(Robin can still hit, if he aims, but if the evil GM adds one
more adverse condition then Robin cannot hit any more).
(Robin might also have Luck points to spend. I haven't decided
what they can do, except allow re-rolls, but it might be legal
to spend a Luck point to lower the TD of a task, on a 1:1 basis.
But a normal could do this as easily as a Talented person, so
the problem still isn't solved)
> >> In Pulp IF, there's no limit to how high can one go. It's something
> >But that's because you apply task difficulty as a modifier to
> >the skill roll. Which results in Talented characters being
> >only quantatively better than normals.
>
> No, task difficulty is the target number (OK, not directly, TN is based
> on how much the the task difficulty is higher than the skill). This is
> because there is no upper limit to skill, and the difficulty of task
> one can accomplish is dependant on how high is the skill.
>
> >In FFRE, a character such as Robin of Locksley would be
> >qualitatively better than a normal.
>
> Than make the open-ededness a result of Talent.
What do you mean?
> Now the poor Robin is mainly quantatively better than normals
> (not because of his skill, but because of his Talent - the sharp, two
> point drop in TD does wonders for the number of successes at medium
> and low difficulties - crunch the numbers and you'll see), and
I have crunched the numbers (I have an old logfile from a QBASIC
program that I can share if you're interested) and you're correct.
> his qualitative advantage only allows him to barely succeed at a
> slighly more difficult tasks than the normals are capable of.
But he can succeed. That's my point. A normal cannot succeed, ever.
It's literally impossible for a normal to succeed at TD 10 and TD 11
tasks.
A normal has no chance.
Robin has a small chance.
> If you say that, for example, Talent allows one to re-roll zeroes
> (or nines, in your case) when rolling against TD of 10+ _once_, and
> Major Talent allows one to re-roll the nines _again_, then Robin of
> Locksley will not be 80% quantatively and 20% qualitatively better
> than normal humans (his main advantage being the much higher number
> of successes he gets due to lower TDs), but he will be 100% qualitatively
> better (he shoots like a normal archer, but can do things normal
> archers cannot do - like reach TD 15).
How would this work?
Should the re-rolled dice then count as 10+d10 or what?
Allowing re-rolls, under normal circumstances, does not
enable a normal to reach anything above TD 9 or a Major Talent
to reach anything above TD 11.
> And if you use both options (Major Talent being both a minus to TD and
> the ability to re-roll for TDs that are still higher than 9) then
> Robin will be not 'slightly qualitatively better', ie. able to compensate
> for 2 more points of bad condition modifiers, he will be considerably
> qualitatively better.
I don't get it, please try to explain in greater detail the
re-roll mechanic.
> [...]
> >> Make the capability to make open-ended rolls an Advantage that is available
> >> only in cinematic campaigns, and voila. The difficulties above 10 will still
> >> drop the number of successess sharply, but FFRE will now scale upwards
> >> without limits (theoretically ;))).
> >
> >But if I allow that then Talented characters looses their
> >ability to be qualitatively better than unTalenteds. Then they
> >can only be quantatively better. Then any random person could,
> >with persistence and decades of training, become as good as
> >a Born archer such as Robin of Locksley.
>
> No, why? This random persons shoots as well as Robin in normal conditions
> (well, one can't get closer to the bullseye than the exact centre of it),
That's true and raises some interesting questions for the
simulation or archery contests.
> but when the J. Random Bowyer shoots at a flying dragon in the midst of a
> raging storm, using only the light from the dragon's fiery breath, he'll
> have absolutely no chance at all of succeeding. And Robin of Locksley will.
Maybe, but first of all I need to understand the system you're
proposing.
> Using the system of -2 to TD, Robin will shoot better than J. Random Bowyer
> in normal conditions, (OK, with TDs above 3), but will be able to do only
> slighly better when it comes to shootin in adverse conditions.
