Stat gain potions considered harfmful.
Ray Dillinger
Stat Gain Potions - number one contributor to "thousanduplet
syndrome", where all the winning characters look alike. With
stat maximums in place, plentiful stat gain potions make all
the characters attain the stat maximums, thereby becoming
identical thousanduplets.
What would the games be like if stat gain potions became so
rare that, in the whole game, you might not find more than
two or three of them?
Answer: The winning characters would be more differentiated,
The fights would be more exciting, and the tactics available
to different characters would be more different.
Just something to consider....
Bear
ABCGi
Or the different races and class combinations lead to a certain limit to
how much "drugs" the PC can hold. Above this limit there is an
increasing chance that the stat up will cause another stat to fall.
This is an opportunity to further differentiate class/races and give a
helping hand to typically weaker classes/races.
--
ABCGi - ab...@yahoo.com
RL LINKS:
http://codemonkey.sunsite.dk/projects/beyond/beyondrelatedlinks_20041117.html
James Bulgin
Well, my roguelike will have no potions of permenant stat gain whatsoever,
athough its possible that a few non-repeatable quests might give a small
permenant stat increase. Starting stats are also less random than in most
roguelikes, but grow continuously over levels with different classes having
distinctly different patterns of growth. I've never liked the traditional
rolling of stats. It has too wide a random range for my tastes, and people will
keep re-rolling until they have near optimal starting stats, anyway. I'd rather
just have everyone start with more or less the same stats per class and do away
with that bit altogether.
Jeff Lait
>
> Stat Gain Potions - number one contributor to "thousanduplet
> syndrome", where all the winning characters look alike. With
> stat maximums in place, plentiful stat gain potions make all
> the characters attain the stat maximums, thereby becoming
> identical thousanduplets.
Not to mention the horror of Stat Gain in *Bands.
IMHO, a system with no stat gain should consider removing random stat
rolls on start entirely. Instead, one's stats would be dependent on
class/race and be fixed, permament, modifiers based on that set up. If
there is no potential for stat advancement and you have stat rolls, one
is just asking for stat-scumming.
(I'd go farther and say Stats considered harmful :>)
--
Jeff Lait
(POWDER: http://www.zincland.com/powder)
Raghar
@typhoon.sonic.net:
>
>
> Stat Gain Potions - number one contributor to "thousanduplet
> syndrome", where all the winning characters look alike. With
> stat maximums in place, plentiful stat gain potions make all
> the characters attain the stat maximums, thereby becoming
> identical thousanduplets.
>
Stats caps?
> What would the games be like if stat gain potions became so
> rare that, in the whole game, you might not find more than
> two or three of them?
>
> Answer: The winning characters would be more differentiated,
> The fights would be more exciting, and the tactics available
> to different characters would be more different.
>
Players would take more time to properly roll up theirs characters.
Antoine
Ray Dillinger wrote:
> Stat Gain Potions - number one contributor to "thousanduplet
> syndrome", where all the winning characters look alike. With
> stat maximums in place, plentiful stat gain potions make all
> the characters attain the stat maximums, thereby becoming
> identical thousanduplets.
>
> What would the games be like if stat gain potions became so
> rare that, in the whole game, you might not find more than
> two or three of them?
>
No stat gain potions in Guild, instead the player gets to increase one
or more of a character's stats every time they level.
This gives a lot of freedom in character development - you can take the
'munchkin' approach and just pump the key stats of your class, or you
can go for a well-rounded character, or you can focus on a secondary
stat for your class as well as the traditional key ones.
I have a feeling that the munchkin approach might make life quite
difficult, as having low secondary stats in Guild can be quite a
hindrance. There is a lot of difference between stat values of 8 and 12
- more, I think, than between 14 and 18 (another irritating thing about
Roguelike stat systems: each additional point of a stat often doesn't
seem to make a lot of difference, particularly at low levels).
A.
Raymond Martineau
>> What would the games be like if stat gain potions became so
>> rare that, in the whole game, you might not find more than
>> two or three of them?
>>
>> Answer: The winning characters would be more differentiated,
>> The fights would be more exciting, and the tactics available
>> to different characters would be more different.
>>
>Players would take more time to properly roll up theirs characters.
That would only apply if characters were rolled. There are a few
roguelikes (some Angband variants in particular) that allow the player to
set their initial stats through a points system.
If there's still a need for an element of randomnexx, then it's still not a
problem - make an initial roll of a certain amount of dice, give the player
additional stat points to spend as necessairy, and continue on as normal.
ru
> Ray Dillinger wrote:
>>
>> Stat Gain Potions - number one contributor to "thousanduplet
>> syndrome", where all the winning characters look alike. With
>> stat maximums in place, plentiful stat gain potions make all
>> the characters attain the stat maximums, thereby becoming
>> identical thousanduplets.
>
> Not to mention the horror of Stat Gain in *Bands.
>
> IMHO, a system with no stat gain should consider removing random stat
> rolls on start entirely. Instead, one's stats would be dependent on
> class/race and be fixed, permament, modifiers based on that set up. If
> there is no potential for stat advancement and you have stat rolls, one
> is just asking for stat-scumming.
some moderate stat gain (maybe as you go up levels, as in crawl) is a
decent way for a player to iron out wrinkles in a character.
smooth over, though - not erase completely.
> (I'd go farther and say Stats considered harmful :>)
okay... i'm interested enough to wonder why you think so, or what you'd
replace them with.
some recent pen & paper rpgs like Fate or Wushu have gone almost
statless (though this is still pretty much restricted to the indie
sector); it's interesting to see the comparison.
--
ru
Mylon
I want to know your thoughts on pen and paper RPing systems.
Particulally two: D&D and GURPS.
In D&D, characters typically follow one path (rogue, fighter, wizard,
cleric), choose an appropriate class, and take that all of the way
through. High level characters tend to be very different from each
other, their tactics differ greatly, and thus their playstyles are very
different and can appeal to different characters. In theory, a
character can try to follow every path at once, but they will be poor
(not even decent) at every such path.
In GURPS, there are no classes. It is a completely point buy system.
If one limits characters to only being able to spend points in "mundane"
traits (except for magic and magic spells) then in theory, with enough
points a character can become a master of all trades (or a
thousanduplet, as you call it). However, characters rarely get enough
points (unless playing a Supers game, which means there's more than just
mundane advantages) to generalize. Thus, characters typically do
specialize, but they can build their characters as they like, taking the
advantages, skills, ect in the proportions that they want, not in the
proportions that the game limits them to. The "point cap" that the game
master indirectly sets is what keeps this thousanduplet syndrome from
occuring.
So is thousanduplet syndrome really a problem of class vs classless, or
is it a problem of classless systems not being "deep" enough and/or
giving too much experience?
Jeff Lait
>
> > (I'd go farther and say Stats considered harmful :>)
>
> okay... i'm interested enough to wonder why you think so, or what
you'd
> replace them with.
I went into this in a fair bit of depth when we were discussing the
merits of Charisma versus Appearance.
By "Stats" I refer herein to that brief numerical summary of character
abilities, such as the ADOM Toughness/Willpower, or AD&D
Strength/Wisdom. I explicitly exclude stuff like Hit Points,
Alignment, Age, Body Weight, Hair Colour, etc.
Why Stats Are Harmful
---------------------
1) Stats are usually created early in game design and then adversely
affect later game design
This is the "Gosh, I better find a use for Charisma" problem. It is
accentuated whenever a point-buy system is added to the statistic
setup. The game developer soon realizes certain statistics are
"useless" so decides to make them useful. This involves creating ad
hoc scenarios purely to exercise these stats.
What should be happening is the exact opposite. A stat should only be
added when it is already useful.
2) Stats encourage needless orthogonality
People designing statistic systems love orthogonality. They want all
stats to be the "same" worth. If they have a mental stat, they need a
corresponding physical stat. One point of strength is considered
equivalent to one point of learning.
Presenting the statistics as a list to the user, often with the same
rolls, and the ability to transfer point between, results in the user
thinking the same. Hence complaints that the Balance stat is under
powered, and the resulting inclusion of tightropes in the dungeon to
justify the existance of the stat.
While users don't expect their character's hair colour to be as
important as their charcter's hit point maximum, for some reason this
assumption breaks down with statistics. I think the fault lies with
the common scale given.
By placing strength and dexterity on the same scale, it suggests to the
user they both represent the same fungible "stat point". In many
roguelikes, it truly is a fungible asset (cf, potions of exchange in
Adom, certain monster attacks in *Bands)
3) Statistics are a hold over from the P&P world
Why are statistics orthogonal? Because that makes rolling up a
character straightforward. Why do we use the same "roll against
Strength" approach for all stats? Because that means players only have
to learn one stat resolution system.
Most importantly, why do we have rolls vs. stats at all? In P&P gaming
sessions, the players and GM are working together to develop a
narrative. When an obstacle is reached, one doesn't want to have to
determine if success or failure is required for the story (the goal is,
of course, to develop a story as a *result* of successes and failures).
Often the nature of the obstacle is one off. The GM determines some
modifiers and the player rolls against their stat.
The question then arises: Why would we have such stat rolls in a
roguelike? The entire gamut of possible actions is already known. The
player can't do anything you haven't programmed, and if you have
programmed it, you are already aware of the roll.
If you have dexterity affecting to-hit and trap evasion, why not just
have a To-Hit statistic and an Evade Trap skill? If strength gives
extra damage and carrying capacity, aren't we better off with a
To-Damage stat and a Carrying Capacity stat?
4) Stat systems introduce a continuous scale where often none is
desired.
A stat system leads directly into the theory that every ability must
lie on a smooth gradient from non-present to fully-skilled. As stated
elsewhere, I am not a big fan of this approach.
The first problem is that it is less fun for the user. When can they
finally use that Heal spell? When they have it at 1% success rate? Or
do they have to wait until 100% success? As a user, I'd rather be
given either the ability to cast heal or not, without the attempt of
the developer to enforce a soft boundrary on me.
The second problem is that it is troubling for the developer. Every
skill needs to be made continuous. You need to use hacks like failure
rates to try and produce a scale of ability for essentially binary
skills (Ie: cast Cure Poison).
One particularly entertaining area is Intelligence. Consider AD&D Int.
Can you use the Int rating to determine what sort of tactics a monster
would use in attacking? An Int 6 human would be assumed to be very
stupid and charge in blindly. But an int 1 wolf pack would exhibit
much superior intelligence. Of course, one can't just assign the
wolves a higher int to reflect this, as then they must suddenly be able
to read & write.
What To Do in a Post-Stats World
--------------------------------
POWDER is "statless" in this sense. Technically there are two stats
(strength and smarts), but they are hidden from the user and only
alterable through polymorph. As such, they can be considered just
properties of the monster type (like whether a monster has a head or
not) rather than AD&D style statistics.
My suggestions for a post-stat game are:
1) Break orthogonality
Stats should specify *units*. Max HP is measured in Hit Points.
Carrying Capacity is in Kilograms. Number of spoken languages is
measured in Languages. Statistics should be things you can measure -
it is a rather ironic turn of fate that its use in roguelikes is to
refer to the very things that can't be measured!
Using units ensures that the user can understand the stat without cross
referencing a manual or spoiler. It also ensures the user realizes
they are not fungible. They won't wonder if they can trade in 10 kilos
of carrying capacity for 10 spell slots.
Breaking orthogonality also means you are less likely to create
"useless" stats as all of your stats were only added when you had a use
for them.
2) Make Secondary Stats Primary
One thing you'll notice repeated above is the replacement of Primary
Stats with Secondary Stats. Primary stats was primarily a book keeping
measure of the P&P days. We do not need it any more. If you wish, you
can always try and calculate an estimated primary stat from the
concrete secondary stats to show the users. Ie, the users strength
score could be reported as some weighted average of their To-Damage
bonus and their Carrying Capacity.
Concrete Suggestion for Developers
----------------------------------
Do not design your stat system first. Design it as a result of other
systems.
Ensure your creature description can handle the addition of arbitrary
named statistics. (Ie, adding another statistic later should not be an
excessive burden)
Make sure each statistic is concrete. It should have units to it. It
should not apply to things not measured in those units.
Ray Dillinger
>
> I want to know your thoughts on pen and paper RPing systems.
> Particulally two: D&D and GURPS.
D&D classes force a particularly extreme character
differentiation, because D&D is, first and foremost,
a social activity. It is not supposed to make sense
to go adventuring with one character. The classes
are interrelated in strengths and weaknesses and need
each other. D&D is designed to help geeks build social
structures and friendships. To some extent, therefore,
the way D&D classes work is not entirely appropriate
to a one-player game.
That said I haven't played D&D since it was AD&D, first edition.
Your description sounds like what I remember, but things may
have changed somewhat.
As of 1st edition AD&D, I felt that D&D classes were too
narrow. There was not much room for one twelfth-level thief,
say, to be different from some other twelfth-level thief. I
remember all kinds of complications and add-ons, mostly published
in Dragon Magazine, that had as their common core trying to
ameliorate this basic problem. What D&D didn't realize at the
time, and may still not, is that their character classes needed
to accomodate a much broader variety of paths of advancement.
