Roguelike developer blog
andrewdoull
This group seems a little quiet at the moment.
So I've put up an article on my blog, suggesting that roguelikes developers
shouldn't bother implementing stats in roguelikes. e.g. get rid of trying to add
Str, Int, Dex, Con, Cha and so on, and spend more time working on the actual
game.
Article is here: http://roguelikedeveloper.blogspot.com
Responses welcome.
Andrew
--
The Roflwtfzomgbbq Quylthulg summons L33t Paladins -more-
"Mr. Candidate, sir, given the overwhelming body of evidence from hundreds of
different scientific fields ranging from archeology to physics to zoology, can
you explain to us how you can seriously believe that the world was created 2,000
years after the Babylonians invented beer?" - Bombula on Slashdot - 17/08/2007
ASCII Dreams: http://roguelikedeveloper.blogspot.com
Unangband: http://unangband.blogspot.com
Krice
> e.g. get rid of trying to add
> Str, Int, Dex, Con, Cha and so on, and spend more time
> working on the actual game.
How then can you tell which monster is stronger than
another if you don't have STR?
I guess STR has to be replaced with something. It could
be something like the mass and/or size of the monster.
andrewdoull
> On 21 elo, 12:18, andrewdoull wrote:
> > e.g. get rid of trying to add
> > Str, Int, Dex, Con, Cha and so on, and spend more time
> > working on the actual game.
>
> How then can you tell which monster is stronger than
> another if you don't have STR?
> I guess STR has to be replaced with something. It could
> be something like the mass and/or size of the monster.
I guess I should have made it clear I'm talking about player stats. I'm also not
talking about all abstractions: anything relating to a game-mechanic is fine. If
you have arm wrestling in your roguelike, an strength indicator is probably
required.
However, if all that strength does is increase damage and carrying capacity,
then why have a strength stat. Increase damage and carrying capacity directly
instead.
Obnoxious User
> On 2007-08-21 12:32:21, Krice <pau...@mbnet.fi> wrote:
>
>> On 21 elo, 12:18, andrewdoull wrote:
>> > e.g. get rid of trying to add
>> > Str, Int, Dex, Con, Cha and so on, and spend more time
>> > working on the actual game.
>>
>> How then can you tell which monster is stronger than
>> another if you don't have STR?
>> I guess STR has to be replaced with something. It could
>> be something like the mass and/or size of the monster.
>
> I guess I should have made it clear I'm talking about player stats. I'm also not
> talking about all abstractions: anything relating to a game-mechanic is fine. If
> you have arm wrestling in your roguelike, an strength indicator is probably
> required.
>
> However, if all that strength does is increase damage and carrying capacity,
> then why have a strength stat. Increase damage and carrying capacity directly
> instead.
>
This kind of purging of unnecessary stats and abstractions have been
going on for some while now in the paper and pen rpg world, and it
has produced some very nice playable systems.
--
Obnoxious User
Krice
> However, if all that strength does is increase damage and carrying
> capacity, then why have a strength stat.
So that you know how much you can carry and do damage.
> Increase damage and carrying capacity directly instead.
Isn't that like inventing two new stats?
Andrew Sidwell
One assumes that if one's game has carrying limits and an amount of
damage dealt per turn, then there are already two stats. Adding a third
is merely adding needless complexity.
--
Andrew Sidwell
http://rephial.org/ -- the home of Angband
Timofei Shatrov
confuse everyone with this message:
>Krice wrote:
>> On 21 elo, 13:47, andrewdoull <andrewdo...@gmail.com> wrote:
>>> However, if all that strength does is increase damage and carrying
>>> capacity, then why have a strength stat.
>>
>> So that you know how much you can carry and do damage.
>>
>>> Increase damage and carrying capacity directly instead.
>>
>> Isn't that like inventing two new stats?
>
>One assumes that if one's game has carrying limits and an amount of
>damage dealt per turn, then there are already two stats. Adding a third
>is merely adding needless complexity.
>
No, it actually decreases complexity. If you have DAM and CAP stats, for
example, but no STR stat then the dimension of the range of possible values is 2
(there are 0 equations and 2 variables).
But when you add STR, and DAM and CAP are calculated based on STR then you
decrease the dimension to 1 (2 equations, 3 variables). That's basically why
stats were invented, they consolidate all possible metrics that a character may
have to a few intuitive parameters such as strength, dexterity etc.
--
|Don't believe this - you're not worthless ,gr---------.ru
|It's us against millions and we can't take them all... | ue il |
|But we can take them on! | @ma |
| (A Wilhelm Scream - The Rip) |______________|
konijn_
> No, it actually decreases complexity. If you have DAM and CAP stats, for
> example, but no STR stat then the dimension of the range of possible values is 2
> (there are 0 equations and 2 variables).
>
> But when you add STR, and DAM and CAP are calculated based on STR then you
> decrease the dimension to 1 (2 equations, 3 variables). That's basically why
> stats were invented, they consolidate all possible metrics that a character may
> have to a few intuitive parameters such as strength, dexterity etc.
Thank you ;)
Jeff Lait
> Hi,
>
> This group seems a little quiet at the moment.
>
> So I've put up an article on my blog, suggesting that roguelikes developers
> shouldn't bother implementing stats in roguelikes. e.g. get rid of trying to add
> Str, Int, Dex, Con, Cha and so on, and spend more time working on the actual
> game.
This is an argument dear to my heart. Stats are Harmful:
http://groups.google.com/group/rec.games.roguelike.development/msg/33eaa49a18b00255
http://groups.google.com/group/rec.games.roguelike.development/msg/5bcaaa8330e58518
In both cases these posts are part of larger threads I encourage
people to read.
--
Jeff Lait
(POWDER: http://www.zincland.com/powder)
andrewdoull
>
>
> > No, it actually decreases complexity. If you have DAM and CAP stats, for
> > example, but no STR stat then the dimension of the range of possible values is 2
> > (there are 0 equations and 2 variables).
> >
> > But when you add STR, and DAM and CAP are calculated based on STR then you
> > decrease the dimension to 1 (2 equations, 3 variables). That's basically why
> > stats were invented, they consolidate all possible metrics that a character may
> > have to a few intuitive parameters such as strength, dexterity etc.
>
> Thank you ;)
So how many pounds / kilograms can a strength 18 character carry?
Oh, that's right. You can't tell without a table lookup of some time.
However, presumably the amount each item weighs is available to you, so you can
figure out how much you are currently carrying.
So suddenly you need to know two things again: your strength, and how much you
can carry.
Similarly, for damage, you need to know (or be able to approximately guess)
monster hit points, and have an idea of how much damage you can inflict.
So you have these less than useful abstractions that take you away from the
actual game mechanics, into some kind of fuzzy 'intuitive' world...
Christophe
> So how many pounds / kilograms can a strength 18 character carry?
>
> Oh, that's right. You can't tell without a table lookup of some time.
>
> However, presumably the amount each item weighs is available to you, so you can
> figure out how much you are currently carrying.
>
> So suddenly you need to know two things again: your strength, and how much you
> can carry.
>
> Similarly, for damage, you need to know (or be able to approximately guess)
> monster hit points, and have an idea of how much damage you can inflict.
>
> So you have these less than useful abstractions that take you away from the
> actual game mechanics, into some kind of fuzzy 'intuitive' world...
Then again, how many points of damage a character that can carry 40 kg
with a single hand would do if he were to punch a goblin in the face ?
Gamer_2k4
> shouldn't bother implementing stats in roguelikes. e.g. get rid of trying to add
> Str, Int, Dex, Con, Cha and so on, and spend more time working on the actual
> game.