You still need to explain the exact nature of your proposed
re-roll mechanic, because I don't get it.
Let's say TD is 15.
Robin is skill 12, Major Talent (the new nomenclature for
FFRE is that this is written "12+2", not "12 (T+2)"), and
he rolls this
012344578899 (vs TD 15 -> TD 13)
Now what do you propose happens?
As far as the current rules are concerned, the highest dice
is 4 below TD. 4 or more below TD is a F-5, known as a
Major Fumble, and Robin should suffer griveous consequences.
Your proposal is that Robin is allowed to re-roll the two 9's
(and also the two 8's since he has not just Talent but Major
Talent) but what's gonna happen to the TD?
If the TD is still 15 then it doesn't matter.
Should the TD be lowered for the purpose of the second roll,
or should the re-rolled dice be enhanced, or what?
Peter Knutsen
"Warren J. Dew" wrote:
>
> John Kim, regarding how the requirement to reverse engineer the system to
> determine attributes when they aren't specifically defined in terms of skill
> groups:
>
> True, but it's a lot easier to reverse-engineer 3 attributes
> than to reverse-engineer 12, I guess. :-)
>
> That's a good point. I personally don't have an objection to people reverse
> engineering a point based system to get the character they want. I'm curious
> about Peter's opinion, here, though. Peter, do you specifically want to
> prevent reverse engineering the system so that there are no surprises in
> character definition?
No, not at all. I'm convinced that the system is good (until
proven otherwise) and I'm not ashamed to let the players see it.
Also, assumption clashes are bad and disruptive, so they should
be avoided.
But I think that sittin down with a player and explaining to
him the aptitude of his character for every single skill in
the game (FFRE will probably have a couple of hundred, many
of them magic skills) is overkill and will result in more
player resentment than any assumption clash.
[...]
> Essentially, you want the number-crunching to be obscure enough that
> the player doesn't realize some important ramifications for her PC.
>
> No. I don't have a problem with players realizing ramifications the player
> considers important; what I'm interested in are the ramifications the player
> considers unimportant.
That was what I failed to say, the last time.
Leszek Karlik
disseminated foul capitalist propaganda:
[...]
>The effect is still there, yes. If your attribute is higher
>than it becomes cheaper to raise the skill. Yes.
>
>But in FFRE the skill starts from zero. In GURPS your attribute
>gives you a head-start.
Starting the skills from zero can be a bad thing - after all, the
'default' skills do exist in Real Life. Apparently, anybody can pick
a gun and point it and shoot. He will probably miss, and might have
problems with figuring out what's happening if the gun doesn't fire,
but a gun that was taken from a dead body who was using it before
becoming dead ;) will not have it's safety on and can be used.
This can be attributed to 'free' rudimentary training in some skills given
to us by our culture (a caveman would not know how to use a gun - though
a Roman Legionnaire might figure it out by assuming it's a wierd looking
crossbow).
[...]
>Finally, in GURPS extreme optimization requires no effort.
>In FFRE you have to spend two minutes reading some rules.
>Plus you have to optimize in a more focused fashion, for
>specific skills rather than for just skills in general.
This is true, and this is what I am aiming at with Idee Fixe.
>> If X < Y then there is a problem. In GURPS, this condition
>> is easy to meet.
>In FFRE it's harder to meet. But it's there, since it would be
>extremely unrealistic for characters to all learn at the same
>speed.
True - but that means that play balance and realism are mutually
exclusive.
Which, of course, makes sense. Some people are poor, sickly, stupid and inept
and some are rich, healthy, smart and skilled. High differences between
PCs are usually disruptive, unfortunately. (Though not always, of course,
and I particularly like Ars Magica's concept of troupe style play. :)).
>> But no system that has attributes influencing skill price and
>> a common pool of points to buy both attributes and skills is
>> completely free from that problem, I'm afraid.
>
>True but what's the alternative?