Our theif could be a safecracker, or a mugger, or an embezzler,
or a cat burglar, or any of several other possibilities - all
thieves, but very different from each other and with different
skill sets- and AD&D was blind to it. Instead, they used some
of these *very* different character subtypes as "level titles"
as though there were a natural progression from one to the
others.
The response of GURPS, and Hero Systems, and the various
"monoclassed" games, was to institute a pure point-based
system where people could buy any combination of skills and
powers the player liked. There is only one character class,
but under its rules, theoretically, you can build anything.
This is, first of all, hard to balance. Read through the
books and you'll see constant references to getting the GM
to approve this or that or the other thing, notes that the
GM may want to block access to this or that power or advantage
or ability, etc etc etc.... If interpreted mechanically,
players can find any number of exploits in the rules for
character construction that give advantages to some class
of utterly nonsensical characters.
Second, it invites players to game the metagame of character
construction instead of the primary game. By making character
construction complicated and not very automatic, it takes
time and attention away from the primary game, frustrating
some players who feel that they are making decisions "blind"
and inviting rules-raping by others who know the consequences
of their decisions all too well.
Third, there's a tradeoff between having a lot of options
and being really good at a very few things forced by the
way the point accounting is done, and I don't like the
point at which these systems typically balance. I think
that the character, at a given point, ought to have a
*lot* of different skills and powers in theme with his class,
and be pretty darn good at most of them, and be neither
invited to take skills wholly inapproprate to his character
type, nor unduly rewarded for concentrating all his efforts
on improving a single power.
> So is thousanduplet syndrome really a problem of class vs classless, or
> is it a problem of classless systems not being "deep" enough and/or
> giving too much experience?
I think "thousanduplet syndrome" happens in any game that
gives few ways of dealing with most kinds of threats, especially
if it's so random as to defy the development of a strategy of
avoiding what you can't handle. This forces the game designer
to fix it so *everyody* has access to those few ways, one way
or another: scrolls and wands and potions give "necessary" spell
capabilities to everyone, and items with key abilities are
*always* possible to find, etc. And finally, every last one
of the winning characters has exactly the same set of
capabilities, albeit maybe with different special effects, and
it's boring.
If you can run into a poison instakill, for example, *ANYWHERE*
below a certain level, and there are no precautions or alternate
routes allowing you to plan ahead to try to reduce the possibility,
and there is no way to survive it except by being poison resistant,
then every winning character is going to be a character with
poison resistance because there's no other way to survive below
that level. And, in that respect, they will become thousanduplets.
Bear
Max Bolingbroke
> 1) Break orthogonality
>
> Stats should specify *units*. Max HP is measured in Hit Points.
> Carrying Capacity is in Kilograms. Number of spoken languages is
> measured in Languages. Statistics should be things you can measure -
> it is a rather ironic turn of fate that its use in roguelikes is to
> refer to the very things that can't be measured!
I found your post very interesting, but now I find myself wondering how
related stats should be handled: carrying capacacity and how hard I can
brain things with a club, both, presumably, being related to muscle size
should probably go up together to an result.
As far as I can see, this could be done explicitly (by increasing
related stats as soon as a stat is increased) or by having some sort of
resultant statistic which is affected by its relatives but cannot be
changed directly. In the example above, for instance, you would have
Strength as a resultant statistic which determines your ability to
batter wolves sensless, and it would be increased by increasing your
carrying capacity, rock-pushing or digging scores. This feels a bit like
a cludge to me though - I'd be interested in hearing others perspectives.
Max
James Bulgin
> Why do you assume the two things are related to one "muscle size"
> parameter?
While there are plenty of factors affecting a character's carrying capacity and
the mount of damage they deal, raw strength is certainly one of them. Really,
when it comes to lifting things, little ought to count more than strength, I
think. Damage varies more depending on the type of weapon, but you can't deny
that a musclebound hulk will hit things a heck of a lot harder with that maul
than someone of average build.
> Carrying capacity could reflect a better packing strategem. A skilled
> backpacker can carry a lot more for a lot longer than an equally fit
> newbie. Likewise, your damage bonus for hitting with a club is likely
> moderated by a lot more than muscle mass! Flexibility is likely as
> important in order to be able to make use of one's mass effectively.
I personally rather that things like being able to pack better are represented
as skills or traits. Something like a 'backpacking' trait that increased
carrying capacity by 50%, for example. Combat could also be affected by various
types of bonus traits in a similar fashion. Personally, I think dexterity ought
to be factored into damage calcutations for many types of weapons in addition
to strength.
> The reason why we want to relate both back to "Muscle Mass" is because
> we have been trained to create that category thanks to our years of D&D
> influence.
True to an extent, however the categories as they exist weren't invented for no
reason. Not to say that they mightn't bear re-examining and some alteration,
but many of them do have a pretty decent grounding in logic, in my opinion.
> While I mentioned Strength as a resultant stat, I certainly would not
> suggest one use that resultant stat to determine wolf-battering
> ability. Rather, one's wolf battering ability would be a seperate
> stat. Say, "Damage Bonus". The Damage Bonus would feed back into
> calculating the approximate "Str" score which is used only for display.
> (Ie, let the user get some sense of how strong the character is)
I don't really see the point of calculting strength from a seperate damage bonus
stat if you've already said that there will be many other factors with a direct
affect on damage bonuses besides strenght. If it isn't an intergral stat as it
is, I really see no way to calculate it in reverse in any fashion that makes
sense. Assasssins type characters can do a lot of damage because they know
where to strike and are deadly accuracte, but that doesn't mean that they are
neccessarily in any way physically strong.
I think that a traditional stat system provides a nice solid base for various
other ability scores to be derived from. Then, multiple other less tangible
factors, such as skill systems or talents can be added to further refine it.
Hansjoerg Malthaner
> Stat Gain Potions - number one contributor to "thousanduplet
> syndrome", where all the winning characters look alike.
Stat gain potions actually render the character rolling at the start of
the game meaningless.
I agree "stat gain potions considered harmful"
My games will not have stat gain potions with permanent effect.
An idea to solve the "thousanduplets syndrom" and still allow stat gain
potions: only allow drink a certain number of potions. This will allow
the player to shape the PC, yet the limit will hinder the player to max
all stats.
> Bear
c.u.
Hajo
Martin Read
>As of 1st edition AD&D, I felt that D&D classes were too
>narrow. There was not much room for one twelfth-level thief,
>say, to be different from some other twelfth-level thief. I
>remember all kinds of complications and add-ons, mostly published
>in Dragon Magazine, that had as their common core trying to
>ameliorate this basic problem. What D&D didn't realize at the
>time, and may still not, is that their character classes needed
>to accomodate a much broader variety of paths of advancement.
>Our theif could be a safecracker, or a mugger, or an embezzler,
>or a cat burglar, or any of several other possibilities - all
>thieves, but very different from each other and with different
>skill sets- and AD&D was blind to it. Instead, they used some
>of these *very* different character subtypes as "level titles"
>as though there were a natural progression from one to the
>others.
Third edition fixes this. Each time you gain a level, you advance
by one level in your current class, or another class you've advanced
in previously, or a new class. So, by 20th level, you could be 10th
level rogue/5th level ranger/5th level sorcerer, or 20th level rogue,
or 5th level rogue/15th level fighter, or whatever.
All the wacky stuff that used to be "class ability" for thieves is now
handled through the skill and feat systems. Rogues (the thief-
replacement) get an *enormous* number of skill points per level - twice
as many as any other adventuring class (before allowing for the effect
of the Intelligence bonus to skill points received) - and the skills
that pertain to doing the sort of things AD&D1Ed thieves did are tagged
as "class skills" (i.e. advance at one rank per skill point, rather than
the one rank per two skill points rate charged for "cross-class" skills)
for thieves (and most of them are *not* class skills for anyone else).
The class to benefit most in terms of diversity is fighters: two 20th
level fighters can now end up looking very, very different. The elegant
elven swashbuckler in his silk shirt, cloak of protection, and bracers
of defence, brandishing a magical rapier with lethal finesse, is just
as dangerous as the sturdy dwarven champion in enchanted full plate,
smiting his foes with a BFAxe.
In a "normal" adventuring context, a party of specialists remains more
effective than a larger number of generalists, of course, because the
party of specialists will have access to the higher-order capabilities
of their specialisms. A 20th-level sorcerer will take a 10th-level
fighter/10th-level sorcerer to pieces, because the pure sorcerer will
have access to more magic and more powerful magic. At 20 vs. 18/2 or
2/18, of course, the advantage might go to the protagonist with the
small amount of off-specialism capability.
--
Martin Read - my opinions are my own. share them if you wish.
Max Bolingbroke
>>Damage varies more depending on the type of weapon, but you can't
>
> deny
>
>>that a musclebound hulk will hit things a heck of a lot harder with
>
> that maul
>
>>than someone of average build.
>
>
> average build person will out damage them. Technique has a huge effect
> on the amount of force that a blow will land with. An untrained muscle
> bound hulk punches with his arm. The skilled average build person
> punches with their entire body.
But you can hardly deny that that raw muscle mass is having an effect -
there are just other factors to take into consideration.The problem that
I am trying to articulate is that of synergy: getting better in some
things will mean you get better at related things. Evidence for this
effect is everywhere - for example, the ability to cook a nutritous
mushroom soup improves your ability to make that tomato soup you like
the look of because you have mastered the use of a hob and saucepan.
This CAN be represented in your system, since you can have stat
increases in some skills cause increases in others, but the problem with
this (as I see it) is that that it is not very intuitive. How many times
do I have to get better at making mushroom soup before I get better at
tomato soup? Does this trend continue at even high proficiences (theres
only so much I can learn about hobs)? It basically requires a huge
burden of data that the user has to digest to play effectively.
This is mitigated to some extent with a characteristic based system
because by increasing a base characteristicall related skills are
increased, since they are computed from this base. Sadly, as you have
rightly pointed out, this is a pretty crude system and has problems of
its own, as some skills will be dependant upon factors not modelled or
several of the factors together.
It's hard to see how to devise a system that takes all of this into
account without overcomplicating things. As a slight evolution of James'
ideas, it's possible to imagine a system with a number of base
statistics that have associated skills. Increasing the base statistic
results in a boost to all the skills associated with it, but this is
costlier than increasing a single skill. Increases in the skills would
count towards to a counter displayed beside the base skill that tells
you how many increases are required before that too increases. I think
this strikes a nice balance between modelling synergy, simplicity, and
allowing the modelling of specalisms - what are your thoughts?
> But then why have a strength *stat* at all? Why not just have traits?
> We could have a Muscle Bound trait which gives bonus to damage and a
> bonus to carrying capacity.
>
This is a great idea! It models those crazy old synergies just fine,
keeping it simple at the same time, and may give the player better
satisfaction because they can build up a clear picture of the character
they are building up. The only fault I can find with it is that it may
require lumping loads of statistics increases together to model all the
effects of such a trait, so you'd have a dirty great list of statistics
being affected by taking a trait along with amount they increase by, but
this could be eliminated, or at least severely reduced, by using a
carefully designed skill set. I like this a lot..
Max
>>I think that a traditional stat system provides a nice solid base for
>
> various
>
>>other ability scores to be derived from. Then, multiple other less
>
> tangible
>
>>factors, such as skill systems or talents can be added to further
>
> refine it.
>
> This is why I say stats are harmful! You do not need a solid base from
> which to derive other ability scores!
>
> What we usually see happen in Roguelikes is that the "less tangible"
> factors end up outweighing the "Stats" factors. Stats really just
> become another skill/trait.
>
> I claim it is best to *start* with the stats just being another
> skill/trait. This avoids putting them in a privelaged position that
> results in crazy stuff like stat gain potions.
ru
> 1) Stats are usually created early in game design and then adversely
> affect later game design
>
> This is the "Gosh, I better find a use for Charisma" problem. It is
> accentuated whenever a point-buy system is added to the statistic
> setup.
i agree. (i've never liked point buy systems, i much prefer templates)
> 2) Stats encourage needless orthogonality
>
> People designing statistic systems love orthogonality. They want all
> stats to be the "same" worth. If they have a mental stat, they need a
> corresponding physical stat. One point of strength is considered
> equivalent to one point of learning.
it's useful to have a set scale to compare a number with, so you can see
what's high and what's low. there's no reason for it to be numeric,
though.
do you remember the old Marvel Superheros rpg? with all that
remarkable/incredible/amazing stuff? i felt it worked very well (despite
or because of large granularity), because it was easier to visualise
incredible vs. amazing, as against 35 vs 45.
Fate (www.faterpg.com) takes an approach where you have a small number
of very granular stats, but you can make them up. i can see an approach
in a roguelike where the player could choose, "strong" twice and "tough"
once. another player could decide to be "weak", "fast", "sneaky",
"perceptive" and "accurate". that's very rough, but you get the idea.
> 3) Statistics are a hold over from the P&P world
well, this kind of thing is slowly eroding with more and more fast-play
systems coming along.
one problem is the roguelike/d&d connection, which weds roguelikes to
pretty much the most archaic p&p system, and all its assumptions.