Well, I agree that you shouldn't make six different stats from the
start and base your game on them, but it's not wrong to use them. For
example, I add the stats as they're used. As soon as I got LOS going,
I made a Perception stat that determines the radius. As soon as I had
my speed system, I added a Dexterity stat that determines how fast you
are. As soon as I added items and inventory, I added a Strength stat
that determined how much you can carry. Maybe I'll rename or condense
these attributes later, but for now, they serve a purpose, and don't
keep me from "working on the actual game."
Another reason for stats is skills. Many RPGs have stat modifiers
that are determined by the stat level. If you have a Strength of 10,
you might have a -1 modifier. If you have a Strength of 20, you might
have a +2 modifier. Then, each skill has a certain stat that affects
it. Literacy, for example, might be an INT skill. That means you add
the Intelligence modifier when making skill checks. Now not only are
the stats balanced and useful, but their uses are also tangible.
Anyone can say, "Oh, if I increase my Toughness, it'll help these five
skills."
So with the proper system, stats aren't harmful at all. They're
really only an issue in Angband type games, many developers are using
different approaches anyway in the interest of originality. I'd say
that stats are fine.
--
Gamer_2k4
Norbert Melzer
> Then again, how many points of damage a character that can carry 40 kg
> with a single hand would do if he were to punch a goblin in the face ?
Not very much... This strong character is simply to slow to hit the small
and fast goblin.
SCNR
Norbert
PS: This is based on my own experiences in ugly real life. My strong
comrades are often slow, and the fast and agile ones are not really strong.
andrewdoull
"You were right: too much weights and not enough speed work." -- Jake.
konijn_
> On Aug 21, 5:18 am, andrewdoull <andrewdo...@gmail.com> wrote:
>
> > Hi,
>
> > This group seems a little quiet at the moment.
>
> > So I've put up an article on my blog, suggesting that roguelikes developers
> > shouldn't bother implementing stats in roguelikes. e.g. get rid of trying to add
> > Str, Int, Dex, Con, Cha and so on, and spend more time working on the actual
> > game.
>
> This is an argument dear to my heart. Stats are Harmful:
>
<SNIP>
It still does not say why for a player there is
benefit to have more skills/abilities/perks and
no stats, it only irks designers.
UI & context matter; when I play on main screen,
I can handle 6 stats. When I check my inventory,
I can see my carry stat ( which is derived from
STR ),when I check my weapon or character screen I
can see my damage bonus if I would be interested
in that.
Anyway, I guess this is a semi-religious old fart
vs. new fart type a thing, so that's the last you
hear of me on this topic ;)
Cheers,
T.
Martin Read
>So how many pounds / kilograms can a strength 18 character carry?
If I were going to write a carrying capacity based on some property of
items more simulationist than item type and item stackability, I would
use a simple formula for relating physical prowess to carrying capacity.
For example, 5kg unencumbered capacity per point of the Body stat (which
stat is also used to determine melee damage bonus and also per-level hit
point gain).
If you have an accompanying document entitled "Levels considered
harmful", then sir, I cordially invite you to eat it, seasoned with the
sauce of your choice.
--
\_\/_/ you take a mortal man and put him in control
\ / and watch him become a god watch people's heads roll
\/ --- Megadeth, "Symphony of Destruction"
R. Dan Henry
Oh, yes. Or go all the way back to "Skills" in 2001. Here's where I
first declared that stats are potentially discardable with an adequate
skill system:
http://groups.google.com/group/rec.games.roguelike.development/msg/b34f5819b885c53c?hl=en&
This was, in fact, the thread that led to changing my hypothetical
original roguelike game to a statless design. Several other of my posts
in that thread contain arguments for the disposability of stats;
however, the thread is about skills, just as advertised. If you've never
read it, I suggest reading the entire thread. There was a lot of good
discussion about skills and skill trees.
--
R. Dan Henry = danh...@inreach.com
If you wish to put anything I post on your website,
please be polite enough to ask first.
R. Dan Henry
wrote:
>Well, I agree that you shouldn't make six different stats from the
>start and base your game on them, but it's not wrong to use them. For
>example, I add the stats as they're used. As soon as I got LOS going,
>I made a Perception stat that determines the radius. As soon as I had
>my speed system, I added a Dexterity stat that determines how fast you
>are. As soon as I added items and inventory, I added a Strength stat
>that determined how much you can carry. Maybe I'll rename or condense
>these attributes later, but for now, they serve a purpose, and don't
>keep me from "working on the actual game."
And as long as those are all you are using them for, there isn't really
a problem (except that Vision, Speed, and Carrying Capacity would be
more clear as names). As soon as you start giving a damage bonus for
Strength (or should that be Dexterity? No wait! Perception helps you
spot the weaknesses in your opponent's defense...) you will start down
the dark path.
Go back and read the threads Jeff and I have pointed to. They cover the
arguments pretty comprehensively.
Billy Bissette
news:46cac771$0$16352$8826...@news.teranews.com:
Pen and paper RPGs are a different beast though. A good GM can make
a good game out of nearly any system. Stats, skills, and the entire
game engine are just aids. Or crutches, in the case of a poor GM.
Most CRPGs and Roguelikes aspire to reach even "poor GM" status.
Which isn't to say you can't make a computer game that removes
those abstractions. But if you really do, you might even start
drifting into another game type, unless you also come up with a
much better emulation of a GM.
Timofei Shatrov
confuse everyone with this message:
>On Tue, 21 Aug 2007 14:49:20 -0000, Gamer_2k4 <game...@gmail.com>
>wrote:
>
>>Well, I agree that you shouldn't make six different stats from the
>>start and base your game on them, but it's not wrong to use them. For
>>example, I add the stats as they're used. As soon as I got LOS going,
>>I made a Perception stat that determines the radius. As soon as I had
>>my speed system, I added a Dexterity stat that determines how fast you
>>are. As soon as I added items and inventory, I added a Strength stat
>>that determined how much you can carry. Maybe I'll rename or condense
>>these attributes later, but for now, they serve a purpose, and don't
>>keep me from "working on the actual game."
>
>And as long as those are all you are using them for, there isn't really
>a problem (except that Vision, Speed, and Carrying Capacity would be
>more clear as names). As soon as you start giving a damage bonus for
>Strength (or should that be Dexterity? No wait! Perception helps you
>spot the weaknesses in your opponent's defense...) you will start down
>the dark path.
He starts down the path which has been employed successfully in all popular
roguelikes. On the other hand, there is no "big" roguelike that I know off
which employs the stat-less approach you're talking about. I think it's clear
which path is better to create "actual games".
andrewdoull
> andrewdoull wrote:
> >So how many pounds / kilograms can a strength 18 character carry?
>
> If I were going to write a carrying capacity based on some property of
> items more simulationist than item type and item stackability, I would
> use a simple formula for relating physical prowess to carrying capacity.
> For example, 5kg unencumbered capacity per point of the Body stat (which
> stat is also used to determine melee damage bonus and also per-level hit
> point gain).
>
> If you have an accompanying document entitled "Levels considered
> harmful", then sir, I cordially invite you to eat it, seasoned with the
> sauce of your choice.
Hmm. I'll point you in the direction of said document
(http://www.gamasutra.com/features/20060222/sirlin_01.shtml) written by someone
with a little more experience in game design than myself.
Of course, it only applies to PvP gaming, which is not what we are talking about
here.