>I try to make optimization expensive (in terms of time it
>requires, and in terms of cost-benefit ratio), hoping to
>raise the treshhold so that fewer players will bother.
That's what I tried to do - but then it turned out that
[...]
>> >Do you use the Skill = (Attribute + Training) method, or something
>> >else?
>> Skill is attribute modified by the skill difficulty plus number of skill
>> levels (easy skills start at attribute -3, moderate at Attribute -6,
>> and hard at Attribute -10).
>So your answer is yes.
Yes. :) I tried to change it to Attribute * 0.7, Attribute * 0.4 and
Attribute * 0.1, which would reduce the head start given by high
attributes, but I was voted out by the dev team. (Since the attributes
change during game play, they tought it too complicated)
This has changed already, though. :)
[...]
>Apart from that, I'd like to point out that I'm not talking
>about cinematics, I'm talking about realism here.
>
>The experience of playing a character that is as smart as a human
>can be, in GURPS, is seriously unrealistic.
Well, yeah, you're right.
[...]
>> Yeah, you're right. I am now thinking about moving to more attributes,
>> separating attribute and skill points, changing the relationship
>> between attributes and skills (though using this solution alone still
>> does not solve the problem) and various other solutions to this problem.
>
>I guess I've made you reconsider your design choices. I should
>probably be proud of that :-)
Probably. :) I've already changed the skill system (the default values of
skills are constant, the price rises as the skill exceeds certain breakpoints
based on attributes) and I'm moving from the original system of four base
and four derived attributes, with skills based only on two, to a system
of nine attributes, with skills based on six attributes.
(The new attributes are: Strength, Willpower and Resistance;
Dexterity, Agility and Reaction; Wits, Intellience and Perception. The six
latter attributes are used for various skills, the first three determine
the character's strength of body and mind).
[...]
>> Well, GURPS is an, ahem, rather old system. :)
>"3rd Edition revised" is only a couple of years old :-)
Unfortunately, main changes since the first edition involve adding
various rules and Advantages. The core mechanics are still intact.
[...]
>> >The reality is that some systems are better than others, and
>> >that some systems may very well be abuse-proof.
>> I guess that's what
>Incomplete sentence?
Yeah, sorry. I don't even remember what I wanted to write, but
it was probably something along the lines 'I guess that what depends
on the definition of abuse'. ;).
[...]
>> If 'talented character with moderate training' has the same effectiveness
>> as 'grizzled veteran with lots of points', and is cheaper (because the
>> character spent more points on attributes and thus could get to the
>> same level of competence as the veteran for an overall price that is
>> lower), that constitutes abuse, IMO.
>
>I agree with that.
>
>I doubt if I can remove the difference (in cheapness and
>effectiveness) completely in FFRE, but I hope to reduce it
>vastly vompared to GURPS.
That should not be difficult - GURPS _is_ an extreme case, in
my observatiosn.
My solution is also reducing the difference, and also including pre-generated
characters in the main rulebook. This, along with some guidelines on
tweaking them to personal tastes should allow somebody who doesn't like
complicated character creation processes to have a character that doesn't,
well, suck.
There's always a possiblity of a situation like in Shadowrun 2nd ed., of
course - the pregenerated characters in SR 2nd ed. are wimpy. They were
done by somebody who obviously had no clue when it comes to SR character
optimization. I mean, I could genrate a better character in 15 minutes
during a boring geography class or something. :) Luckily, this has
changed in SR 3rd edition - the characters there are OK, they are pretty
efficient and were made by somebody who knows his stuff. ;)
In Idee Fixe, I'm going to make the pregenerated characters. That should
keep them rather well optimized for the tasks they are supposed to
perform. :)
[...]
>If a complete newbie to the system, who has understood the
>attribte list (=he knows the difference between Agility and
>Dexterity, for instance) and embraced the notion that
>attributes helps you learn skills faster/easier, makes a character
>that is 10% or 15% more expensive than a character made by a
>system expert who has gone to great lengths to optimize, then
>I don't think it's a problem. Because it's only a 10% or 15%
>difference.