> If you have dexterity affecting to-hit and trap evasion, why not just
> have a To-Hit statistic and an Evade Trap skill? If strength gives
> extra damage and carrying capacity, aren't we better off with a
> To-Damage stat and a Carrying Capacity stat?
i think that there are two general assumptions, right or wrong.
first is one of applicability. "strength" is a very general description
that is designed to apply to many tasks, as opposed to skills which
apply to a much narrower set of activities.
the old star wars d6 rpg from WEG used this approach explicitly: skills
were specialisations of a stat. if you didn't have a required skill you
used the value of the governing stat instead, but if you had the skill,
you just used the skill (due to their nature, a skill would always be
higher than its governing stat).
the other is the idea that stats represent some kind of intrinsic
quality of the character. this idea is often represented by the
advancements system, where skills are likely to change much more than
stats. stats in this sense reduce to knacks for certain sets of skills.
> 2) Make Secondary Stats Primary
>
> One thing you'll notice repeated above is the replacement of Primary
> Stats with Secondary Stats. Primary stats was primarily a book keeping
> measure of the P&P days. We do not need it any more. If you wish, you
> can always try and calculate an estimated primary stat from the
> concrete secondary stats to show the users. Ie, the users strength
> score could be reported as some weighted average of their To-Damage
> bonus and their Carrying Capacity.
quite a few console rpgs do this, listing "atk" and "def" and so on,
rather than strength or toughness.
--
ru
Brent Ross
Jeff Lait <torespon...@hotmail.com> wrote:
// 3) Statistics are a hold over from the P&P world
//
// Why are statistics orthogonal? Because that makes rolling up a
// character straightforward. Why do we use the same "roll against
// Strength" approach for all stats? Because that means players only have
// to learn one stat resolution system.
//
// Most importantly, why do we have rolls vs. stats at all? In P&P gaming
// sessions, the players and GM are working together to develop a
// narrative. When an obstacle is reached, one doesn't want to have to
// determine if success or failure is required for the story (the goal is,
// of course, to develop a story as a *result* of successes and failures).
// Often the nature of the obstacle is one off. The GM determines some
// modifiers and the player rolls against their stat.
//
// The question then arises: Why would we have such stat rolls in a
// roguelike? The entire gamut of possible actions is already known. The
// player can't do anything you haven't programmed, and if you have
// programmed it, you are already aware of the roll.
Actually, I find things to be exactly the opposite for pretty much the
reasons you're giving here. The co-operative narrative storytelling
of roleplaying sessions is best served by minimizing the amount of dice
rolling and stat checks (and the book keeping of some systems adds further
hinderences). The success of failure of an action is typically better
handled simply be what fits the character attempting, the preparations
taken, and the narrative. Some of my favourite systems have no dice
rolling (Amber) or encourage avoiding it (Time Lord, Fudge, Fung Shui).
When you add stats and dice rolling, a well planed narrative approach
can be completely ruined... boss vampires turned into lawn chairs,
dungeons quickly eliminated due to stone to mud being using on the floor
of the lake overtop of it, a lucky theif accidentally opening a door
that wasn't suppost to be open yet. The players can easily destroy a
week of GM preparation time with an oddball plan and a lucky roll. At
least when the GM rolls something unexpectly behind the screen they
can ignore it... but sometimes it's hard to explain to a player that
their incredible success roll doesn't matter.
Roguelikes on the other hand use the computer for GM (thus the book
keeping is largely handled and less of a issue), and have a strong
natural lean to rollplay over roleplay. We're along way from having
a computer that can apply an arbritary imaginative player's plan and
decide success merely on it's merit and a narrative worth. And so
computers have to take the simpler GM road... present some obstacles,
give the player's random chances of success, and let the fate of the
dice drag the narrative somewhere. However, the game still has to
be playable by humans, and there is such a thing as information
overload.
So, in short... dealing with the "entire gamut of options" is really the
realm of tabletop roleplaying. With human level intelligence in the GM,
the results of any possible player action can be evaluated to a result
without dice: Is what the character doing possible given their past and
description? Is it sensible? Does it improve the story? In RLs, you
understand all the options, only in that the game arbiter is extremely
limited in it's capacity to deal with the unexpected or create or even
judge narrative. That typically means out-sourcing the roleplaying to
the heads of the sentient players and using rollplaying chance methods to
help inspire that (otherwise we would be getting into the realm of CRPGs
with their very fixed narratives and corresponding lower replay value).
// If you have dexterity affecting to-hit and trap evasion, why not just
// have a To-Hit statistic and an Evade Trap skill? If strength gives
// extra damage and carrying capacity, aren't we better off with a
// To-Damage stat and a Carrying Capacity stat?
Simple... Strength and dexterity are merely representative of the
transferable parts of those skills. Strength is the measure of ability
to exert force and both damage and carrying involve the exerting of force.
Similar with dexterity which is a measure of physical control. The reason
you wouldn't want to go into a lot of things like a Carrying Capacity stat
would be information overload (which wouldn't be a better off situations).
When you have a bunch of things that aren't largely independant it can be
quite reasonable to tie them up into a single abstraction that simplifies
things (at the expense of some accuracy of the model). And so strength
isn't such a bad thing, it's just a higher level abstraction and a valid
choice.
// What To Do in a Post-Stats World
// --------------------------------
//
// POWDER is "statless" in this sense. Technically there are two stats
// (strength and smarts), but they are hidden from the user and only
// alterable through polymorph. As such, they can be considered just
// properties of the monster type (like whether a monster has a head or
// not) rather than AD&D style statistics.
//
// My suggestions for a post-stat game are:
//
// 1) Break orthogonality
//
// Stats should specify *units*. Max HP is measured in Hit Points.
// Carrying Capacity is in Kilograms. Number of spoken languages is
// measured in Languages. Statistics should be things you can measure -
// it is a rather ironic turn of fate that its use in roguelikes is to
// refer to the very things that can't be measured!
Acutally, they can be... Strength can be measured in exerted force,
which, via abstractions and other attributes (size, shape, equipment
being used) can be converted into carrying capacity or extra damage
from additional force no greater than the concept of "hit point" or
carrying capacity measured merely in "kilograms". Similarly, dexterity,
intelligence, wisdom, and constitution can all be tested and given
measurements with scientific methods. Charisma is also possible, but
the results would be a bit more subjective. [1]
The stat abstraction isn't bad in itself... taking a set of values
and assigning a range of possibilitis to each is one way to simplify
the system. Sure the dexterity factor applied to-hit then implies a
certain range of dexterity factor in trap avoidance/disarming. That
might not seem appropriate in all cases, but there are other factors
involved... dexterity is merely the transferable one. Centaurs have
large bodies that hinder their ability to dodge bolt traps... should
that be represented by having a separation of trap avoidance and to-hit
(which isn't as affected)? But isn't that just the same problem
that getting rid of dexterity and splitting it in the first place (it
overloads and abstracts away a differentiation into a single number)?
Personally, I think it's better to target the causes... in this case a
highly dexterous centaur can easily manipulate things, but their body
size and shape prevents that from being a factor in agility.
// 2) Make Secondary Stats Primary
//
// One thing you'll notice repeated above is the replacement of Primary
// Stats with Secondary Stats. Primary stats was primarily a book keeping
// measure of the P&P days. We do not need it any more. If you wish, you
// can always try and calculate an estimated primary stat from the
// concrete secondary stats to show the users. Ie, the users strength
// score could be reported as some weighted average of their To-Damage
// bonus and their Carrying Capacity.
Which is one way to look at it. I prefer to have my primary things be the
transferable factors. Carrying capacity isn't as transferable a thing
as stength. To back calaculate them as secondary is pointless... if
you ever needed to do such a thing (beyond merely for display), then
it's a sign that your design is backwards. You probably should have
decided on the equivalence classes between the different representations
(ie associate carrying capacities with extra force damage) and apply
adjustments to differentiate them appropriately (centaurs get extra
capacity from size and shape, some weapons are better at focusing force
than others). The primaries should be more general and abstract than
the secondaries that are derived from them.
All said, there are a lot of good ideas and concepts in the original
post... it's certainly a take and approach that can work. However,
I simply cannot agree that stats are necessarily a bad abstraction and
something that should definitely be avoided. Stats, used correctly
and avoiding the issues and examples of poor implementation brought up
in the post, are as valid an abstraction as the split presented (since
they are largely the same thing, the only real difference being in the
level of abstraction).
Brent Ross
[1] We could try to measure it in milli-yorks (the amount of charisma
required to inspire ten men to march up and down a hill), but that's not
really testable in a lab. For those who prefer Beauty, that's measured
in milli-helens (amount of beauty required to launch one ship), which
is also a bit tricky for accurate lab readings.
Brent Ross
Jeff Lait <torespon...@hotmail.com> wrote:
// Max Bolingbroke wrote:
// > Jeff Lait wrote:
// > > 1) Break orthogonality
// > >
// > > Carrying Capacity is in Kilograms. Number of spoken languages is
// > > measured in Languages. Statistics should be things you can measure
// -
// > > it is a rather ironic turn of fate that its use in roguelikes is to
// > > refer to the very things that can't be measured!
// >
// > I found your post very interesting, but now I find myself wondering
// how
// > related stats should be handled: carrying capacacity and how hard I
// can
// > brain things with a club, both, presumably, being related to muscle
// size
// > should probably go up together to an result.
//
// Why do you assume the two things are related to one "muscle size"
// parameter?
Because there is at least some transferable factor between the two...
the this that absolutely cannot be assumed is that they are independant.
// Carrying capacity could reflect a better packing strategem. A skilled
// backpacker can carry a lot more for a lot longer than an equally fit
// newbie.
Which would be a skill factor. Which is different simply for the reason
that in such a situation it's possible that the skilled backpacker
could be capable of packing so efficient that neither could carry it.
At some point an amount of mass has to be lifted and carried, and no
amount of skill in orderliness, balance, or efficiency is going to undo
that reality and need to exert force.
// Likewise, your damage bonus for hitting with a club is likely
// moderated by a lot more than muscle mass!
Ah, but who's dictating that it's the only factor? Or how it's even
applied? On thing is sure from physics, however, factors like
leverage and speed are important to weapons like two-handed swords. [1]
At some point there's the question of the character's ability to
swing the sword at speed (or perhaps effectively at all) and achieve
better success... that's where the the strength factor lies: in the
ability to physically exert force to achieve greater impact.
// Flexibility is likely as
// important in order to be able to make use of one's mass effectively.
Which is a different unrelated factor... one of which is more important
to weapons like rapiers where this is even more the case. [2] This is
the "quality of hit" factor, which also matters, but it's important to
notice that a high impact glancing blow and a no impact accurate blow
are probably about equally ineffective. At the very least, you can't
apply any sort of finesse to a massive club that you can hardly lift.
At some point you need to have the strength to actually be able to apply
the forces you desire to the weapon.
// The reason why we want to relate both back to "Muscle Mass" is because
// we have been trained to create that category thanks to our years of D&D
// influence.
That could be... however, I can't see tossing it out automatically
because some implementations are bad. Force is a very important part
of real world physics... strength can be a reasonable abstraction of
that if used correctly.
Brent Ross
[1] Twohanded swords evolved over time along with armour. The longer
blades provided more leverage which combined with the mass of the metal
allowed for staggering amounts of impact. The extra mass and leverage
required more strength (and endurance), but that was partially offset
by better metallurgy (which allowed for thiner, and lighter blades with
better edges to compensate).
[2] Rapiers came after twohanded swords, largely because armour had
become fairly obsolete against strong bows. Stength has much less
of a role, but does have a small support factor in blade control.
Ray Dillinger
>
> I am not trying to deny that. As I said at the beginning, I am arguing
> against AD&D style "Primary Stats". Not against statistics, or against
> traits, or against skills.
>
> Raw muscle mass definitely has an effect. It, however, is not some
> "special" effect.
>
Actually, I think it is. What I'm using is a class system
where someone's class, and primary stats, and actions,
determines the proportions of experience spent on various
stats and skills in an underlying point based system.
And primary stats have the further effect of modifying
the price of certain skills so that the experience spent
on them goes further or not as far.
So -- as a fighter, there is a minimum you'll be spending
per level on strength and speed. If your primary
weapon is a "light" sword or a bow, there's more that
you'll be spending on speed, where if your primary
weapon is a club, you'll be spending more on strength.
Natch, you'll also be spending a minimum amount on your
primary weapon proficiency. But each point of strength
or nimbleness changes the cost of your primary weapon
proficiency, making the points already spent go further.
So there's a synergistic effect.
And finally, there's a point of diminishing returns
with statistics; the price of increasing a statistic
asymptotically approaches infinity as the statistic
itself asymptotically approaches double its starting
value. Eventually there's a cutoff, where the points
that would otherwise be spent won't even get one point
of a stat - and spending on stats stops.
I did this as a way of fighting "thousanduplet syndrome" -
I'm deliberately using stats to differentiate characters
within their classes.
>>The problem that
>>I am trying to articulate is that of synergy: getting better in some
>>things will mean you get better at related things.
>
>
> Exactly! Weight training will improve your carrying capacity and
> abillity to wield weapons effectively. This doesn't make weight
> training a primary stat, however.
"Pack training" is a skill whose price depends mostly on
strength, somewhat on speed (agility) and a little bit on
intelligence. It reduces endurance penalties for carrying
around your gear, and can up to double the amount of gear
you carry before being "burdened," etc.