Andrew
PS: I should point out I'm defending a position in order to provoke people to
think about why they have stats. I'll prefer to play that hypocritical card
rather than eat my own words.
andrewdoull
> On Tue, 21 Aug 2007 12:31:27 -0700, R. Dan Henry tried to
> confuse everyone with this message:
>
> >On Tue, 21 Aug 2007 14:49:20 -0000, Gamer_2k4
> >wrote:
> >
> >>Well, I agree that you shouldn't make six different stats from the
> >>start and base your game on them, but it's not wrong to use them. For
> >>example, I add the stats as they're used. As soon as I got LOS going,
> >>I made a Perception stat that determines the radius. As soon as I had
> >>my speed system, I added a Dexterity stat that determines how fast you
> >>are. As soon as I added items and inventory, I added a Strength stat
> >>that determined how much you can carry. Maybe I'll rename or condense
> >>these attributes later, but for now, they serve a purpose, and don't
> >>keep me from "working on the actual game."
> >
> >And as long as those are all you are using them for, there isn't really
> >a problem (except that Vision, Speed, and Carrying Capacity would be
> >more clear as names). As soon as you start giving a damage bonus for
> >Strength (or should that be Dexterity? No wait! Perception helps you
> >spot the weaknesses in your opponent's defense...) you will start down
> >the dark path.
>
> He starts down the path which has been employed successfully in all popular
> roguelikes. On the other hand, there is no "big" roguelike that I know off
> which employs the stat-less approach you're talking about. I think it's clear
> which path is better to create "actual games".
DoomRL. You only live once. Rogue. [1]
[1] IIRC (I believe it has 1 stat, which avoids a lot of the problems we are
talking about).
And for "actual games" - one of my big design criteria is 'did Zelda do it?' I
think you can argue that is quite a successful franchise, although not
necessarily in the roguelike genre.
David Damerell
>So I've put up an article on my blog,
So post it to the group.
--
David Damerell <dame...@chiark.greenend.org.uk> flcl?
Today is Olethros, August - a weekend.
David Damerell
>This kind of purging of unnecessary stats and abstractions have been
>going on for some while now in the paper and pen rpg world, and it
>has produced some very nice playable systems.
But they're trying to do something completely different - unsurprisingly,
since roguelikes and other single-player computer games aren't
roleplaying games.
That mess of numbers that isn't necessary for playing a character is often
what provides structure to a tactical challenge.
Brog
>
> He starts down the path which has been employed successfully in all popular
> roguelikes. On the other hand, there is no "big" roguelike that I know off
> which employs the stat-less approach you're talking about. I think it's clear
> which path is better to create "actual games".
Roguelikes already reached their peak long ago (in Rogue*) and that
had exactly one stat. (actually four: hp, max hp, strength, max
strength. one sounds better.)
I don't get why all you people are messing with that perfect formula.
Stick with what's been proven to work, dammit!
It's clear which path has been used most often in the past to create
"actual games". Maybe the next big roguelike will not have a unified
stat system. I see no reason to claim that one way is better than the
other, but one has been done to death and the other is reasonably
fresh.
It is always good to question assumptions. And there is always
somebody who complains that they're obviously necessary because they
have worked in the past. Game developers should be prototyping new
game mechanics, not reimplementing what has been done before.
Brog
* If you doubt this, ask yourself: how many other roguelikes have
spawned an entire genre of games?
andrewdoull
> Quoting andrewdoull :
> >So I've put up an article on my blog,
>
> So post it to the group.
I'd previously asked about posting a development blog to rgrd on a regular
basis. People said I should maintain it as a separate blog, which I have done
so.
I have then posted occasionally (but I think no more than once a month) to the
group, when I've put up a design article that may be of interest to people on
here.
If anyone else expresses an interest, I can post it here. But since the article
contains links back to other blog articles that I have written, and is probably
best read in conjunction with them, I suspect the original location is best.
Having said that, the rgrd community generates plenty of response on the group,
so people here aren't exactly missing out on the discussion (and frankly, Usenet
is a better place for follow up discussion of this kind).
As an aside, Jeff Lait, and R Dan Henry - I'd like to copy the links that you
have provided and reference them in a follow up article with your permission.
David Damerell
>He starts down the path which has been employed successfully in all popular
>roguelikes. On the other hand, there is no "big" roguelike that I know off
>which employs the stat-less approach you're talking about. I think it's clear
>which path is better to create "actual games".
It may not be Big, but POWDER's an actual game and it's a lot of fun.
I've been pondering stats in large discrete chunks like POWDER skills. At
the start of the game, your character is one of "strong", "smart", or
"fast", rather than a mass of 18/140 and 53%...
David Damerell
>David Damerell <dame...@chiark.greenend.org.uk> wrote:
>>Quoting andrewdoull :
>>>So I've put up an article on my blog,
>>So post it to the group.
>I'd previously asked about posting a development blog to rgrd on a regular
>basis.
Which is a bad idea. But if you're got a specific article, as you have
here, which you think is worth discussion on rgr.dev - post it on rgr.dev.
>best read in conjunction with them, I suspect the original location is best.
No, it's not. The location I'm looking in is best. What you're saying is
"I'm going to make my article harder to read than any other on rgr.dev".
David Damerell
>On 2007-08-21 19:58:20, Martin Read <mpr...@chiark.greenend.org.uk> wrote:
>>If you have an accompanying document entitled "Levels considered
>>harmful", then sir, I cordially invite you to eat it, seasoned with the
>>sauce of your choice.
>Hmm. I'll point you in the direction of said document
>(http://www.gamasutra.com/features/20060222/sirlin_01.shtml)
Which is about _multiplayer games_. Yes, levels and levelling are a vast
bleeding ulcer in the MMO genre (and it's all down to the "RPG"
terminological idiocy); I seem to remember writing something on the
subject myself for the purpose of getting hate mail from the City of
Heroes communities...
What's that tell us about roguelikes? Not a lot.
Gerry Quinn
> I guess I should have made it clear I'm talking about player stats. I'm also not
> talking about all abstractions: anything relating to a game-mechanic is fine. If
> you have arm wrestling in your roguelike, an strength indicator is probably
> required.
>
> However, if all that strength does is increase damage and carrying capacity,
> then why have a strength stat. Increase damage and carrying capacity directly
> instead.
What I think is the problem with stats is allowing them to change, as
in Angband. Abstract stats at the start are usable IMO. Suppose, for
example, you are rolling/choosing a new character. You might allow
some points be put into variations on the standard, like so:
Trait (point cost)
------------------
Consequence(s)
Large (cost 1)
--------------
+ 15% melee damage
+ 5% ranged damage
- 10% stealth
Huge (cost 2)
-------------
+ 30% melee damage
+ 10% ranged damage
- 40% stealth
Can smash open some chests and doors
Stealthy (cost 1)
-----------------
+ 25% stealth
Large and huge would be mutually exclusive, i.e. 1 or 2 points into a
'Size' trait.
Of course there would be other attractive ways to spend the points, but
combinations of the above would be attractive to warriors - indeed it
is improbable that a melee type would not pick Large or Huge. A
conservative warrior might pick Large and Stealthy. A semi-warrior
might just take Large and save a point for a bit of magic or whatever.
- Gerry Quinn
Gerry Quinn
> Hmm. I'll point you in the direction of said document
> (http://www.gamasutra.com/features/20060222/sirlin_01.shtml) written by someone
> with a little more experience in game design than myself.
>
> Of course, it only applies to PvP gaming, which is not what we are talking about
> here.
It's basically a rant that misses the point that WoW quite simply is
not intended to be the same kind of game as his beloved Street Fighter.
It is *not* a test of skill: it is implicit that unskilful players will
be able to level up and explore much of the game's content.