>
>In GURPS, the difference is often one of many hundred percent.
>This is where the problem is.
Yeah, yeah, you're of course right. I hope I can get the difference
down to something about 10 or 20%, too. :)
[...]
>> >Characters with advantages will still exist, be legal and be
>> >necessary. Anybody with above-average quality of appearance
>> >(Attractive, Very Handsome, Mindbogglingly Beautiful) needs
>> >an Advantage.
>>
>> This could be an attribute, you know. :)
>
>It probably will be an extrinsic ("fifth category") attribute.
>
>It will be extrinsic because it is trivially easy to alter it,
>using Illusion magic, Shapechange magic or plastic surgery.
So are other attributes. Strength is trivially easy to alter with
Body magic, Shapechange magic or muscle grafts in cyberpunkish world. :)
However, most of those changes are going to be temporary, won't they?
>I could probably do something such as
>
>3 being Normal
>4 being Attractive (similar to the GURPS advantage)
>5 being Beautiful/Handsome (again similar to GURPS in spirit)
>6 being Very B/H
>7 being Incredibly B/H
>8 being Mindbogglingly B/H
>
>with lower-than-3 representing reduced attractiveness.
>
>A similar trait scale could exist for Social Status, Legal
>Status, Wealth and maybe others as well.
Social and Legal status as well as wealth can change easily, just
by travelling. Appearance - well, the standards of beauty can also
change, but not by much. The universal categories of human attractiveness
when it comes to human _face_ do not change much.
It is easy to _reduce_ Appearance, too, but then, so is Agility. Just
cut up the sinews, and voila, you have made somebody a crippled man,
reducing many of his physical attributes.
So, Appearance has a good claim to be an intrinsic (as opposed to Status or
Wealth, which are extrinsic) attribute.
[...]
>Some GMs may feel, for odd reasons, a need to be able to set things
>up so that PC Gabriel has exactly a 55% chance of hitting the
>pear with his hand crossbow. But I don't understand that mentality.
>
>Why do you need to know the exact probability? It should be
>enough to know the difficulty of the task (whether it's TD 4,
>TD 5, TD 6, TD 7, TD 8 or what) and then tell that to the player.
Yeah, that's true.
[...]
>> (Now, pulp IF, with it's compressed scale, leaves no room for
>> uncertainity - the scale is even tighter than in FFRE, with 'normal'
>> humans compressed to 0-5 scale. It also fits with the style of
>> the setting. :)))
>What does 0 represent on this compressed scale?
Small kids and cripples (0-3 'normal' attribute).
[...]
>> >d100 is nanoscale, 1-20 is microscale.
>> OK, a matter of nomenclature I guess.
>Or sophistry on my behalf :-)
Well, this _is_ UseNet. :>
[...]
>> >Actually it's TD 10 and 12, since I use the nonstandard d10-
>> >reading 0=0 convention instead of the standard 0=10.
>> Oh, yeah.
>I know this is confusing and I wish I could change it but
>it would make computerized tools much less useful
It's not really that confusing, I just forgot about it.
[...]
>> OK, let me elaborate.
>>
>> We have Robin of Locksley. He is shooting at a distant target, in a bad
>> lighting. TD of 11, but with his Major Talent it goes down to 9.
>> He rolls his Archery skill of, IIRC, 12. 011223455689. A hit.
>> Now, he also is shooting at a distant target, this time in plain
>> daylight, but with a crosswind. Again, TD of 11, and this time he
>> rolls 00223577899. A hit, with 2S. So, confidently, he reaches for
>> another arrow and... well.. the sun has just set. Distant target, TD 10,
>> crosswind, TD 11, poor lighting, TD 12. Ooops. Even Locksley can't do that.
>> There's a _sharp_ drop in hit probability, to a nice round 0.