Once you have spent n experience points on pack training,
raising any of those stats increases the amount of skill
gained per n experience points, thereby raising your skill
with pack training. And for as long as it can be improved,
the amount spent per level on pack training mostly depends
on the weight of your pack.
Improving strength and speed is of course going to do the
same thing to your weapon skills at the same time.
> So are you suggesting that there should be a Ck (Cooking) statistic
> beside Str and Wis?
Nah. Different order. Cooking is a skill. The primary stats
are what determines the experience price of skills.
> Increasing Carrying Capacity would *never* give you a ToDamage bonus.
> The entire point is to make so-called secondary statistics primary to
> give us the freedom to escape assumptions like "Carrying Capacity is a
> result of Strength"
I think you're dismissing a good idea because of an overly
simplistic (bad) implementation of that idea in D&D. I'm
perfectly happy to have the pack-training skill (which the
user never sees) determine burdened, overburdened, etc,
levels (which the character does see) and in turn have its
cost (and therefore level) determined by strength (and
agility, and intelligence) and practice with carrying gear.
> This also results in the Stat Gain system. When we build a bunch of
> "base characteristics" we end up with point-buy, stat gain potions, and
> the belief that these stats are fungible.
Eh. Or not. Depends on the game designer, doesn't it? I
mean, it's an easy thing to fall into if you engage in faint
and fuzzy thinking, but there's nothing inherent in the
existence of base stats that forces people to commit errors
of logic. I mean, I know for sure that different stats have
different value. They have different value during the
character design process and throughout the game, and I
*don't* allow players to exchange one for another on a
point-per-point basis.
I also know that the point-buy system I'm modelling, if
subject to complete player control and choice, would
result in wildly unbalanced characters intended to
abuse the synergies I'm awarding for stats. But taking
that level of control from the player, automating it to
behave in some "reasonable" fashion, and removing the
esoterica like "pack training skill" from the interface
as bookkeeping details that the player's not interested
in, results in interesting and varied characters without
opening up the point-buy system to "rules raping."
Bear
Brent Ross
Jeff Lait <torespon...@hotmail.com> wrote:
// Brent Ross wrote:
// > In article <1103010617....@c13g2000cwb.googlegroups.com>,
// > Jeff Lait <torespon...@hotmail.com> wrote:
// > // Max Bolingbroke wrote:
// > // Carrying capacity could reflect a better packing strategem. A
// skilled
// > // backpacker can carry a lot more for a lot longer than an equally
// fit
// > // newbie.
// >
// > Which would be a skill factor. Which is different simply for the
// reason
// > that in such a situation it's possible that the skilled backpacker
// > could be capable of packing so efficient that neither could carry it.
// > At some point an amount of mass has to be lifted and carried, and no
// > amount of skill in orderliness, balance, or efficiency is going to
// undo
// > that reality and need to exert force.
//
// Measuring Carrying Capacity with Kilograms was obviously foolish of me,
// as then there is no sense in having a packing strategy effect it. I
// should have said Generic Encumberance Units :>
It's understandable given the typical roguelike representation of carrying
capacity as purely mass (which is typically not very well done, and as
you've said before, is largely used to have strength do something).
// I really don't see the distinction, at the game level, between skills
// and traits and stats. There may a difference to the user in how they
// are represented, and how easy they are to change. Ie, saying "packing
// is a skill" isn't any different than saying "lifting heavy weights is a
// skill". Both are things which you acquire with lots of practice :>
Certainly there's not a lot of difference at some level since they
all qualify as factors which get applied via mechanics to get results.
The difference only really comes out (and then only if you want it to)
in the design of how the physical laws of the game world work.
Just about everything can be broken down into having factors that come
from transferable talents, practised skill in the thing being done,
or general knowledge of the subject at hand. Bringing any one of those
things to the task at hand is better than bringing none at all (and the
more, the better). The way I look at it, strength is a transferable
talent here... the strong can compensate for some poor organisation
though raw power. Experience in packing is different in that it's a skill
factor... skills tend to be less transferable than talents and knowledges,
as well as more specific and thus more useful when they apply. Or just
rack it up to overpowering with brawn versus compensating with brain if
you prefer.
As for lifting heavy weights... there is such a thing as raw untrained
strength, I should know because I leg pressed 610 lbs on my first day
(and could possibly have done the stack if I had really tried... I took
it slow and did that a little while later... that strength came from
biking up a rather large hill on a regular basis). A lot of it also
comes from willpower and confidence as well (average people have been
known to move things weighing more than a ton when under pressure)... all
of which can be trained, but which some people more naturally have and
can bring to the table immediately. However, you are quite right in
that I wasn't going to be doing any impressive clean and jerk lifts
on the first day... that requires a lot more skill and practise to get
right because it's far trickier and requires technique (due to the fact
that balance and timing are very important there). So again we have
a balance between the concept of the raw, transferable power and the
practised, specific skill. And so lifting heavy weights has factors
of both... add in some knowledge from a learned coach and you can push
the limits. But any one of them alone won't get you to that level:
practicing the technique of lifting doesn't generate the raw muscle
power you need... you need to use different exercises to gain that,
and those exercises of course, won't get you the specific technique you
need... and all of this is improved and made more effective simply by
the knowledge of what the most effective methods are.
// > // Likewise, your damage bonus for hitting with a club is likely
// > // moderated by a lot more than muscle mass!
// >
// > Ah, but who's dictating that it's the only factor? Or how it's even
// > applied? On thing is sure from physics, however, factors like
// > leverage and speed are important to weapons like two-handed swords.
// [1]
// > At some point there's the question of the character's ability to
// > swing the sword at speed (or perhaps effectively at all) and achieve
// > better success... that's where the the strength factor lies: in the
// > ability to physically exert force to achieve greater impact.
//
// Agreed. Which shows that strength is one of many factors. Indeed, in
// most roguelikes it seems the Strength modifier ends up being drowned by
// other modifiers anyways. The so-called "Primary Stat" ends up having
// only an ornamental role for ascension level characters.
Depends on what you want from (or perhaps better put as how you define)
"primary", I suppose. If by "primary" you mean most directly significant
stats, then the problems you mentioned are big issues, unless, of course,
the level of abstraction is at that high level. There's not much problem
with a system where strength has only five levels of strength (very
weak, weak, average, strong, very strong) and each is very specifically
defined to it's effects. This makes it easy to control the player into
playing their character in the designed way (for example, if only the
very strong can use the largest most damaging weapons then that becomes
their special priviledge... if those weapons are a real advantage it
becomes their path). Since the aqbstraction is so high, it doesn't
matter so much if the things represented by strength are so different.
Another way to look at "primary", however, is as the base elemental
attributes (ie similar to primary colours). As such, they're not
necessarily the most important factors so much as transferable natural
factors and limitations. As such they typically encourage certain
strategies and ease specific avenues of advancement in the beginning
and midgame (until specific skills become dominant). As such, what you
want is the best set of factors for combining to generate the results
you want (and that's typically a minimal set... as too many factors cause
problems [1]). This doesn't stop one from increasing the weight of
such stats for the end game by having them more effect "capability"
than "ability" or "magnitude" [2], by having stat requirements for the
very best things. In this way the elemental nature of the primary stats
influences the development paths in a subtle way, and then, at the end
becomes relevant again in limiting the access to some of the top level
things (albeit in a blunt fashion).
Brent Ross
[1] If you have too many factors involved in calculating secondary
values here, you can easily end up with having to apply all or most
of them in very small and meaningless ways. In general, it's better
to ignore some factors that seem correct merely from "common sense"...
the larger weight of the remaining factors will be better felt by the
player and add more to the value of the game than a system where large
changes are needed in several areas before any benefit is achieved.
[2] This comes from a model I sometimes use to think about how factors
should be applied in a RL situation (Capability-Ability-Magnitude).
Capability refers to being able to actually do something (ie a character
cannot attack with a twohanded sword if they do not have a twohanded
sword, or are too small to use one, or don't have enough strength to
swing it effectively). Ability is the factor that measures our ability
to acomplish things we're capable of, and Magnitude is the weight of the
result (both of these might also have factors for being too small or
too weak with twohanded swords, but the intent is that the two-handed
sword is feasible here so those factors are small and not absolute).
It's useful to note that Ability tends to be a better place to apply
transferable traits, and Magnitude a better place for specific ones.
This is an easy way to allow the player to change their character's path
to a related one without the character suddenly becoming very incompetent
(they'll just be a little less effective). It also gains the designer
benefit in that magnitude tends to be something you want to rein in,
and by having that largely the demain of specific skills that need to
be independantly developed helps put down some limits.
Antoine
Jeff Lait wrote:
> [snip]
>
> Stats should specify *units*. Max HP is measured in Hit Points.
> Carrying Capacity is in Kilograms. Number of spoken languages is
> measured in Languages. Statistics should be things you can measure -
> it is a rather ironic turn of fate that its use in roguelikes is to
> refer to the very things that can't be measured!
>
> Using units ensures that the user can understand the stat without
cross
> referencing a manual or spoiler. It also ensures the user realizes
> they are not fungible. They won't wonder if they can trade in 10
kilos
> of carrying capacity for 10 spell slots.
One disadvantage of the 'units' system is that it makes the attributes
of the character very clear and explicit, removing any air of mystery.
This may not be what the game designer wants.
If I start a game and find I have a Sanity statistic of 17 or a
Perception of 14, I may think "kewl, I wonder what that does". What's
the alternative - "Number of times you can see an eldritch thing
without going mad: 6 times" and "Distance you can see: 12 spaces"?
Seems a bit pedestrian.
In similar vein, I would find a Strength statistic of 15 more
interesting / mood enhancing than the rather mundane "you can carry 80
kilograms".
Antoine
Ray Dillinger wrote:
> Mylon wrote:
>
> >
> > I want to know your thoughts on pen and paper RPing systems.
> > Particulally two: D&D and GURPS.
>
> D&D classes force a particularly extreme character
> differentiation, because D&D is, first and foremost,
> a social activity. It is not supposed to make sense
> to go adventuring with one character. The classes
> are interrelated in strengths and weaknesses and need
> each other. D&D is designed to help geeks build social
> structures and friendships. To some extent, therefore,
> the way D&D classes work is not entirely appropriate
> to a one-player game.
>
Hence GUILD, a game with a party of multiple characters, strongly
differentiated by class.
For example, a warrior will never be able to use spellbooks, wands, or
rods, no matter how high their level, and a mage will never be a good
melee fighter (I would think a high level wizard who hadn't chosen
their equipment with melee combat in mind might well not be a match for
a small group of orcs in melee).
Jeff Lait
Max Bolingbroke wrote:
> Jeff Lait wrote:
> > 1) Break orthogonality
> >
> > Stats should specify *units*. Max HP is measured in Hit Points.
> > Carrying Capacity is in Kilograms. Number of spoken languages is
> > measured in Languages. Statistics should be things you can measure
-
> > it is a rather ironic turn of fate that its use in roguelikes is to
> > refer to the very things that can't be measured!
>
> I found your post very interesting, but now I find myself wondering
how
> related stats should be handled: carrying capacacity and how hard I
can
> brain things with a club, both, presumably, being related to muscle
size
> should probably go up together to an result.
Why do you assume the two things are related to one "muscle size"
parameter?
Carrying capacity could reflect a better packing strategem. A skilled
backpacker can carry a lot more for a lot longer than an equally fit
newbie. Likewise, your damage bonus for hitting with a club is likely
moderated by a lot more than muscle mass! Flexibility is likely as
important in order to be able to make use of one's mass effectively.
The reason why we want to relate both back to "Muscle Mass" is because
we have been trained to create that category thanks to our years of D&D
influence.
> As far as I can see, this could be done explicitly (by increasing
> related stats as soon as a stat is increased) or by having some sort
of
> resultant statistic which is affected by its relatives but cannot be
> changed directly. In the example above, for instance, you would have
> Strength as a resultant statistic which determines your ability to
> batter wolves sensless, and it would be increased by increasing your
> carrying capacity, rock-pushing or digging scores. This feels a bit
like
> a cludge to me though - I'd be interested in hearing others
perspectives.
While I mentioned Strength as a resultant stat, I certainly would not
suggest one use that resultant stat to determine wolf-battering
ability. Rather, one's wolf battering ability would be a seperate
stat. Say, "Damage Bonus". The Damage Bonus would feed back into
calculating the approximate "Str" score which is used only for display.
(Ie, let the user get some sense of how strong the character is)
Jeff Lait
> On 2004-12-14 08:50:17, "Jeff Lait" <torespon...@hotmail.com>
wrote:
>
> > Why do you assume the two things are related to one "muscle size"
> > parameter?
>
> While there are plenty of factors affecting a character's carrying
capacity and
> the mount of damage they deal, raw strength is certainly one of them.
Really,
> when it comes to lifting things, little ought to count more than
strength, I
> think.
Carrying capacity is not being able to lift things.
> Damage varies more depending on the type of weapon, but you can't
deny
> that a musclebound hulk will hit things a heck of a lot harder with
that maul
> than someone of average build.