It *is* a massively multiplayer online role-playing game, and there are
soft issues that cannot reasonably be covered by software or simple
rules. Maybe the author should devote some time to learning how to
deal with such soft rules... or if he can't manage it, go back to
Street Fighter.
- Gerry Quinn
Brog
> In article <fagtdf$at...@news.vol.cz>, andrewdo...@gmail.com says...
I don't want to start an argument or anything, but: all of the "soft
rules" mentioned could easily be hardcoded (except the ones that
shouldn't be there in the first place). Players should not have to
play under additional restrictions to what the game code itself
imposes. It really is silly.
Yes, the rest is just ranting about WoW duelling not being a serious
competitive game, but point 6 is entirely valid. (and yes, WoW should
be considered harmful.)
Brog
Gerry Quinn
dame...@chiark.greenend.org.uk says...
> Quoting Timofei Shatrov <gr...@mail.ru>:
> >He starts down the path which has been employed successfully in all popular
> >roguelikes. On the other hand, there is no "big" roguelike that I know off
> >which employs the stat-less approach you're talking about. I think it's clear
> >which path is better to create "actual games".
>
> It may not be Big, but POWDER's an actual game and it's a lot of fun.
>
> I've been pondering stats in large discrete chunks like POWDER skills. At
> the start of the game, your character is one of "strong", "smart", or
> "fast", rather than a mass of 18/140 and 53%...
I like that too, but of course that tends to overlap with the class
concept, if you're using that.
If your game has warriors, thieves and wizards, there seems little
point in giving the player clearly suboptimal 'choices' for strong,
fast and smart.
If you don't have warriors, thieves and wizards, but you do have
strong, fast and smart, aren't you just engaged in a re-labelling
exercise?
- Gerry Quinn
Ilya Bely
andrewdoull wrote in message <faeake$1flc$1...@news.vol.cz>:
> Article is here: http://roguelikedeveloper.blogspot.com
I'll start with a bit of rant joining voices demanding to post
article body here instead of link. I have to add that the provided
link refers to the site, not the article, so if someone will be
reading this thread later, he will find it extremely difficul to read
the article as the front page of the site will be changed by that time
and contain different records.
> If a ring of Strength (+3) gives +5 to hit and increases the character's
> carrying capacity by 25 lbs, call it a ring of Strength (+5 to hit, +25
> lbs) instead
Well, that rests on assumption that attack bonus is the same in every
situation. It doesn't have to be true.
For example, in my 7DRL if there were a ring of Str (+1), by your
sugggestion it would be called something like the following
a Ring of Strength (+3 atk with 2-handed weapons and hammers, +2 with axes
and bows, +2 bow range and +1 atk with other weapons)
Essentialy, such a ring would have a partial replication of the item
table, growing longer as new items are added to the game.
Actually, a care was taken that equipment choice would be very stat
dependant. A character with +2 str would take a good use of two-handed
sword, while another one with -1 str (stats can be negative there, and
zero is the base value) would be much better with something else that do
not depend on strength much.
The bad thing is that it was too complex for a 7DRL and did nothing but
stole time that could be better spent on other parts of the game, but
that is a completely different story.
If it was not a 7DRL, but a full game, where strength could affect
more than just combat, I can easily see it becoming a lot longer
(..., +5kg inventory limit, +15kg weightlifting, +1 muscular look,
+3 break door, +1 shield bash, +2 digging, +1 smithing)
If your stat is used for only two things, probably replacing it with two
values as you suggested does make sense. But when stats are few and each
has a lot of uses, it only seem to overcomplicate things.
I'm not trying to say that stat system is better, but that a single
example, while valid, should not be induced to cover everything.
--
May your code work forever and never have a bug.
At your service, Ilya Bely
i e l y @ l . c o
l y a b g m a i m
Gerry Quinn
crys...@gmail.com says...
Point 6 is both silly and exaggerated. Making a character called
"roflcopter" will not get you banned, but it will likely get you told
to pick a new name. Similarly for terrain exploits etc., or paying a
visit to IF Airport. I suppose you might be banned if you kept on
doing it after warnings.
So tell me - how would you hard-code choice of names? You *could* make
players choose from a pre-generated list, I suppose, but that is hardly
an ideal solution.
And not all exploits are easy to fix. It is necessary to take a
pragmatic attitude. Most people *are* capable of learning how to play
multi-player games with 'soft' rules. For the rest, there is Street
Fighter, or other games with environments so simpler that exploits can
practically be hard-coded out. (Though I see Sirlin himself has a rule
about causing the game to crash when in a losing position...)
Maybe Blizzard should be more pro-active in fixing loopholes, but they
are hardly the "fascists" Sirlin accuses them of being.
Googling on related subjects finds a large amount of Sirlin's rantings.
Suffice it to say that I consider them to be inane for the most part.
But heh, it's a free world. When he makes his own game he can make the
rules for it. Until then, he needs to learn to play by the rules as
they stand.
- Gerry Quinn
Jeff Lait
> As an aside, Jeff Lait, and R Dan Henry - I'd like to copy the links that you
> have provided and reference them in a follow up article with your permission.
You may consider my permission already granted for such actions now
and in the future.
--
Jeff Lait
(POWDER: http://www.zincland.com/powder)
andrewdoull
> Quoting andrewdoull :
> >David Damerell wrote:
> >>Quoting andrewdoull :
> >>>So I've put up an article on my blog,
> >>So post it to the group.
> >I'd previously asked about posting a development blog to rgrd on a regular
> >basis.
>
> Which is a bad idea. But if you're got a specific article, as you have
> here, which you think is worth discussion on rgr.dev - post it on rgr.dev.
>
> >best read in conjunction with them, I suspect the original location is best.
>
> No, it's not. The location I'm looking in is best. What you're saying is
> "I'm going to make my article harder to read than any other on rgr.dev".
As requested:
Stats are a great convention in role-playing games. They allow players to
distinguish between strong fighters and smart wizards, wise priests and dextrous
thieves, with a minimum of effort and as a game mechanic directly feed into the
class and skill-based systems that I have discussed previously. Dungeons &
Dragons has its Str, Int, Wis, Dex, Con and Cha, Fallout has its SPECIAL system,
and so on. I bet 99% of you know what the D&D terms I just wrote are
abbreviations for, without necessarily having played the role-playing game
itself, because the stat systems have become an equal convention in computer
RPGs.
There's a recent discussion on rec.games.roguelike.angband, started by the new
maintainer of Angband about making Cha (short for Charisma, for those of you who
don't know) more important. The Cha stat at the moment just affects the prices
in shops in-game, and has little impact once the player starts getting a
multi-million gold piece purse. Andrew Sidwell has (correctly) seen that there's
not much of a point having a stat that is useless for most of the game, and the
suggestions made have fall into either getting rid of the stat completely, or
increasing its in-game utility (or a radical few suggesting no change at all).
I've got a slightly different perspective, having gone through a similar thought
process for Unangband. A big part of the problem is that for a large part of the
game of Angband, there are no interesting choices to be made around stats. Stats
are affected in 3 ways: various monster attacks can drain your stats, and force
you to buy stat recovery potions; you can find potions of gain stat in the
dungeon, which permanently increase your stat; and you can find various items
which increase your stat while you are wearing the item.
And most of the time, there's only one sensible option. There is little to no
incentive to not drink a potion of Intelligence for instance (although Andrew
has added some potions from Eyangband which increase one stat at the expense of
another). You may end up with choices around which stat improvement item to
wear, but it'll almost inevitably be a ring of Con, along with (maybe) a spell
stat item. The only points at which stat choices are important is character
creation. And its not a particularly interesting set of choices there, either,
because each class finds one stat so much more useful than the others.