>True if you see it from a success-centric point of view.
>If you take a step back, you will see that the most likely
>outcome only changes from 1S (barely a hit) to f-1 (barely
>a miss).
>Not much difference. It's true that f-1 is not a hit and that
>it never does any damage. But Robin can still keep shooting,
>confident that he will get no Fumles (at least not any of
>the bad ones, like F-4 and F-5).
But the problem is that he never hits... never, even if he shoots
a thousand arrows.
And a normal guy who has a low skill and usually misses a TD 8 target
will hit it if he shoots a thousand arrows...
>> >In FFRE, a character such as Robin of Locksley would be
>> >qualitatively better than a normal.
>>
>> Than make the open-ededness a result of Talent.
>
>What do you mean?
[...]
>> If you say that, for example, Talent allows one to re-roll zeroes
>> (or nines, in your case) when rolling against TD of 10+ _once_, and
>> Major Talent allows one to re-roll the nines _again_, then Robin of
>> Locksley will not be 80% quantatively and 20% qualitatively better
>> than normal humans (his main advantage being the much higher number
>> of successes he gets due to lower TDs), but he will be 100% qualitatively
>> better (he shoots like a normal archer, but can do things normal
>> archers cannot do - like reach TD 15).
>
>How would this work?
>
>Should the re-rolled dice then count as 10+d10 or what?
Yes. Well, they should count as 8+d10, otherwise you go from
rolling 9 to rolling 11, with no possibility of rolling 10.
This is the problem with Shadowrun dice mechanics - TN 7 is equal to TN 6,
since if you roll a six on a six-sided dice, this is also a seven, since
you cannot roll a zero on a six-sided dice.
No, wait. I have just realized that you CAN roll zero on a ten sided dice
in FFRE. Then the re-rolled dice should count as 9+d10 - so rolling 9 will
not also mean automatically rolling 10.
>Allowing re-rolls, under normal circumstances, does not
>enable a normal to reach anything above TD 9 or a Major Talent
>to reach anything above TD 11.
Well, re-roll is not what I meant by 'open-ended'. I guess you haven't
played either Shadowrun or Star Wars, right? The Wild Die in WEG Star
Wars, or all the dice in Shadowrun are open-ended - you reroll the six
and add the re-rolled number to it.
[...]
>> but when the J. Random Bowyer shoots at a flying dragon in the midst of a
>> raging storm, using only the light from the dragon's fiery breath, he'll
>> have absolutely no chance at all of succeeding. And Robin of Locksley will.
>Maybe, but first of all I need to understand the system you're
>proposing.
Well, I thought you had some contact with Shadowrun or D6. I see you haven't,
otherwise the detailed explanation is unneeded.
>> Using the system of -2 to TD, Robin will shoot better than J. Random Bowyer
>> in normal conditions, (OK, with TDs above 3), but will be able to do only
>> slighly better when it comes to shootin in adverse conditions.
>
>You still need to explain the exact nature of your proposed
>re-roll mechanic, because I don't get it.
>
>Let's say TD is 15.
>
>Robin is skill 12, Major Talent (the new nomenclature for
>FFRE is that this is written "12+2", not "12 (T+2)"), and
>he rolls this
>
>012344578899 (vs TD 15 -> TD 13)
>
>Now what do you propose happens?
>
>As far as the current rules are concerned, the highest dice
>is 4 below TD. 4 or more below TD is a F-5, known as a
>Major Fumble, and Robin should suffer griveous consequences.
>
>Your proposal is that Robin is allowed to re-roll the two 9's
>(and also the two 8's since he has not just Talent but Major
>Talent) but what's gonna happen to the TD?
>
>If the TD is still 15 then it doesn't matter.
>
>Should the TD be lowered for the purpose of the second roll,
>or should the re-rolled dice be enhanced, or what?