If the musclebound hulk has not learned how to swing properly, then the
average build person will out damage them. Technique has a huge effect
on the amount of force that a blow will land with. An untrained muscle
bound hulk punches with his arm. The skilled average build person
punches with their entire body.
> > Carrying capacity could reflect a better packing strategem. A
skilled
> > backpacker can carry a lot more for a lot longer than an equally
fit
> > newbie. Likewise, your damage bonus for hitting with a club is
likely
> > moderated by a lot more than muscle mass! Flexibility is likely as
> > important in order to be able to make use of one's mass
effectively.
>
> I personally rather that things like being able to pack better are
represented
> as skills or traits. Something like a 'backpacking' trait that
increased
> carrying capacity by 50%, for example. Combat could also be affected
by various
> types of bonus traits in a similar fashion.
But then why have a strength *stat* at all? Why not just have traits?
We could have a Muscle Bound trait which gives bonus to damage and a
bonus to carrying capacity.
> Personally, I think dexterity ought
> to be factored into damage calcutations for many types of weapons in
addition
> to strength.
Which goes to show how useless the AD&D stats are. If strength affects
to hit and weapon recovery (strong characters have more control over
heavy weapons) and dexterity affects damage done, we start to wonder
why we have these different stats at all.
> > The reason why we want to relate both back to "Muscle Mass" is
because
> > we have been trained to create that category thanks to our years of
D&D
> > influence.
>
> True to an extent, however the categories as they exist weren't
invented for no
> reason. Not to say that they mightn't bear re-examining and some
alteration,
> but many of them do have a pretty decent grounding in logic, in my
opinion.
The reason they were invented was to simplify the book keeping for P&P
play. It is easier to say "I have Strength 18" then "I have +3 damage,
carrying capacity of 200 stones, and can bend bars on a 15"
Bookkeeping is no longer an excuse in a computerized implementation.
> I think that a traditional stat system provides a nice solid base for
various
> other ability scores to be derived from. Then, multiple other less
tangible
> factors, such as skill systems or talents can be added to further
refine it.
This is why I say stats are harmful! You do not need a solid base from
which to derive other ability scores!
What we usually see happen in Roguelikes is that the "less tangible"
factors end up outweighing the "Stats" factors. Stats really just
become another skill/trait.
I claim it is best to *start* with the stats just being another
skill/trait. This avoids putting them in a privelaged position that
results in crazy stuff like stat gain potions.
Jeff Lait
> Jeff Lait wrote:
> >>Damage varies more depending on the type of weapon, but you can't
> >
> > deny
> >
> >>that a musclebound hulk will hit things a heck of a lot harder with
> >
> > that maul
> >
> >>than someone of average build.
> >
> >
> > If the musclebound hulk has not learned how to swing properly, then
the
> > average build person will out damage them. Technique has a huge
effect
> > on the amount of force that a blow will land with. An untrained
muscle
> > bound hulk punches with his arm. The skilled average build person
> > punches with their entire body.
>
> But you can hardly deny that that raw muscle mass is having an effect
-
> there are just other factors to take into consideration.
I am not trying to deny that. As I said at the beginning, I am arguing
against AD&D style "Primary Stats". Not against statistics, or against
traits, or against skills.
Raw muscle mass definitely has an effect. It, however, is not some
"special" effect.
> The problem that
> I am trying to articulate is that of synergy: getting better in some
> things will mean you get better at related things.
Exactly! Weight training will improve your carrying capacity and
abillity to wield weapons effectively. This doesn't make weight
training a primary stat, however.
> Evidence for this
> effect is everywhere - for example, the ability to cook a nutritous
> mushroom soup improves your ability to make that tomato soup you like
> the look of because you have mastered the use of a hob and saucepan.
So are you suggesting that there should be a Ck (Cooking) statistic
beside Str and Wis?
> This CAN be represented in your system, since you can have stat
> increases in some skills cause increases in others,
That would not be how I would represent it.
Increasing Carrying Capacity would *never* give you a ToDamage bonus.
The entire point is to make so-called secondary statistics primary to
give us the freedom to escape assumptions like "Carrying Capacity is a
result of Strength"
> This is mitigated to some extent with a characteristic based system
> because by increasing a base characteristicall related skills are
> increased, since they are computed from this base. Sadly, as you have
> rightly pointed out, this is a pretty crude system and has problems
of
> its own, as some skills will be dependant upon factors not modelled
or
> several of the factors together.
This also results in the Stat Gain system. When we build a bunch of
"base characteristics" we end up with point-buy, stat gain potions, and
the belief that these stats are fungible.
> It's hard to see how to devise a system that takes all of this into
> account without overcomplicating things. As a slight evolution of
James'
> ideas, it's possible to imagine a system with a number of base
> statistics that have associated skills. Increasing the base statistic
> results in a boost to all the skills associated with it, but this is
> costlier than increasing a single skill. Increases in the skills
would
> count towards to a counter displayed beside the base skill that tells
> you how many increases are required before that too increases. I
think
> this strikes a nice balance between modelling synergy, simplicity,
and
> allowing the modelling of specalisms - what are your thoughts?
I would rather there be a hierarchy of skills where base skills will
improve derived skills. Ie, both mushroom and tomatoe soup
construction derive from some cooking skill. Having one's base cooking
proficiency based on Intelligence or what not seems foolish.
> > But then why have a strength *stat* at all? Why not just have
traits?
> > We could have a Muscle Bound trait which gives bonus to damage and
a
> > bonus to carrying capacity.
>
> This is a great idea! It models those crazy old synergies just fine,
Woohoo! We have communication :>
> keeping it simple at the same time, and may give the player better
> satisfaction because they can build up a clear picture of the
character
> they are building up. The only fault I can find with it is that it
may
> require lumping loads of statistics increases together to model all
the
> effects of such a trait, so you'd have a dirty great list of
statistics
> being affected by taking a trait along with amount they increase by,
but
> this could be eliminated, or at least severely reduced, by using a
> carefully designed skill set. I like this a lot..
Thank you. This was the primary point I was trying to get at by saying
D&D style Primary Statistics are Harmful.
Jeff Lait
True. I wouldn't claim they are dependent. I phrased that very
poorly. A better phrasing may be:
"Why do you assume the two things are primarily derived from one muscle
size parameter?"
It is the notion of *primary* statistics that I trying to attack here.
> // Carrying capacity could reflect a better packing strategem. A
skilled
> // backpacker can carry a lot more for a lot longer than an equally
fit
> // newbie.
>
> Which would be a skill factor. Which is different simply for the
reason
> that in such a situation it's possible that the skilled backpacker
> could be capable of packing so efficient that neither could carry it.
> At some point an amount of mass has to be lifted and carried, and no
> amount of skill in orderliness, balance, or efficiency is going to
undo
> that reality and need to exert force.
Measuring Carrying Capacity with Kilograms was obviously foolish of me,
as then there is no sense in having a packing strategy effect it. I
should have said Generic Encumberance Units :>
I really don't see the distinction, at the game level, between skills
and traits and stats. There may a difference to the user in how they
are represented, and how easy they are to change. Ie, saying "packing
is a skill" isn't any different than saying "lifting heavy weights is a
skill". Both are things which you acquire with lots of practice :>
> // Likewise, your damage bonus for hitting with a club is likely
> // moderated by a lot more than muscle mass!
>
> Ah, but who's dictating that it's the only factor? Or how it's even
> applied? On thing is sure from physics, however, factors like
> leverage and speed are important to weapons like two-handed swords.
[1]
> At some point there's the question of the character's ability to
> swing the sword at speed (or perhaps effectively at all) and achieve
> better success... that's where the the strength factor lies: in the
> ability to physically exert force to achieve greater impact.
Agreed. Which shows that strength is one of many factors. Indeed, in
most roguelikes it seems the Strength modifier ends up being drowned by
other modifiers anyways. The so-called "Primary Stat" ends up having
only an ornamental role for ascension level characters.
> // Flexibility is likely as
> // important in order to be able to make use of one's mass
effectively.
>
> Which is a different unrelated factor... one of which is more
important
> to weapons like rapiers where this is even more the case. [2] This
is
> the "quality of hit" factor, which also matters, but it's important
to
> notice that a high impact glancing blow and a no impact accurate blow
> are probably about equally ineffective.
I'm referring here more to the ability to not just use your arm
strength in the hit, but being able to use your entire body. I bring
it up precisely because it shows the "ToDamage" statistic is not
primarily derived from any of the standard D&D stats. So why do we
call those stats primary?
> // The reason why we want to relate both back to "Muscle Mass" is
because
> // we have been trained to create that category thanks to our years
of D&D
> // influence.
>
> That could be... however, I can't see tossing it out automatically
> because some implementations are bad. Force is a very important part
> of real world physics... strength can be a reasonable abstraction of
> that if used correctly.
I'm not tossing out a "Muscle Mass" statistic. I'm tossing out the
"Strength is a Primary Stat". And, more precisely, the idea that there
*are* primary stats.
Jeff Lait
I loved that system of naming! Not due to visualization (incredible vs
amazing has no relative image in my mind. Incredible Hulk vs Amazing
Spiderman might imply Incredible was stronger, but I'm sure it could
also start a flame war :>), but because of the granularization and
introduction of a convenient shorthand.
POWDER's strength and smarts stats were intentionally granular and
named rather than just being numbers. Ie, it is SMARTS_FOOL, not
SMARTS_3.
> Fate (www.faterpg.com) takes an approach where you have a small
number
> of very granular stats, but you can make them up. i can see an
approach
> in a roguelike where the player could choose, "strong" twice and
"tough"
> once. another player could decide to be "weak", "fast", "sneaky",
> "perceptive" and "accurate". that's very rough, but you get the idea.
That sounds very much like what I'd like to see in a roguelike. One
would have to have a fixed list of stats, but it would help break away
from the fungiblity aspect. (Especially if a template system is used
rather than a point buy, the latter suggesting that strong is
equivalent to sneaky)
> > 3) Statistics are a hold over from the P&P world
>
> well, this kind of thing is slowly eroding with more and more
fast-play
> systems coming along.
>
> one problem is the roguelike/d&d connection, which weds roguelikes to
> pretty much the most archaic p&p system, and all its assumptions.
True. I'm being unfair to P&P, which as you point out has advanced.
> > If you have dexterity affecting to-hit and trap evasion, why not
just
> > have a To-Hit statistic and an Evade Trap skill? If strength gives
> > extra damage and carrying capacity, aren't we better off with a
> > To-Damage stat and a Carrying Capacity stat?
>
> i think that there are two general assumptions, right or wrong.
>
> first is one of applicability. "strength" is a very general
description
> that is designed to apply to many tasks, as opposed to skills which
> apply to a much narrower set of activities.
Roguelikes, however, tend to have a small number of tasks.
> the other is the idea that stats represent some kind of intrinsic
> quality of the character. this idea is often represented by the
> advancements system, where skills are likely to change much more than
> stats. stats in this sense reduce to knacks for certain sets of
skills.
Of course, this thread is started precisely because stats tend to
change to the stat caps in roguelikes :> I think in Nethack your stats
change by more units than your skills in an average ascension game :>
I'm not convinced we need to model intrinsic qualities seperate from
skills. Related skills should be given base skills which allow one to
advance across the entire field if this sort of synergy is desired.
(This also avoids silly questions like whether Str or Dex should be
used for rope climbing. IF there was an Athletics skill for how
athletic the character was which influenced rope climbing, mountain
climbing, etc, the problem would evaporate)
The stats imply default skill levels I feel is often an attempt to
justify the stats more than attempt to fix skills.
> > 2) Make Secondary Stats Primary
> >
> > One thing you'll notice repeated above is the replacement of
Primary
> > Stats with Secondary Stats. Primary stats was primarily a book
keeping
> > measure of the P&P days. We do not need it any more. If you wish,
you
> > can always try and calculate an estimated primary stat from the
> > concrete secondary stats to show the users. Ie, the users strength
> > score could be reported as some weighted average of their To-Damage
> > bonus and their Carrying Capacity.
>
> quite a few console rpgs do this, listing "atk" and "def" and so on,
> rather than strength or toughness.
Exactly. ADOM also does this, if by accident, when it allows you to
see your weapons To Hit and To Damage bonus. People use this rather
than their Strength for determining their weapon stats.
Jeff Lait
Mike Blackney
>
> I would rather there be a hierarchy of skills where base skills will
> improve derived skills. Ie, both mushroom and tomatoe soup
> construction derive from some cooking skill.
Wouldn't the existence of base skills have you end up with point-buy,
skill gain potions, and the belief that these skills are fungible? :)
> Having one's base cooking proficiency based on Intelligence or what
> not seems foolish.
That's because intelligence as a stat is foolish.
But base proficencies aren't. Give me two poor twins who have been
separated and locked in tight, dark cells for all of their eighteen
years of life and when you let them out I guarantee that the better
packer will have higher strength, more inventiveness or better luck.
Then lock the better packer back in a cell for another five years while
we teach the other how to pack and lift bags of loot, and the tables
will possibly be turned. /Possibly/.
This is why skills and stats make sense to be distinct.