Unangband has a different way of doing stats. In Un, each spell casting class
has 3 separate stats that affect spell casting: for learning spells, gaining
mana and reducing spell failure %. All stats are also useful for warrior type
classes: e.g. Int reduces the chance of thrown (or fired - not yet implemented)
items breaking and increases to-hit, Wis increases armour class, Cha does 'crowd
control'. There's two different stats that contribute to player hit points (Con
and Siz) but each has differing side-effects, that makes warriors prefer Siz and
spell casters prefer Con. There's an additional stat Agility that adds to
speed.
And there are no stat gain potions: instead you gain a set number of stat gains
every couple of levels to distribute as you wish.
But how did Angband end up in a position where a game-play mechanic doesn't have
much of an in-game impact. Stats are made to seem so important in Angband that a
fair amout of screen real estate is devoted to displaying them. And Angband is
not the only computer RPG out there that makes this decision.
A key point to make, is that stats are an RPG convention and its possible that
like other conventions I've attacked, they may not make the most sense in a
computer game setting. Stats are a great short cut to starting to develop a
character for the purpose of role-playing, and as I've pointed out previously,
not a heck of a lot of role-playing actually occurs in computer games. And more
importantly, they are a way of distinguishing different characters from each
other.
So why have stats in a single-character game at all?
Stats kind of make sense in games where you control multiple characters. You
want to differentiate each character as much as possible, and if you've already
agreed with me and got rid of classes and skills in the game, then stats are
probably the only way you have of doing this.
But in a single player game, you don't need to make this distinction. You're
you, and your distinguished from everything else in the game by virtue of the
fact you don't have to kill yourself or pick yourself up (Or as I've seen in
some .sig files: roguelikes are about killing letters to collect punctuation).
So the primary functions of stats are useless in this environment. Instead, you
should be directly exposing the game-mechanics to the player. If a ring of
Strength (+3) gives +5 to hit and increases the character's carrying capacity by
25 lbs, call it a ring of Strength (+5 to hit, +25 lbs) instead. At the start of
the game, roll-up carrying capacity, to-hit and to-dam bonus and spell failure
chance directly. You don't need to worry about these abstract and ill-defined
concepts of stats to confuse the player.
All you end up doing with stats is forcing the player to understand and memories
tables of numbers in order to figure out the weird and non-intuitive stat break
points that you've developed as a part of the game.
So get rid of stats completely. Statistically speaking, I'm sure you'll feel
even more liberated.
Gamer_2k4
> * If you doubt this, ask yourself: how many other roguelikes have
> spawned an entire genre of games?
Angband. It has more variants than Rogue does, and it uses the six
attribute system that everyone "hates."
--
Gamer_2k4
Jeff Lait
> In article <T1*lr...@news.chiark.greenend.org.uk>,
> damer...@chiark.greenend.org.uk says...
>
> > Quoting Timofei Shatrov <g...@mail.ru>:
> > >He starts down the path which has been employed successfully in all popular
> > >roguelikes. On the other hand, there is no "big" roguelike that I know off
> > >which employs the stat-less approach you're talking about. I think it's clear
> > >which path is better to create "actual games".
>
> > It may not be Big, but POWDER's an actual game and it's a lot of fun.
>
> > I've been pondering stats in large discrete chunks like POWDER skills. At
> > the start of the game, your character is one of "strong", "smart", or
> > "fast", rather than a mass of 18/140 and 53%...
>
> I like that too, but of course that tends to overlap with the class
> concept, if you're using that.
>
> If your game has warriors, thieves and wizards, there seems little
> point in giving the player clearly suboptimal 'choices' for strong,
> fast and smart.
I agree with this. Warrior/thief/wizard becomes a proxy for strong/
fast/smart, so any such bonuses should just be applied on level up.
> If you don't have warriors, thieves and wizards, but you do have
> strong, fast and smart, aren't you just engaged in a re-labelling
> exercise?
That seems somewhat unfair. One could claim any classless game is
just a "relabeling" exercise. Choosing to learn a magic skill rather
than a sword skill is just relabeling the mage/warrior choice.
The point of classless games is to let the player find their own
optimum that matches how they want to play. I think it is
particularly useful in single player/single character games where the
adventurer is almost demanded to be a jack-of-all-trades. One could,
thus, play a wizard that relies on the sword so chooses Strong as an
attribute.
As for discrete chunks, I think my view on this is clear. As a
player, I want meaningful choices that don't require me to read
manuals to figure out all of the implications. Unfortunately, us
developers seem to think spreadsheet analysis adds to the complexity
of the game as it adds to the complexity of the code. My rule of
thumb is to always try to keep in mind the noise floor. Any factor
below a certain amount will be indistinguishable from noise from the
RNG. Will a +1% bonus to damage *ever* be noticed by a player? How
many trials would they need to perform with a 1d8 longsword to notice
the +1% bonus? Especially when their feedback is how many hits it
takes to kill a creature that potentially has a random number of hit
points?
I think this is why many people want more visible numbers in a
roguelike (damage done per hit, % chance to hit) - they fear that
there are these +1% bonuses that they are missing because they have
only this granular view of the mechanics. The answer isn't to provide
these immersion breaking details. It is to reassure them that they
aren't missing such bonuses.
My goal is thus to save up the +1% increments until they can become
something substantial that *is* noticeable. One's pact with the
players should be: "If you don't notice a difference between A & B,
there isn't a difference between A & B" rather than "differentiating"
items by adding incredibly subtle effects. (Note I use the word
"goal" - I wouldn't claim I've been entirely successful :>)
andrewdoull
> On Aug 22, 4:37 am, Brog wrote:
> > * If you doubt this, ask yourself: how many other roguelikes have
> > spawned an entire genre of games?
>
> Angband. It has more variants than Rogue does, and it uses the six
> attribute system that everyone "hates."
Um. We've only established that I "hate" it for the purpose of writing an
article about "hating" it, and that I'm doing that on the basis that I'm
questioning the utility of Angband trying to find additional utility for one of
those attributes, and have some (limited) experience in having done it myself,
due to the fact I've written one of those variants.
Luckily some people seem to agree with me.
Andrew
R. Dan Henry
wrote:
>What I think is the problem with stats is allowing them to change, as
>in Angband.
ADOM, Nethack, and Crawl. Of these, Crawl is the only one which
seriously limits the stat-increase game.
R. Dan Henry
>* If you doubt this, ask yourself: how many other roguelikes have
>spawned an entire genre of games?
Moria. It not only spun off multiple variants, but one of these,
Angband, became a major roguelike that became the basis for an explosion
of roguelikes (some of which have gone quite a way from their origin).
Angband has also influenced game that have not shared code with it. So,
I count Moria as *one* in answer to your question.
Elethiomel
> Roguelikes already reached their peak long ago (in Rogue*) and that
> * If you doubt this, ask yourself: how many other roguelikes have
> spawned an entire genre of games?
Strawman. If you consider Rogue to be a roguelike, it is the first
roguelike. That is, Rogue is the original inspiration for all
roguelikes. Being the inspiration for other games, or being the first
of its genre, is not good enough to be considered the peak of
perfection. That's like saying all modern serial-production cars are
just trying to live up to the perfection of the Ford model T.
Oliver Wong
"Gerry Quinn" <ger...@indigo.ie> wrote in message
news:MPG.213646384...@news1.eircom.net...
> In article <1187785956.3...@l22g2000prc.googlegroups.com>,
> crys...@gmail.com says...
>> On Aug 23, 12:19 am, Gerry Quinn <ger...@indigo.ie> wrote:
>> > In article <fagtdf$at...@news.vol.cz>, andrewdo...@gmail.com says...