Well, each re-rolled die counts as 9+d10, so if we assume that
Robin rolls, on the second roll of 4d10, 0246, the final result
will be 01234457 9 11 13 15, for 1S. Wow! Robin _did_ suceed!
(OK, now here we have a problem with two-digit numbers... Sorry.
Traveller used a hexadecimal system for recording character data,
but this won't work here. :))
Now, if we assume that the re-rolled dice are added to the number
they previously rolled, and that he rolled 6 on the dice that has
rolled 8 previously, than his highest result is going to be 14,
not 15, which will still be a failure, BUT an F-1 instead of
the catastrophic F-4.
Peter Knutsen
I can't really envision that happening in FFRE.
> I didn't know I was going to be a biologist until fairly
> late in my school career, but in retrospect there were lots of
> details consistent with this. I just hadn't realized that
> fondness for bugs had anything to do with a career path.
> If I'd played a character for some years, and all the little
> details were in place--the fondness for bugs, the analytic
> puzzle-solving skills, the curiosity, the quick grasp of
> logical relationships--but when the character finally took
> up the skill she was actively bad at it because of my
First of all, you can't be actively bad at something without
having an Incompetence, and an Incompetence is an explicitly
player-chosen disadvantage. Similarly you can't be actively
good at something without having a Talent which is an
explcitily player-chosen advantage.
What you can be is be slower or faster at learning a skill
than the average person.
And as you've read in a previous thread, the theoretically
possible span of learning speed is huge (a factor of 70 or
so - more if you count in characters who have 0 in some
of the relevant attributes).
But in reality the learning speed difference will be smaller.
Let's say that you've created a typical amazon warriorine with
a tendency to go berzerk in combat (just to name an example).
You don't want to do the "dumb fighter" stereotype so you
give her average Intelligence (3) but low Willpower (2).
She's a good woodswoman, however, so she needs to be
Perceptive (5). You also define her physical attributes in
such a way that she becomes a good fighter but these are
irrelevant here.
Then for some odd reason she ends up in a present-day Western
country and falls in love with a biology professor. She's
desperately looking for a way to get his attention and
impress him, so she decides to study biology.
Okay, the Aptitude formula for Biology (the same for physics,
an almost the same as for chemistry) is IIIIWWP, meaning that
Intellect plays a large roll (weight 4), Willpower plays a
medium role (weight 2) and Perception plays a minor role
(weight 1). So we calculate Apt = (3+3+3+3+2+2+5)/7 = 3.
Her Aptitude(Biology) is 3, which happens to be exactly the
same as the average member of the population. What good
fortune!
However, her classmates are not average. The average Biology
student at the University is probably Intelligence 5, Will 4,
Perception 4 (with some variation - a few may be INT 4. Some
may be INT 4, WIL 5, and so on). So the average Apt of her
classmates is (5+5+5+5+4+4+4)/7 = 4.571.
Now, at first Apt has no direct importance (it will, later),
but Learning Speed (LS) will. And you find LS this way:
LS = (Apt*Apt)/9.
The Amazon warriorine has a LS of 1, meaning that she learns
as fast as the average person. (convenient, eh?).
Her average classmate, however, has an LS of 2.32, so they
learn more than twice as fast as the average person. Since
the University enviroment is geared for such people, the
system expects people to have a LS of maybe 2.25 or 2.5 or
so.
So your character learns at 43% the speed of her classmates.
But she's not doomed.
First of all, for the sake of simplicity we'll assume that
all students arrive at the University knowing no biology
whatsoever (yes, I know this is absurd).
Also there are regular classes which must be attended, and
assignments which must be solved on time. But individual
students can choose to put in extra effort, by using more
spare time on reading and solving training problems.
Assuming Danish standards, a biology study at the university
is "normed" to take 5 years (at this point you'd get your
Kandidatgrad or Master's Degree) but in reality the average
time it takes a student to graduate is (approximately) 6.5
years.