--
michaelblackney at hotmail dot com
http://aburatan.sourceforge.net/
Latest version 0.95 2-5-4
Jeff Lait
> "Jeff Lait" <torespon...@hotmail.com> communicated:
> >
> > I would rather there be a hierarchy of skills where base skills
will
> > improve derived skills. Ie, both mushroom and tomatoe soup
> > construction derive from some cooking skill.
>
> Wouldn't the existence of base skills have you end up with point-buy,
> skill gain potions, and the belief that these skills are fungible? :)
Except in this case the developer will be able to ensure that if they
want fungibility, it is sensible to have fungibility.
The groupings are driven by the skills, not the skills driven by the
groupings.
> > Having one's base cooking proficiency based on Intelligence or what
> > not seems foolish.
>
> That's because intelligence as a stat is foolish.
But so is strength. What stats are we left with then?
> But base proficencies aren't. Give me two poor twins who have been
> separated and locked in tight, dark cells for all of their eighteen
> years of life and when you let them out I guarantee that the better
> packer will have higher strength, more inventiveness or better luck.
>
> Then lock the better packer back in a cell for another five years
while
> we teach the other how to pack and lift bags of loot, and the tables
> will possibly be turned. /Possibly/.
I think 5 years of training would easily offset 5 years of being locked
in a box.
> This is why skills and stats make sense to be distinct.
But, how are they different? If you discount stats like intelligence,
there seems to be no difference between a Strength stat and a Cooking
skill. Sure, cooking is more narrow, but there is not any reason why
we can't have broad skills (Ie: Athletics)
I often see zany things written like "Skills can change faster than
stats..." which really make me wonder what people are thinking of.
What is the difference between working hard at the gym to train up your
strength, or jogging to train your endurance, from writing roguelikes
to train your C++? Skills are generally trained on the order of
*years*. Given *years* to work with, I think it is not unreasonable to
expect stats to be changed as greatly. (I think people get confounded
by the "First 90% is 10% of the effort effect. One can very swiftly go
from zero programming to some programming. Since there is not much of
an equivalent with physical stats (we already got to "some strength" a
while ago), we conclude that since it is easy to get another 10%
programming skill when one a small ability, and hard to get 10%
strength, strength must be somehow different. We forget that if we
spend 10 years programming, that additional 10% becomes significantly
harder to achieve)
One might want a system to represent "quick learner". Ie, one
character may be faster to acquire a new combat move than another. Or
one might want a system to allow differential caps to skill. Ie, one
character can never really get the hang of the longsword. Such a
system belongs entirely outside of the traditional AD&D statistics,
however. Consider how Nethack accomplishes this with class specific
weapon restrictions.
The "primary stat bar" is a hangover from AD&D which should be
abolished. Sooner we fill those characters with actual relevant data,
the better.
Mike Blackney
>
> Mike Blackney wrote:
> >
> > "Jeff Lait" <torespon...@hotmail.com> communicated:
> > >
> > > I would rather there be a hierarchy of skills where base skills
> > > will improve derived skills. Ie, both mushroom and tomatoe
> > > soup construction derive from some cooking skill.
> >
> > Wouldn't the existence of base skills have you end up with
> > point-buy, skill gain potions, and the belief that these skills
> > are fungible? :)
>
> Except in this case the developer will be able to ensure that if they
> want fungibility, it is sensible to have fungibility.
>
> The groupings are driven by the skills, not the skills driven by the
> groupings.
Why can't I use this intelligent approach to stat design? How do base
stats necessarily differ from base skills?
> > > Having one's base cooking proficiency based on Intelligence or
> > > what not seems foolish.
> >
> > That's because intelligence as a stat is foolish.
>
> But so is strength. What stats are we left with then?
Okay, I take it back. Intelligence is a poor name for a stat and a
poor name to give in an example. If you're talking about D&D stats,
cooking proficiency more rightly should come under wisdom or even
charisma. If you're talking about a single intelligence stat, then my
point remains the same.
And I disagree that strength is a foolish stat.
> > But base proficencies aren't. Give me two poor twins who have been
> > separated and locked in tight, dark cells for all of their eighteen
> > years of life and when you let them out I guarantee that the better
> > packer will have higher strength, more inventiveness or better
> > luck.
> >
> > Then lock the better packer back in a cell for another five years
> > while we teach the other how to pack and lift bags of loot, and
> > the tables will possibly be turned. /Possibly/.
>
> I think 5 years of training would easily offset 5 years of being
> locked in a box.
Even if the one let out of the box to train is two feet shorter and has
unusually weak limbs?
> > This is why skills and stats make sense to be distinct.
>
> But, how are they different? If you discount stats like
> intelligence, there seems to be no difference between a Strength
> stat and a Cooking skill. Sure, cooking is more narrow, but there
> is not any reason why we can't have broad skills (Ie: Athletics)
You're right - there isn't any reason. But there also isn't any reason
to call it a skill when you could call it Innate Athletic statistic.
We're discussing the ancient nature vs. nurture, and it seems that you
believe that everything is nurture.
> The "primary stat bar" is a hangover from AD&D which should be
> abolished. Sooner we fill those characters with actual relevant
> data, the better.
Change the stats and call them base skills and you've still got the
primary skill bar. If you categorise them strength skills, dexterity
skills, etc. you've got the same system under a different name. You
see that, don't you? It's poor stat design you should be arguing
against, not some awful nature of stats.
Jeff Lait
> Jeff Lait communicated:
> >
> > The groupings are driven by the skills, not the skills driven by
the
> > groupings.
>
> Why can't I use this intelligent approach to stat design? How do
base
> stats necessarily differ from base skills?
Different name. A diamond by another name is cubic zirconium.
I also want to reiterate what I said at the beginning:
> By "Stats" I refer herein to that brief numerical summary of
character
> abilities, such as the ADOM Toughness/Willpower, or AD&D
> Strength/Wisdom. I explicitly exclude stuff like Hit Points,
> Alignment, Age, Body Weight, Hair Colour, etc.
Of course "stats" could be designed well. However, so long as people
call them "stats" they are liable to start from the basis of AD&D and
this notion of their being "primary stats".
> And I disagree that strength is a foolish stat.
A "statistic" should be measurable. Weight is a good stat. Hair
colour a good stat. Why is it that we use "statistic" to refer to the
"general tendency" and an large meta grouping of actual statistics?
If you want "Max Bench Press" as a stat, that is fine. However, most
systems do not have Strength to mean *only* Max Bench Press. They
consider it to be some mystical underlying strength trait which affects
all strength related tasks. (And, oddly enough, can't be trained, as
it is considered to be innate as opposed to skills.)
> > > But base proficencies aren't. Give me two poor twins who have
been
> > > separated and locked in tight, dark cells for all of their
eighteen
> > > years of life and when you let them out I guarantee that the
better
> > > packer will have higher strength, more inventiveness or better
> > > luck.
> > >
> > > Then lock the better packer back in a cell for another five years
> > > while we teach the other how to pack and lift bags of loot, and
> > > the tables will possibly be turned. /Possibly/.
> >
> > I think 5 years of training would easily offset 5 years of being
> > locked in a box.
>
> Even if the one let out of the box to train is two feet shorter and
has
> unusually weak limbs?
I think the one stuck in the box will end up shorter by virtue of no
exercise. I also thought you were starting with "twins" which suggests
that there wouldn't be a large "nature" based variance.
> > > This is why skills and stats make sense to be distinct.
> >
> > But, how are they different? If you discount stats like
> > intelligence, there seems to be no difference between a Strength
> > stat and a Cooking skill. Sure, cooking is more narrow, but there
> > is not any reason why we can't have broad skills (Ie: Athletics)
>
> You're right - there isn't any reason. But there also isn't any
reason
> to call it a skill when you could call it Innate Athletic statistic.
> We're discussing the ancient nature vs. nurture, and it seems that
you
> believe that everything is nurture.
I am *not* discussing nature vs nurture. I would reject the notion of
"nature vs nature" as some axis along which you can split performance.
Statistics is an attempt to claim that there is some "nature" component
and some seperate "nurture" component. The two things are not
seperable in such a trivial fashion. Some people can sleep all day and
still be strong. Others can work in the gym and be strong. Both, in
the end, are strong.
> > The "primary stat bar" is a hangover from AD&D which should be
> > abolished. Sooner we fill those characters with actual relevant
> > data, the better.
>
> Change the stats and call them base skills and you've still got the
> primary skill bar.
If they are called skills, hopefully the developer will be smart enough
to realize they shouldn't chew up valuable screen real estate to give
the player turn by turn accounting of them.
> If you categorise them strength skills, dexterity
> skills, etc. you've got the same system under a different name. You
> see that, don't you? It's poor stat design you should be arguing
> against, not some awful nature of stats.
I am arguing against poor stat design. Poor stat design is inherited
from the connotations implicit in the term "stat". It is inherited
from this assumption that there exists a discrete "Nature" component,
and that we need to model this inside a game.
The term "stat" is particularly bad when it sneaks onto the same
listing as HP and MP. The latter are *real* primary statistics.
We do not need a centralized "Dexterity" stat to modify 100 unrelated
things by +/- 1. The idea that we need that is because we are used to
systems that have it. Systems have that because they *started* with a
Dexterity stat and then went hunting for "uses" for it.
Why do you think we end up in perennial Charisma vs Appearance style
debates? Arguments if Strength or Dexterity should affect to-hit
bonuses? Actual debates if Intelligence, Wisdom, or Charisma should
affect cooking more! It is because the underlying mechanics is
backwards! It is not traditional primary stats like Str that build
secondary stats like Carrying Capacity. It is Carrying Capacity that
determines Strength!
To put it succinctly: We say someone has the "Strong" trait because
they can lift a heavy object.
I'd much rather have things like "Strong" and "Nimble" treated as Feats
or Skills. I'm all for predispositions for characters, so a "Muscle
bound" character will have better innate abilities. This is not,
however, the traditional AD&D Str attribute.
Mike Blackney
>
> Mike Blackney wrote:
> >
> > do base stats necessarily differ from base skills?
>
> Different name. A diamond by another name is cubic zirconium.
So your plan is going to change things little. Like how the homosexual
community has historically renamed itself every decade to reduce the
stigma attached to the word, your new word will eventually be clumped
with the same bad rap simply because it refers to the same object.
> I also want to reiterate what I said at the beginning:
>
> > By "Stats" I refer herein to that brief numerical summary of
> > character abilities, such as the ADOM Toughness/Willpower,
> > or AD&D Strength/Wisdom. I explicitly exclude stuff like Hit
> > Points, Alignment, Age, Body Weight, Hair Colour, etc.
>
> Of course "stats" could be designed well. However, so long as people
> call them "stats" they are liable to start from the basis of AD&D and
> this notion of their being "primary stats".
>
> > And I disagree that strength is a foolish stat.
>
> A "statistic" should be measurable. Weight is a good stat. Hair
> colour a good stat. Why is it that we use "statistic" to refer to
> the "general tendency" and an large meta grouping of actual
> statistics?
It's a name. Call them 'traits' and 'skills' then?
> If you want "Max Bench Press" as a stat, that is fine. However,
> most systems do not have Strength to mean *only* Max Bench Press.
> They consider it to be some mystical underlying strength trait
> which affects all strength related tasks.
Well your proposed 'Athletics' skill is just as unrealistic. I could
be a master sprinter and long-jumper but a lousy rings and pole-vaulter
simply because I have low upper arm strength. All abstractions will
have their limits.
> (And, oddly enough, can't be trained, as it is considered to be
> innate as opposed to skills.)
Played Nethack lately? And don't get out of it by saying that you
meant 'Most Systems' because you're trying to argue that skill systems
are fundamentally flawed. That most systems are flawed only proves
that most system designers didn't think/care about the realism.
> > > I think 5 years of training would easily offset 5 years of
> > > being locked in a box.
> >
> > Even if the one let out of the box to train is two feet shorter
> > and has unusually weak limbs?
>
> I think the one stuck in the box will end up shorter by virtue of no
> exercise. I also thought you were starting with "twins" which
> suggests that there wouldn't be a large "nature" based variance.
You can't dispute the hypothetical facts. In the first eighteen years,
these non-identical twins grew differently. One is two feet taller
than the other and has naturally stronger limbs. With five years of
packing practice the shorter, innately weaker twin may have trouble
keeping up.
> I am *not* discussing nature vs nurture. I would reject the notion
> of "nature vs nature" as some axis along which you can split
> performance.
>
> Statistics is an attempt to claim that there is some "nature"
> component and some seperate "nurture" component. The two things are
> not seperable in such a trivial fashion. Some people can sleep all
> day and still be strong. Others can work in the gym and be strong.
> Both, in the end, are strong.
But one has training.
> > > The "primary stat bar" is a hangover from AD&D which should be
> > > abolished. Sooner we fill those characters with actual relevant
> > > data, the better.
> >
> > Change the stats and call them base skills and you've still got the
> > primary skill bar.
>
> If they are called skills, hopefully the developer will be smart
> enough to realize they shouldn't chew up valuable screen real estate
> to give the player turn by turn accounting of them.
?? If the player needs to check their abilities every hundred turns,
why not put it on the status bar? Because their called skills? What
sense is that?
> > If you categorise them strength skills, dexterity
> > skills, etc. you've got the same system under a different name.
> > You see that, don't you? It's poor stat design you should be
> > arguing against, not some awful nature of stats.