>> >
>> > > Hmm. I'll point you in the direction of said document
>> > > (http://www.gamasutra.com/features/20060222/sirlin_01.shtml)
>> > > written by someone
>> > > with a little more experience in game design than myself.
>> >
>> > It *is* a massively multiplayer online role-playing game, and there
>> > are
>> > soft issues that cannot reasonably be covered by software or simple
>> > rules.
>>
>> I don't want to start an argument or anything, but: all of the "soft
>> rules" mentioned could easily be hardcoded (except the ones that
>> shouldn't be there in the first place). Players should not have to
>> play under additional restrictions to what the game code itself
>> imposes. It really is silly.
>
> Point 6 is both silly and exaggerated. Making a character called
> "roflcopter" will not get you banned, but it will likely get you told
> to pick a new name. Similarly for terrain exploits etc., or paying a
> visit to IF Airport. I suppose you might be banned if you kept on
> doing it after warnings.
>
> So tell me - how would you hard-code choice of names? You *could* make
> players choose from a pre-generated list, I suppose, but that is hardly
> an ideal solution.
>
> And not all exploits are easy to fix. It is necessary to take a
> pragmatic attitude. Most people *are* capable of learning how to play
> multi-player games with 'soft' rules. For the rest, there is Street
> Fighter, or other games with environments so simpler that exploits can
> practically be hard-coded out. (Though I see Sirlin himself has a rule
> about causing the game to crash when in a losing position...)
If people naming their characters "roflcopter" was such a big problem
that they had to make a soft rule about it, perhaps the developers could
create a list of blacklisted names, and include "roflcopter" in that list.
At least that would prevent any new character getting created with that
name.
As new "undesirable" names are discovered, the players controlling
such characters could be asked to change those names, and the moderators
could add those names to the blacklist.
I see so-called "soft rules" as a necessary evil: Developers should
enforce as many rules as possible in the actual code, but there are some
stuff for which our technology simply isn't good enough to enforce yet.
Censoring abuse language is one example, as it probably requires AI to
determine meaning and context.
- Oliver
Ray Dillinger
>
> What I think is the problem with stats is allowing them to change, as
> in Angband.
That's a small part of what I call the "thousanduplet problem."
When all your ascended characters look so alike that any two
of them could be identical twins -- which typically happens
in games that permit/encourage you to max out all stats,
collect all intrinsics, and master all skills -- you don't
really have distinct playing experiences after some nebulous
point in the midgame where all the individuality gets ground
out of characters. 'Cause nobody really wants to be different;
they all want to be The Winning Character (and there's typically
only one way to be that). So they turn into thousanduplets.
When you allow stats to change, the problem in my mind is that
under "usual" implementations they all max out (and thereby
become identical).
So I think it may be better to have some kind of dynamic balance
where there are points of diminishing returns; No matter how many
strength potions you quaff, for example, you never get more than
double your initial or "unaided" strength. Or certain opposed
resistances cannot simultaneously affect you, eg, you can't be
resistant to both heat and cold at the same time. That cold-
resistance item will not work at all unless it's stronger than
your heat-resistance item, and in that case your heat resistance
will not work at all while you've got cold resistance on. And
so on.
Force 'em to choose.
Bear
Krice
> Um. We've only established that I "hate" it for the purpose
> of writing an article about "hating" it
I tried to play UnAngband, but the UI from stone age was
really annoying. Now when I have developed my game
and try Nethack, Angband or any major roguelike I always
wonder why those games have empty game world and nothing
else to do than fight generic monsters. It's boring!
It's no wonder that roguelike scene is really dying,
because people demand more from gameplay than
20 years ago.
R. Dan Henry
<andre...@gmail.com> wrote:
>As an aside, Jeff Lait, and R Dan Henry - I'd like to copy the links that you
>have provided and reference them in a follow up article with your permission.
I'd appreciate it if you'd forward what you want to copy to me and let
me proofread it first. I tend to write to Usenet at odd hours and some
of what I write is a bit embarrassing when looked at fully awake.
R. Dan Henry
wrote:
>On Wed, 22 Aug 2007 09:29:01 +0000 (UTC), andrewdoull
><andre...@gmail.com> wrote:
>
>>As an aside, Jeff Lait, and R Dan Henry - I'd like to copy the links that you
>>have provided and reference them in a follow up article with your permission.
>
>I'd appreciate it if you'd forward what you want to copy to me and let
>me proofread it first. I tend to write to Usenet at odd hours and some
>of what I write is a bit embarrassing when looked at fully awake.
Oh, wait, you asked permission to use the *link*. That doesn't even make
sense. Those are Google Groups links; we have no authority whatsoever
over the links. (I would suggest reading the threads in full yourself;
you may find some other post(s) in there that you prefer.)
David Damerell
>If your game has warriors, thieves and wizards, there seems little
>point in giving the player clearly suboptimal 'choices' for strong,
>fast and smart.
Are they? I mean, sure, hack-fantasy wizards all have to be smart, but we
don't have to be hack-fantasy. A strong wizard can cast more spells
without becoming exhausted; a fast one is better able to defend themselves
from attacks. A strong warrior hits hard, but a fast one can choose their
battles and a smart one will do well against intelligent opponents.
Some combinations might be suboptimal, but it doesn't have to be a
non-choice with every class having an accompanying best stat.
--
David Damerell <dame...@chiark.greenend.org.uk> Distortion Field!
Today is First Monday, August.
Gerry Quinn
ow...@castortech.com says...
Unfortunately there are 26^10 ten letter names, the majority of which
are unacceptable. So this is not a sensible solution. It is much more
reasonable to encourage players to pick suitable names for their
characters. If they can't, they aren't capable of playing properly in
a MMORPG anyway.
> I see so-called "soft rules" as a necessary evil: Developers should
> enforce as many rules as possible in the actual code, but there are some
> stuff for which our technology simply isn't good enough to enforce yet.
> Censoring abuse language is one example, as it probably requires AI to
> determine meaning and context.
I don't see how you can be aware of that, and not be aware that it
applies equally to character names.
Actually, from reading this thread and some of Sirlin's stuff I am
inclined to think that 'soft rules' in MMORPGs are a good thing insofar
as they may discourage Sirlin and the kind he lionises from playing the
game.
- Gerry Quinn
Gerry Quinn
torespon...@hotmail.com says...
> My goal is thus to save up the +1% increments until they can become
> something substantial that *is* noticeable. One's pact with the
> players should be: "If you don't notice a difference between A & B,
> there isn't a difference between A & B" rather than "differentiating"
> items by adding incredibly subtle effects. (Note I use the word
> "goal" - I wouldn't claim I've been entirely successful :>)
I agree with that principle - you can see from my example of 'large',
'huge' and 'stealthy' traits the sort of granularity I would aim at.
- Gerry Quinn
Gerry Quinn
dame...@chiark.greenend.org.uk says...
> Quoting Gerry Quinn <ger...@indigo.ie>:
> >If your game has warriors, thieves and wizards, there seems little
> >point in giving the player clearly suboptimal 'choices' for strong,
> >fast and smart.
>
> Are they? I mean, sure, hack-fantasy wizards all have to be smart, but we
> don't have to be hack-fantasy. A strong wizard can cast more spells
> without becoming exhausted; a fast one is better able to defend themselves
> from attacks. A strong warrior hits hard, but a fast one can choose their
> battles and a smart one will do well against intelligent opponents.
>
> Some combinations might be suboptimal, but it doesn't have to be a
> non-choice with every class having an accompanying best stat.