Furthermore I'll assume that the average student is as
described above (Apt 4.571) and puts in 30 hours of work
per week 45 weeks a year. This means that an entire study
up to Master's Degree level of proficiency requires
30 * 45 * 6.5 * 2.32 = 20358 hours of study for the normal
person (an average biology student can do it in 8775 hours).
So your amazone warriorine could choose to spend 50 hours
a week studying instead of 30. That way, instead of being
43% as fast as her classmates, she'll be 72% as fast. Or
she could study 60 hours a week, or pay for special
tuition (you learn faster if you have a teacher to help
you than if you study alone).
Chances are she won't keep the same pace as her classmates.
But she isn't doomed.
She can complete the study, if she wants to.
And anyway, the game (I almost said story!!!!) can have a happy
ending if the professor becomes impressed by her determination.
Assuming that she does complete the study, she has a Biology 5
skill, the same as all the others who graduate (except for a
few who put in extra work and thus have Biology 6) and she
will perform just as well as the others. She rolls 5d10, they
roll 5d10.
No difference. Except she had to spend a bit more time getting
there.
And keep in mind that this is the case of a character who is
not at all suited for learning knowledge skills.
If you have made an "above average smarts character" with
INT 4, WIL 4, PER 4, then she would have learned almost as
fast as the other students. (Apt 4 instead of Apt 4.571 and
LS 1.78 instead of LS 2.32).
> lack of understanding of the mechanics system, well, I
> don't think I would find that a helpful or entertaining
> form of surprise.
As I wrote in reply to another poster, chance are you won't
get such nasty surprises. The expected variation is probably
a factor of 50% at most. You find out that your character
learns a skill 50% faster or 50% slower than what you would
expect. Usually the factor is smaller, like 10% or 20%.
> But if the game's on a different level of detail maybe this
> wouldn't happen, and so the problem wouldn't arise. We tend
> to play very high-detail and very slowly.
I'm not sure what you mean here, can you try to explain?
> Mary Kuhner mkku...@eskimo.com
--
Peter Knutsen
Peter Knutsen
Have you thought about asking the players if they'd like to try this?
I think that doing a campaign wherein the players discover how
magic works could be fun, given the right group of players.
But normally I would not do such a thing. I'm 100% pro rules
visibility.
> Neel
--
Peter Knutsen
Steve Mading
: So far so good.
Not necessarily.
: When they emerge Beatrice has a skill of 92% and Chris has a skill
: of 74%.
: And my suspension of disbelief evaporates instantly.
: I need people with high aptitude to learn faster than people
: with normal aptitude. I need people with low aptitude to
: learn slower than normal people. Chris is severely retarded.
: He needs constant care to ensure that he's kept healthy and
: well fed. Beatrice is a genius, probably smarter than all of
: us rgfa'ers added together.
: Beatrice should emerge, after having spent the same amount of
: effort as Chris, with a skill at least 10 times higher.
: Preferably 30 or 100 times higher.
I wouldn't want to go so far as to skew things that far out
of whack, but I could see the argument for adjusting the scale
a bit. Nobody said that the number of points the two people
got for those years of schooling is the same. I never made
the implication that points == time spent.
: If you don't believe me then it's because you've never
: met and interacted with a person who is severely retarded.
:> you *have* to have the ranks in training or your high int doesn't help.
: Yes but it's wrong that getting the ranks requires the same
: effort for everyone.
: Also, your Mace example is problematic in that only strength
: plays a role. What about hand-eye coordination? The Mace
: skill is presumably about hitting something.
That has nothing to do with the system itself, it's just a bad
choice for what attribute to base that one skill it on. My
first instinct was to do something that would have made the
example more complex, so I left it out - namely to make the
base be an average of two stats - str and dex.
:> Now, my examples above assuemd a percentile system. But the same
:> type of idea could be done on some other sort of scale. The key
:> thing is to make sure that the typical ranks in a skill and the
:> typical attribute ranges are on approximately the same scale.