>
> I am arguing against poor stat design. Poor stat design is inherited
> from the connotations implicit in the term "stat". It is inherited
> from this assumption that there exists a discrete "Nature" component,
> and that we need to model this inside a game.
Again, see Nethack.
> To put it succinctly: We say someone has the "Strong" trait because
> they can lift a heavy object.
I think that since said someone is strong before they lift the object,
the order is just fine how it is. I'm not saying that your skill-only
system is wrong, but defending stats because I honestly disagree that
they are as flawed as you make out.
James Bulgin
> Different name. A diamond by another name is cubic zirconium.
Actually, a cubic zirconia is a popular variety of fake diamond, but is
certainly not a diamond itself.
I don't really have anything to add to the discussion, but I just felt like
mentioning that. I appologize for being picky :P
Jeff Lait
> Jeff Lait communicated:
> >
> > Mike Blackney wrote:
> > >
> > > Why can't I use this intelligent approach to stat design? How
> > > do base stats necessarily differ from base skills?
> >
> > Different name. A diamond by another name is cubic zirconium.
>
> So your plan is going to change things little. Like how the
homosexual
> community has historically renamed itself every decade to reduce the
> stigma attached to the word, your new word will eventually be clumped
> with the same bad rap simply because it refers to the same object.
I think you have been missing the point I was trying to make. I've
clearly done a poor job of communicating, so apologize. My fault lies
with excessive rhetoric on my part. Inside a thread named after the
infamous "Goto considered harmful", I thought it on topic to present a
rabid view of anti-stats.
A name *does* change things. An athletics skill will mean a very
different thing to the *developer* than a strength stat. This means
the *developer* will create a different thing. The rseult is thus not
a strength stat renamed to athletics, but something actually different.
> > I also want to reiterate what I said at the beginning:
> >
> > > By "Stats" I refer herein to that brief numerical summary of
> > > character abilities, such as the ADOM Toughness/Willpower,
> > > or AD&D Strength/Wisdom. I explicitly exclude stuff like Hit
> > > Points, Alignment, Age, Body Weight, Hair Colour, etc.
> >
> > Of course "stats" could be designed well. However, so long as
people
> > call them "stats" they are liable to start from the basis of AD&D
and
> > this notion of their being "primary stats".
> >
> > > And I disagree that strength is a foolish stat.
> >
> > A "statistic" should be measurable. Weight is a good stat. Hair
> > colour a good stat. Why is it that we use "statistic" to refer to
> > the "general tendency" and an large meta grouping of actual
> > statistics?
>
> It's a name. Call them 'traits' and 'skills' then?
Certainly! Of course, this implies you have rolled them in with a
larger "trait" system or "skill" system, rather than just renaming your
stat line. Then these will be seen properly a secondary
characteristics rather than primary characteristics.
s/stat/trait/ of course will not suffice. But, if one has a system of
traits, and adds to it "stat-like" predispositions, one would not end
up with the STR: 18 status line.
> > If you want "Max Bench Press" as a stat, that is fine. However,
> > most systems do not have Strength to mean *only* Max Bench Press.
> > They consider it to be some mystical underlying strength trait
> > which affects all strength related tasks.
>
> Well your proposed 'Athletics' skill is just as unrealistic. I could
> be a master sprinter and long-jumper but a lousy rings and
pole-vaulter
> simply because I have low upper arm strength. All abstractions will
> have their limits.
So why start a design by making arbitrary abstractions?
Don't say you can make a design that starts with
skills/hitpoints/attack values and then derives primary statistics.
They would not then be primary statistics, now would they? A system
with primary statistics is a system that went from the primary to the
secondary, not one that went the other way around.
> > (And, oddly enough, can't be trained, as it is considered to be
> > innate as opposed to skills.)
>
> Played Nethack lately? And don't get out of it by saying that you
> meant 'Most Systems' because you're trying to argue that skill
systems
> are fundamentally flawed. That most systems are flawed only proves
> that most system designers didn't think/care about the realism.
1) I don't think or care about realism.
2) I did say most systems.
3) Nethack's training, which I was aware of when I was writing all
this, only puts into sharp relief the pointlessness of having primary
stat systems. Nethack turns the stats into the equivalent of skills,
resulting in stat-capped characters being the norm rather than the
exception. They thus no longer represent any innate trait. Of course,
my objection *isn't* an nature vs nurture one. It is that when you
realize you have built a nethack style system, you obviously aren't
modelling innate traits any more. So why are we calling them primary
statistics and defending them with language based on innate traits?
Why not throw them out and go with an explicit trait/skill system which
has the advantage of not promoting orthogonality or artificial,
fungible, scales?
> > > > I think 5 years of training would easily offset 5 years of
> > > > being locked in a box.
> > >
> > > Even if the one let out of the box to train is two feet shorter
> > > and has unusually weak limbs?
> >
> > I think the one stuck in the box will end up shorter by virtue of
no
> > exercise. I also thought you were starting with "twins" which
> > suggests that there wouldn't be a large "nature" based variance.
>
> You can't dispute the hypothetical facts. In the first eighteen
years,
> these non-identical twins grew differently. One is two feet taller
> than the other and has naturally stronger limbs. With five years of
> packing practice the shorter, innately weaker twin may have trouble
> keeping up.
Well, I can't very well say anything then, can I? Presumeably, the
"fact" that the innately weaker twin can't keep up is one of the things
I can't dispute, no? Why not just add the hypothetical fact that the
training twin dies at the end of the training? Then I'll have no way
to argue he'd be the better pack bearer.
> > > > The "primary stat bar" is a hangover from AD&D which should be
> > > > abolished. Sooner we fill those characters with actual
relevant
> > > > data, the better.
> > >
> > > Change the stats and call them base skills and you've still got
the
> > > primary skill bar.
> >
> > If they are called skills, hopefully the developer will be smart
> > enough to realize they shouldn't chew up valuable screen real
estate
> > to give the player turn by turn accounting of them.
>
> ?? If the player needs to check their abilities every hundred turns,
> why not put it on the status bar? Because their called skills? What
> sense is that?
Why do they need to check their abilities every hundred turns? Or, why
do they need a constant running tally of Strength, but their Sword
Skill is considered irrelevant?
The reason is because they have been enshrined as "primary stats" with
the implication that everything derives from them. Placing them on the
stat bar is the first step to destruction. One then usually thinks:
"Ah, these are on the stat bar all the time. I guess I better have
them change more often to justify their presense!" We then get stat
gain potions, stat exercising, and potions of exchange.
> > To put it succinctly: We say someone has the "Strong" trait because
> > they can lift a heavy object.
>
> I think that since said someone is strong before they lift the
object,
> the order is just fine how it is. I'm not saying that your
skill-only
> system is wrong, but defending stats because I honestly disagree that
> they are as flawed as you make out.
Well, then we can stop this discussion right now. When I say "stats
considered harmful", they clearly aren't some virulent poison that
would destroy any game they touch. One need merely note all the
successful games with stats.
I merely mean what I posted to begin with: new roguelike authors would
do well to rethink the reflexive inclusion of primary stats.
The Sheep
> Jeff Lait wrote:
> It's hard to see how to devise a system that takes all of this into
> account without overcomplicating things.
How about having a `standard tree of skills' reversed, with it's roots
pointing into the sky?
Define lots of `primitive stats', like Muscle Bound, Coordination, etc.
Maybe even take into account arms and legs separately. Then compute
`base stats', that serve as bases for actual attributes.
You get the actual attribute values by adding (multiplying, etc.) the
skills to the base stats.
Yes, it's complicated, but we don;t have to show it to the player -- he's
only interested in base stats and skills. Primitive stats are for us to
bound attributes together when they are related. Abilities are used to
get actual results of actions, and there are as many of them as possible
types of actions. Both list are pretty long, but the list of `base stats'
and `skills' in between may be much shorter.
It's like a multi-layer neural network.
--
Radomir @**@_ Bee! Create your own computer role playing
`The Sheep' ('') 3 game using the great H-World engine!
Dopieralski .vvVvVVVVVVvVVvv.v. http://h-world.simugraph.com/
R. Dan Henry
wrote:
>Stat Gain Potions - number one contributor to "thousanduplet
>syndrome", where all the winning characters look alike. With
>stat maximums in place, plentiful stat gain potions make all
>the characters attain the stat maximums, thereby becoming
>identical thousanduplets.
>
>What would the games be like if stat gain potions became so
>rare that, in the whole game, you might not find more than
>two or three of them?
Alternatively, how about: You start with a given set of stats and
every 10 points of stat gained by potions, training, etc. doubles the
cost per point. That is to say, your first ten Potions of Gain
Flexibility increase your flexibility by ten, but the next ten only
increase it by five. Character A starts with a stat of 8, Character B
with a stat of 15. After 100 potions, each has gained 37 points of
stat, so Character A has a 45 stat and Character B has a 52 stat. This
accommodates the possibility of huge numbers of stat gain potions,
while the relativity to base stats is a general method to maintain
character differentiation.
R. Dan Henry
danh...@inreach.com
Brent Ross
R. Dan Henry <danh...@inreach.com> wrote:
// On Sun, 12 Dec 2004 09:00:46 GMT, Ray Dillinger <be...@sonic.net>
// wrote:
// >Stat Gain Potions - number one contributor to "thousanduplet
// >syndrome", where all the winning characters look alike. With
// >stat maximums in place, plentiful stat gain potions make all
// >the characters attain the stat maximums, thereby becoming
// >identical thousanduplets.
// >
// >What would the games be like if stat gain potions became so
// >rare that, in the whole game, you might not find more than
// >two or three of them?
//
// Alternatively, how about: You start with a given set of stats and
// every 10 points of stat gained by potions, training, etc. doubles the
// cost per point.
Moria/Angband already does that to an extent (potions give progressively
less benefit the closer you get to 18/100). It's not a solution in
itself, and actually tends to increase aberrant behaviours. It will
increase stat gain potion scumming... munchkins will simply grit their
teeth and work twice as long for the benefit. It might also further
increase autoroll scumming for stats at the begining of the game (and if
you have a point system, it will increase the value of twinking it)... the
end result here is that all starting characters will look more alike, too.
// Character A starts with a stat of 8, Character B
// with a stat of 15. After 100 potions, each has gained 37 points of
// stat, so Character A has a 45 stat and Character B has a 52 stat.
In a lot of systems those characters have gone from having different
stats to pretty much the same, simply because the 7 point difference
is less that 1/6th of 45, but is almost double the original 8. And if
52 is worth almost twice the value of 45, then one has to question who
intuitive the stats system is (and should it be be retooled to present
a number more suitable for comparisons, like 84). This ratio effect is
due to the nature of the function 1/X, and so I call it the 1/X problem.
It's actually really common and it needs to be watched out for. In short,
the lession is that you need to be careful using ranges that vary too much
relatively or get too small (those last few steps can get pretty big[1]).
At the very least you need to be wary of the way the scales and ratios
are distorting.
// This
// accommodates the possibility of huge numbers of stat gain potions,
// while the relativity to base stats is a general method to maintain
// character differentiation.
If you want huge numbers of stat gain potions, my suggestion would be to
limit them to being temporary or non-stable (ie mutations that might get
removed by later effects). Giving out permanent boosts to stats from
an available plentiful source is just asking the players to scum them
for stat levels they don't even really need, and then complain about how
long the stat gain portion of the game is. If you design your game to
not encourage or include that, it will probably end up being more fun.
It's also easier to design a game when such variables are more static.
The problem with allowing stats to climb largely is that you end up
having to balance late game things for far more powerful characters, thus
causing a feedback loop to further encourage all characters to become
more powerful. A game with limited development doesn't have to worry
about stat munchkins, and so can provide a more consistant and interesting
game that plays off each character's strengths and weaknesses.
Brent Ross
[1] For example consider a "cost" (mana, food, whatever) with 1000 units
to spend (diff is the # extra units purchased compared with cost + 1).
cost 1 2 3 4 5 6 7 8
amount 1000 500 333 250 200 166 142 125
diff +500 +167 +83 +50 +34 +24 +17 +14
As this chart shows, -1 to the cost increases dramatically. For example,
if we look at these as food costs, we can see that the benfit from 3 to
2 is equivalent to the penalty from 3 to 6 (both are a difference of 167
turns on 1000 food units). The factor involved here is, 1/(x * (x + 1)),
or on the order of 1/x^2... which as shown in the chart above, starts to
flaten out pretty quickly so if you can avoid the bottom few values you
can drastically reduce your distortion. Push the granularity up further
and you can change the system to one where the differences are constant,
the amounts linear, and the costs calculated/amortized to maintain that
(and remove the distortion almost completely). The other option is
to know and work with the curve as is.
Ray Dillinger
> Alternatively, how about: You start with a given set of stats and
> every 10 points of stat gained by potions, training, etc. doubles the
> cost per point. That is to say, your first ten Potions of Gain
> Flexibility increase your flexibility by ten, but the next ten only
> increase it by five. Character A starts with a stat of 8, Character B
> with a stat of 15. After 100 potions, each has gained 37 points of
> stat, so Character A has a 45 stat and Character B has a 52 stat. This
> accommodates the possibility of huge numbers of stat gain potions,
> while the relativity to base stats is a general method to maintain
> character differentiation.