True, but I think you are tweaking the conventional meanings of the
terms a bit here... I have never seen a CRPG, for example, in which
mana is proportional to strength.
- Gerry Quinn
Gerry Quinn
> So the primary functions of stats are useless in this environment. Instead, you
> should be directly exposing the game-mechanics to the player. If a ring of
> Strength (+3) gives +5 to hit and increases the character's carrying capacity by
> 25 lbs, call it a ring of Strength (+5 to hit, +25 lbs) instead.
You are carrying:
...
(c) A Ring of Strength (+5 to hit, +25 pounds carried, +5% to break
open chests, +10% to bash down a door, +20% chance of winning in Basho
the Troll's arm-wrestling contest, +3% chance of breaking free from an
anaconda, +2% to stun wen wielding a mace....
...
- Gerry Quinn
Christophe
You are carrying a ring of Strength (+1+(curStr%10==0?+2*curStr+1:0))
What, I am not allow to have str damage bonus equal to str+(str/10)²?
And don't ask me how to write it with more than one Strength rign
equipped :)
David Damerell
>dame...@chiark.greenend.org.uk says...
>>Are they? I mean, sure, hack-fantasy wizards all have to be smart, but we
>>don't have to be hack-fantasy.
>terms a bit here...
I'm going to have to, though, if I'm going to have strong/fast/smart and
warriors/thieves/wizards and not have it be a non-choice; and I think
there's a lot to be said for _not_ following the hack fantasy cliches
blindly.
[There is something to be said for them, too. Everyone knows what a
wizard is, and that dragons are dangerous. But there's room for a happy
middle ground.]
I was struck by this when playing Majesty. Everything in that is
completely a cliche, so for example all wizards are old. What did they do
beforehand?
"Morning, Bob. How's things? My arthritis is killing me."
"Pretty good, Fred. I woke up this morning and, fuck me, I can cast
fireballs."
>I have never seen a CRPG, for example, in which mana is proportional
>to strength.
But there are plenty where "finesse" fighters coexist with meatshields.
R. Dan Henry
<torespon...@hotmail.com> wrote:
>On Aug 22, 8:43 am, Gerry Quinn <ger...@indigo.ie> wrote:
>> If your game has warriors, thieves and wizards, there seems little
>> point in giving the player clearly suboptimal 'choices' for strong,
>> fast and smart.
>
>I agree with this. Warrior/thief/wizard becomes a proxy for strong/
>fast/smart, so any such bonuses should just be applied on level up.
The alternative is to use stats as play style modifiers. Either as
replacements for classes (as has been suggested in previous discussions)
or as David suggests, for different subtypes within classes. A strong
warrior uses big two-handed weapons, a fast warrior is a swashbuckler
with a light blade, a smart warrior has a lot of different skills to
fall back on.
>> If you don't have warriors, thieves and wizards, but you do have
>> strong, fast and smart, aren't you just engaged in a re-labelling
>> exercise?
Yes and no. The stats make nicer sliding values and gaining smarts or
strength on level up seems more natural than switching from thief to
wizard to barbarian while in the dungeon doing the same real job all the
time (POWDER nicely gets around the plausibility issue here with tying
its classes to the gods)*. It also means you can put points into
intelligence for the item-recognition or skill-slot gains, without being
labeled a wizard when it isn't the spells you want.
>The point of classless games is to let the player find their own
>optimum that matches how they want to play. I think it is
>particularly useful in single player/single character games where the
>adventurer is almost demanded to be a jack-of-all-trades. One could,
>thus, play a wizard that relies on the sword so chooses Strong as an
>attribute.
Or puts points into Strength because he wants to be the kind of wizard
who hauls around a ton of staves, wands, crystals, potions, and arcane
tomes.
>My goal is thus to save up the +1% increments until they can become
>something substantial that *is* noticeable. One's pact with the
>players should be: "If you don't notice a difference between A & B,
>there isn't a difference between A & B" rather than "differentiating"
>items by adding incredibly subtle effects. (Note I use the word
>"goal" - I wouldn't claim I've been entirely successful :>)
The problem is that until you about double effectiveness, a fairly
random system, combined with selective memory, is going to prevent
correctly noticing differences. If you're happy de-randomizing your
combat system, sure, go ahead, but otherwise I think you are
unnecessarily (but not necessarily harmfully) limiting the things you
allow in your game.
* - Oh, wait, this is replying to Jeff. I'd better balance that comment
to avoid inflating his ego. :-) Well, I finally tried POWDER and was
de-immersed by the casting of Acid Splash when the backsplash destroyed
my Book of Acid. I'd think if anything would have been acid-proofed,
it'd have to be a *magic book that contains a spell that's likely to get
acid on it!*.
R. Dan Henry
wrote:
>In article <O7B*7R...@news.chiark.greenend.org.uk>,
>dame...@chiark.greenend.org.uk says...
>> Some combinations might be suboptimal, but it doesn't have to be a
>> non-choice with every class having an accompanying best stat.
>
>True, but I think you are tweaking the conventional meanings of the
>terms a bit here... I have never seen a CRPG, for example, in which
>mana is proportional to strength.
No, but I have seen PnP games where spell-casting was tied to physical
exhaustion and CON (or whatever the game's equivalent) would be helpful
to a wizard type.
R. Dan Henry
<andre...@gmail.com> wrote:
>A key point to make, is that stats are an RPG convention and its possible that
>like other conventions I've attacked, they may not make the most sense in a
>computer game setting. Stats are a great short cut to starting to develop a
>character for the purpose of role-playing, and as I've pointed out previously,
>not a heck of a lot of role-playing actually occurs in computer games.
This is where I disagree with you the most. Stats don't really help with
role-playing. All intelligence is homogenized into a single number, for
example, whereas in a single paragraph description, you can distinguish
easily between "bookish", "witty", or "a careful reasoner". The
advantage of stats is mechanical. If you come across a situation for
which there is no specific rule/skill, you can fall back on checking
against a stat. Vorag the Mutilator wants to tear the Book of Evil
Summoning in half to thwart the Purple Necromancer? Make a Strength
check. Need to know if an NPC solves a puzzle? Intelligence check. A
character wants to build a house of cards? Dexterity check.
But in a computer game, the player can only do those things you program
into it. You can make your skill system comprehensive. You don't need a
fall-back mechanic.
Radomir 'The Sheep' Dopieralski
R Dan Henry wrote:
> But in a computer game, the player can only do those things you program
> into it. You can make your skill system comprehensive. You don't need a
> fall-back mechanic.
Actually if you somehow manage to implement enough direct control
(a la Mario, where you have to actually control your character to
jump over the pit, not just select "jump over this pit please"
from the menu), then you can observe your players doing things
you never actually thought of. Of course, one can argue that you
programmed them into it, by introducing in the the equations and
parameters and what not. But you didn't know all the possible ways
of using them.
It's like the Lego blocks don't have everything you can actually
build from them pre-designed.
--
Radomir `The Sheep' Dopieralski <http://sheep.art.pl>
() ascii ribbon campaign - against html e-mail
/\ <www.asciiribbon.org> - against proprietary attachments
Jeff Lait
>
> * - Oh, wait, this is replying to Jeff. I'd better balance that comment
> to avoid inflating his ego. :-) Well, I finally tried POWDER and was
> de-immersed by the casting of Acid Splash when the backsplash destroyed
> my Book of Acid. I'd think if anything would have been acid-proofed,
> it'd have to be a *magic book that contains a spell that's likely to get
> acid on it!*.
I've never claimed that the wizards of the POWDER world were
particularly bright. Remember the leading cause of death among
necromancers is Finger of Death accidents!