:> (So if a X is an average attribute score, then X ranks is an average
:> amount to put in a skill), and then to make sure that the scale
:> in which a skill check is made is roughly N times as large as
:> the scale in which the attributes and ranks fit. (Where N is the
:> total that all natural and learned factors have to add to in your
:> skills.)
: I'm not playing if Chris can learn at the same speed as
: Beatrice. It's beyond absurdity.
I never said that was the case.
Steve Mading
: On Thu, 15 Mar 2001 07:46:48 +0100, Peter Knutsen <pe...@knutsen.dk>
: disseminated foul capitalist propaganda:
: [...]
:>The effect is still there, yes. If your attribute is higher
:>than it becomes cheaper to raise the skill. Yes.
:>
:>But in FFRE the skill starts from zero. In GURPS your attribute
:>gives you a head-start.
: Starting the skills from zero can be a bad thing - after all, the
: 'default' skills do exist in Real Life. Apparently, anybody can pick
: a gun and point it and shoot. He will probably miss, and might have
: problems with figuring out what's happening if the gun doesn't fire,
: but a gun that was taken from a dead body who was using it before
: becoming dead ;) will not have it's safety on and can be used.
But which "default" skills exist in real life is largely a function
of what society you grew up in. I think a better solution would
be to have a 'default starting package' of skills that all characters
in a culture are required to take. Characters from another culture
have a different starting package. It makes sense to say that even
the poorest American with no car still has a tiny amount of 'free'
skill in driving a car, just from the sheer exposure to cars
that you get in such a society. Everyone knows the basics even
if they've never done it themselves. Most driving training courses
in the US are really about getting comfortable physically with what
you already know mentally about driving. Now, the same might not be
true for a midevil swordsman, or a Bantu tribesman, or a man from a
future world where all transport is computer controlled.
Nis Haller Baggesen
>
>
>
> I'm curious: do you think of real people in terms of attributes, or
> only characters?
>
more than) as in terms of skills. I oftenthink of my friends as being
smart/humorous/quick/strong etc. as much as I think about their skills.
Of course this may be becuase many of my frinds study natural sciences
like myself, so skill is not such a distinguishing characteristic. Also
with strangers I am much more likely to think :"He looks a bit dim" or
"He look quite agile", rather than think, "he look like a construction
worker" (These three examples do not necessarily describe the same
person).
Lance Dyas
them... Unless its blatantly obvious. In fact traits tend to be specific
The bear guzzlin, Jock or body builder for instance gets a mention for
rolling paunch and bulging biceps, short clipped hair ... if I meet them
in their societal role they become the cop... or PE teacher with etc.
Or in their personal relation the husband of such and such who looks
like
they might do drugs...
They are npcs, even so Age, size, other identifying feature and some
general stereotyp
e and a preferance of theirs or context specific information about how
they interact
with the ME the PC is almost always logged.
Close friends THE OTHER PCs might get details like strong grip and high
pain tolerance,
loves to climb dangerous places, wirey and thin and near sighted, sleeps
very soundly,
intellectual with a crisp memory and perception, little ambition (not
much of a risk
taker, insufficient self confidence though able to work hard), strong
mechanical
aptitude, well educated in Mathematics... with some physics and
computers, loves to drive
in bad weather, fan of star trek, insulin dependent but loves REAL coke.
Vulnerable
to sex appeal.
Yes the preceding describes one person.
PC details are Usually "discovered" over quite a while as they start out
much like the npcs.
Even the PC detailed traits are of the vague sort [ somewhat,
definitely, very and extremely]
Most however have role archetypes and stereotypes with obvious visible
snippets thrown in
--
*****************************************************************
Dragonlord Garthans - Decision Driven Gaming - Role Play center
http://incolor.inetnebr.com/lancelot/roleplay
Want custom art for your RPG characters?
*****************************************************************