>
I'm doing something different, but I don't know yet how well the
game will balance with it.
The more stat potions you consume, the closer you get
(asymptotically) to a stat maximum level. You never actually
reach the maximum, but you can get within a point or two of it
if you're persistent.
This isn't "thousanduplet syndrome" because the maximum isn't a
fixed constant; the maximum is twice your initial stat roll.
If you started off as a weakling, you'll never be *very*
strong, no matter how many !oStrength you drink. Conversely,
if you started off as with huge strength, the first few strength
potions will put you well beyond what are normally human maximums
-- but as a human, you'll still never be as strong as, say, a
frost giant.
The magic formula is:
L = T + ((I * (X + 2P)) / (X + P))
Where L is the current effective stat, I is the initial stat,
T is training (depends on your class/level and behavior), X
is a constant that depends on character race, and P is the
number of statgain potions so far drunk.
Bear
Atriel
I, as a player, *want* stats. Lots of them.
Stat GAIN and final stat as is in angband is what愀 bad.
cheers
--
I will hold the candle till it burns up my arm.
I'll keep taking punches until their will grows tired.
I will stare the sun down until my eyes go blind.
I won't change direction and I won't change my mind...
How much difference does it make?
Atriel
stats. STATS are *NEEDED*. if there´s no stats, the game will degrade to a
boring gauntlet-alike turn based.
It´s impossible to try to categorize how things are in real life, how do i hit a
monster and the effects of say, mind and body etc etc in the to-hit. In real
life we can die with 1 blow.
I agree that stat gain sux, and no final variation sux.
So, here´s what i am intending to do to end stat gain.
I have plans to make a small variation, with changes only in the stats.
My coding is pretty rusty, but i have sources and buggy compiles
all over my HD.
Here´s a brief review of some things am planning:
Stats will always be POINT BASED on start. 60 points distributed at will, but
minimum 3 and maximum 17 to any starting stat except CHA.
CHA costs half point (one stat point = 2 CHA points)
Potions of STR, DEX, CON, INT, WIS, CHA and Augmentation do NOT raise the MAX
VALUE of a stat. They pump the stat TEMPORARILY to a buffer over the maximum
stat value. Effects that lower the stat affect this "buffer" first. The buffer
slowly fades away, until stat = max stat.
CHA is a highly volatile stat; Magic / elemental attacks will affect it. Yea,
you´ll remember for a long time how you got hurt.
All stats *slowly* come back to normal with time, if lower than MAX stat value.
All stats come back to normal with time, if higher than MAX stat value.
2 Stat points per level up. These points may be used to permanently
raise MAX STR, DEX, CON, INT, WIS, CHA.
CHA costs half point to raise (one stat point = 2 CHA points)
There is a cap that is (CLvl*2/5)+20 for all stats but CHA. (40, but tables
account up to 50)
Racial and Class modifiers are applied AFTER cap (Maximize Mode always on)
Points can be stored for later use.
Impact of CHA in stores will be considerably raised.
Infinite RANDARTS make-able in town for huge money.
That´s it. Of course, the XP table are completely redone.
Yeah. Each char will be very different, and completely based
on the player taste for stats.
Heavy weapon users will need less DEX due to lower possible blows.
Light weapon users will need more DEX and so on...
Mages will put in INT, and may be able to fight a lil
if they pump STR/DEX, or have full HP pumping CON,...
The player will have to plan the spending of stats based on
items he can / will use, etc etc.
If someone want to use it, go ahead :) if someone makes something
cool, ill probably use too :)
Brent Ross
// ack
// I, as a player, *want* stats. Lots of them.
I'd suggest going back and actually reading the post you're responding
to. The rejection of "stats" here is against the concept of STR, INT,
CON, etc as primary stats. Primary being defined as most important
factor in various different calculations. His proposal was that the
derived/secondary stats should be the primary ones. In other words,
instead of having strength and applying it to carrying capacity and
extra melee damage... just define carrying capacity and extra melee
damage as stats, and don't bother with strength. So he's pretty much
talking about exploding the stats into all their parts (ie stength becomes
strength-carry and strength-extra-damage). If you want lots of stats,
he was giving them to you (and then some).
My points were mostly based on the fact that I consider primary here to
be more about the ability to derive, than the requirement that they be
necessarily large factors. For example, strength is somewhat important
to carrying capacity, but size and shape of the character are also
very important aspects... at least as large as, if not more so, than
strength. The benefit of the traditional stats here isn't in that they
represent the sum of the characters capabilities... but are more about
transferable ability and capabilities (ie they're more like attributes
like size and body shape and parts). For example, a smarter PC can use
their intelligence to help develop their spell casting and cast spells
a little easier, but things like skills and foci should be the real
factors in how well magic works. Stats give flexibility, not power here.
Of course, both systems are useful and worth thinking about.
Brent Ross
Jeff Lait
> Funny, this thread started with topic about evils of STAT GAIN, not
evils of
> stats.
I apologize for committing the horrible sin of moving a thread off
topic.
> STATS are *NEEDED*. if there´s no stats, the game will degrade to a
> boring gauntlet-alike turn based.
Stats, meaning some quantitative analog of hit points, mana points,
etc, *are* needed. Of course, gauntlet had these stats, IIRC.
Stats, meaning AD&D Str/Dex/etc are most definitely not needed! My
argument is not so much that AD&D stats are bad things, as the belief
that they are *necessary* is bad.
> It´s impossible to try to categorize how things are in real life,
how do i hit a
> monster and the effects of say, mind and body etc etc in the to-hit.
In real
> life we can die with 1 blow.
>
> I agree that stat gain sux, and no final variation sux.
> So, here´s what i am intending to do to end stat gain.
> I have plans to make a small variation, with changes only in the
stats.
> My coding is pretty rusty, but i have sources and buggy compiles
> all over my HD.
> Here´s a brief review of some things am planning:
>
> Stats will always be POINT BASED on start. 60 points distributed at
will, but
> minimum 3 and maximum 17 to any starting stat except CHA.
> CHA costs half point (one stat point = 2 CHA points)
And you go to support my position that Stats are considered Harmful.
Why not throw out the CHA stat entirely rather than make it some
bastard half point stat? Why not base point costs based on skew from
the average rather than flat rates (There by resulting in the
guaranteed 17s in the "key" stats)
> Heavy weapon users will need less DEX due to lower possible blows.
> Light weapon users will need more DEX and so on...
Why require the user to do this min-maxing? It seems we could throw
out Str & Dex and have a single Physical Prowess stat. This Heavy vs
Light distinction could then be accounted for by having Heavy vs Light
combat skills.
Jeff Lait
Brent Ross wrote:
> My points were mostly based on the fact that I consider primary here
to
> be more about the ability to derive, than the requirement that they
be
> necessarily large factors. For example, strength is somewhat
important
> to carrying capacity, but size and shape of the character are also
> very important aspects... at least as large as, if not more so, than
> strength. The benefit of the traditional stats here isn't in that
they
> represent the sum of the characters capabilities... but are more
about
> transferable ability and capabilities (ie they're more like
attributes
> like size and body shape and parts). For example, a smarter PC can
use
> their intelligence to help develop their spell casting and cast
spells
> a little easier, but things like skills and foci should be the real
> factors in how well magic works. Stats give flexibility, not power
here.
Thank you for the nice summary. The quoted text of your view shows we
are not very far off in our views.
Atriel
> Stats, meaning some quantitative analog of hit points, mana points,
> etc, *are* needed. Of course, gauntlet had these stats, IIRC.
Why are needed hit points? This is the most untruestat that there is.
In real life, It愀 possible to die with only 1 well done hit, a hit that
like, would behead the char.
> Stats, meaning AD&D Str/Dex/etc are most definitely not needed! My
> argument is not so much that AD&D stats are bad things, as the belief
> that they are *necessary* is bad.
I agree with you... they are not necessary. but there are players that
want to play a game and just kill stuff and stay alive and drink a lot
of potions to get strong and not tweak at ALL their characters,
just keep going ahead.
And there are players that like tweaking. IMHO the best
tweaking Angband var today is TOME. Sangband is cool, but broken.
With no limits, every character looks the same in the end in Sang.
But these vars have only skills, stats still max with potions...
> And you go to support my position that Stats are considered Harmful.
> Why not throw out the CHA stat entirely rather than make it some
> bastard half point stat? Why not base point costs based on skew from
> the average rather than flat rates (There by resulting in the
> guaranteed 17s in the "key" stats)
Because Cha is in a long term convention. I like it. Lots of people like it.
It愀 cheap, and lots of people want to make a sexy char. :)
As money will be of use in town (randart buying), CHA will be useful.
There will be people that will leave it at 3 and be happy with a
Half Troll Warrior.
There will be people that will make a HE something that will buy
everything cheap. and have lots of pets.
Yeah. CHA will have major effect in shops that the current effects.
But it wont help much against that orc pit.
> > Heavy weapon users will need less DEX due to lower possible blows.
> > Light weapon users will need more DEX and so on...
> Why require the user to do this min-maxing? It seems we could throw
> out Str & Dex and have a single Physical Prowess stat. This Heavy vs
> Light distinction could then be accounted for by having Heavy vs Light
> combat skills.
Veteran triathlon/heptathlon athletes know that there is a balance
in their muscles.
When you start exercising (this is TRUE real life):
*There are exercises
that make the muscle fast, explosive, but not very powerful.
Generally, short range sprints/repetitions etc. The muscular
cells are allocated in a way that the glucogen will be used very
fast but with few reserves.
Of course, a guy that trains his muscle to be explosive probably
already trained his reflexes to use his muscles the right way.
DEX is this.
*There are exercises that make the muscle resistant, but not very
fast nor powerful. Generally, long sprints. Glucogen in this
case will be accumulated between the cells.
This woud be like, Stamina... a derivate from CON. (vide DiabloII)
*There are exercises that make the muscle powerful, but somewhat
clumsy. Like heavy weight lifting. Bodybuilders are this way.
Glucogen is mostly intracelular and there are more cells too.
In this case, resistance also suffers.
This is STR, of course.
There is a point that the athlete reachs where improving the 3 ways
is ALMOST IMPOSSIBLE. So if he wants to lean towards one
way, he惻l have to give up the other 2. Or stay balanced
forever at like 60% of the full power he would have
if hed chosen a way. But if he choose one way, he will
almost certainly degrade the others, or at least not develop
them fully.
IMHO, two skills "heavy combat skills" and "light combat skills"
are horrible, just because it depends on the BODY of the char.
Or else, a Marathon sprinter maybe could be good with a
Maul if he had "heavy combat skills"...
The skills will be chosen in the class. If the guy wants to fight,
he惻l chose a warrior. With heavy weapons, pump STR
and enough DEX to give the maximum n of blows with
his weapon. Heavier weapon max attacks will be like 4 blows
ala [O].
So, no. a single physical prowess stat is very, very bad.
About Min-maxing: This is not very hard to explain.
There would be huge variability of character builds.
Cheers
David Damerell
>JEff Lait:
>>Stats, meaning some quantitative analog of hit points, mana points,
>>etc, *are* needed. Of course, gauntlet had these stats, IIRC.
>Why are needed hit points? This is the most untruestat that there is.
>In real life,
Jeff is of course arguing from gameplay, not realism. In any combat-heavy
game [1], taking damage needs to be treated highly unrealistically.
>There will be people that will leave it at 3 and be happy with a
>Half Troll Warrior.
>But it wont help much against that orc pit.
JOOI, do you really think the world needs another Angband variant?
>>Why require the user to do this min-maxing? It seems we could throw
>>out Str & Dex and have a single Physical Prowess stat. This Heavy vs
>>Light distinction could then be accounted for by having Heavy vs Light
>>combat skills.
>Veteran triathlon/heptathlon athletes know that there is a balance
>in their muscles.
>When you start exercising (this is TRUE real life):
OK; what's your gameplay argument?
[1] Yes, I know roguelikes don't have to be, and I'm sure one can think of
an exception, but you know what I mean.
--
David Damerell <dame...@chiark.greenend.org.uk> Kill the tomato!
R. Dan Henry
>I want to know your thoughts on pen and paper RPing systems.
>Particulally two: D&D and GURPS.
Both broken in several places [1], although GURPS at least is only
seriously broken in two. Unfortunately, these are at the heart of the
system. One is "stats", where their worth is all out of balance and
the DEX and INT stats dominate skill for the most part. The second is
the point-buy system. Unlike Champions/HERO where points were sensibly
based purely on game-balance (although, of course, one can always
argue the actual point values, it isn't *conceptually* broken), the
interchangeable character points represent three things: game balance
value (special abilities/flaws), easy/difficulty of learning (skills)
and demographic distribution (stats). Mixing in the "realism" factors
creates points that can be freely exchanged between areas in which
they change meaning and balance is broken. Notoriously, one can have
better skills by sinking most of the points into stats and barely
buying a few key skills.
[1] D&D less than it used to be, given the change list I've seen for
D&D3E, but it is still a clumsy beast. I will also note that neither
is so broken that I would consider them unplayable, but why not just
play a better game?
R. Dan Henry
danh...@inreach.com