>From the encyclopedia:
When one considers the years of study required to learn this most
potent spell, one would think the users would be wise. Nonetheless,
careless Finger of Death accidents claim dozens of necromancers every
year! Do not become another statistic!
And, of course, we have the likes of Prakal as mentioned in the Entomb
spell description:
It is said that Prakal, the Earth Mage, perfected a version of this
spell which does not require Dig as a prerequisite. We can only guess
what it was. Prakal's mummified remains, found buried in solid rock,
have told us nothing.
Of course, thanks to the RNG, it is possible for your tome of acid to
be made of glass or gold, in which case it will be safe from such
backsplash. In other games, be wary!
Ray Dillinger
> True, but I think you are tweaking the conventional meanings of the
> terms a bit here... I have never seen a CRPG, for example, in which
> mana is proportional to strength.
It would be a neat system, IMO, if magic were powered by your
physical metabolism. Every spell gives you extra hunger, spread
over the next 24 hours.
So if you use a spell that costs, say, 10 mana - you'd have to
eat an extra ten meals within in the next day to make up for it.
Wizards would always be hungry. Hungrier than trolls.
Bear
Gerry Quinn
says...
Actually, Crawl is a bit like that.
- Gerry Quinn
Gerry Quinn
danh...@inreach.com says...
> >My goal is thus to save up the +1% increments until they can become
> >something substantial that *is* noticeable. One's pact with the
> >players should be: "If you don't notice a difference between A & B,
> >there isn't a difference between A & B" rather than "differentiating"
> >items by adding incredibly subtle effects. (Note I use the word
> >"goal" - I wouldn't claim I've been entirely successful :>)
>
> The problem is that until you about double effectiveness, a fairly
> random system, combined with selective memory, is going to prevent
> correctly noticing differences. If you're happy de-randomizing your
> combat system, sure, go ahead, but otherwise I think you are
> unnecessarily (but not necessarily harmfully) limiting the things you
> allow in your game.
But is it a loss to leave out things that players won't notice?
- Gerry Quinn
R. Dan Henry
<torespon...@hotmail.com> wrote:
>On Aug 23, 3:44 pm, R. Dan Henry <danhe...@inreach.com> wrote:
>>
>> * - Oh, wait, this is replying to Jeff. I'd better balance that comment
>> to avoid inflating his ego. :-) Well, I finally tried POWDER and was
>> de-immersed by the casting of Acid Splash when the backsplash destroyed
>> my Book of Acid. I'd think if anything would have been acid-proofed,
>> it'd have to be a *magic book that contains a spell that's likely to get
>> acid on it!*.
>
>I've never claimed that the wizards of the POWDER world were
>particularly bright. Remember the leading cause of death among
>necromancers is Finger of Death accidents!
Now, you see what happens when you don't have an INT requirement for
spell-casting?!? I think the balance of argument has just swung back in
favor of the pro-stats side.
R. Dan Henry
wrote:
>In article <b1krc355dt1ha3tno...@4ax.com>,
Yes, if you mean things players won't notice when left with nothing but
their own subjective impressions. You eliminate a lot of tactical gaming
choices. +5% is generally significant, but not enough for subjective
notice. +25% could easily escape notice, with a fairly random combat
system. I really don't think you can count on the player having a good
idea of superiority from use until you have a doubling of effectiveness
and that's a lot courser grain than I'd want for most purposes (now for
something like a superhero game, a logarithmic scaling makes sense).
Jeff Lait
> On Fri, 24 Aug 2007 13:10:04 +0100, Gerry Quinn <ger...@indigo.ie>
> wrote:
>
> >danhe...@inreach.com says...
> >> On Wed, 22 Aug 2007 13:36:31 -0000, Jeff Lait
> >> <torespondisfut...@hotmail.com> wrote:
>
> >> >My goal is thus to save up the +1% increments until they can become
> >> >something substantial that *is* noticeable. One's pact with the
> >> >players should be: "If you don't notice a difference between A & B,
> >> >there isn't a difference between A & B" rather than "differentiating"
> >> >items by adding incredibly subtle effects. (Note I use the word
> >> >"goal" - I wouldn't claim I've been entirely successful :>)
>
> >> The problem is that until you about double effectiveness, a fairly
> >> random system, combined with selective memory, is going to prevent
> >> correctly noticing differences. If you're happy de-randomizing your
> >> combat system, sure, go ahead, but otherwise I think you are
> >> unnecessarily (but not necessarily harmfully) limiting the things you
> >> allow in your game.
>
> >But is it a loss to leave out things that players won't notice?
>
> Yes, if you mean things players won't notice when left with nothing but
> their own subjective impressions. You eliminate a lot of tactical gaming
> choices. +5% is generally significant, but not enough for subjective
> notice. +25% could easily escape notice, with a fairly random combat
> system. I really don't think you can count on the player having a good
> idea of superiority from use until you have a doubling of effectiveness
> and that's a lot courser grain than I'd want for most purposes (now for
> something like a superhero game, a logarithmic scaling makes sense).
First, I am less pessimistic than you about people's ability to notice
bonuses. I think +5% to hit is likely on the boundary of threshold,
while I think you would put it even higher.
Second, there is more than one way to give +5% effectiveness. If the
increased effectiveness is a change-of-kind rather than just
incremental, it is a lot more noticeable. Learning a fireball spell
is like this - you are X% more effective than when you only knew
sticky flames and magic missile (the prereqs), but regardless of the
magnitude of X, people notice it.
One thing I was very impressed with Crawl was the "You head-butt the
orc" type messages. Head butting could be hidden as an extra 5% bonus
one would never notice to attack, but adding an explicit message means
one will notice it.
The parry skill in POWDER provides about a 10% dodge bonus. However,
it is not required for the player to notice the slight change in hit
ratios - successful parries are described as such.
Magic weapons tend to telegraph their bonus in a similar fashion: +1
is +1 so is better :> Following Crawl's lead, POWDER auto-ids the
bonus after some combat, ensuring no one has to try and determine the
5% differential from experiment.
I also disagree about the importance over all of 5% in a roguelike.
They aren't, generally, games of attrition. Encounters tend to be
binary: Survive or Die. Contrast with a strategy game like Galactic
Civilizations where I will happily grab a 1% bonus because the game
tends to be all about compound interest.
Radomir 'The Sheep' Dopieralski
Jeff Lait wrote:
> On Aug 24, 2:18 pm, R. Dan Henry <danhe...@inreach.com> wrote:
>> Yes, if you mean things players won't notice when left with nothing but
>> their own subjective impressions. You eliminate a lot of tactical gaming
>> choices. +5% is generally significant, but not enough for subjective
>> notice. +25% could easily escape notice, with a fairly random combat
>> system. I really don't think you can count on the player having a good
>> idea of superiority from use until you have a doubling of effectiveness
>> and that's a lot courser grain than I'd want for most purposes (now for
>> something like a superhero game, a logarithmic scaling makes sense).
> First, I am less pessimistic than you about people's ability to notice
> bonuses. I think +5% to hit is likely on the boundary of threshold,
> while I think you would put it even higher.
With non-random combat, the players are likely to notice bonuses when
the number of hits required to kill a monster changes. For example, if
you kill a monster with 2 hits, the smallest bonus is 200%. If it's 3
hits, 150% will be enough. "I was killing mutant robots with two hits,
now I kill them with a single blow." Now, if there are several kinds of
monsters on a level and they differ in the amount of hit points, you can
have smaller thereshold: "I was killing both mutant robots and zombie
flamingos with two hits, but now I can kill zombie flamingos with a single
blow, while mutant robots still require two hits." Of course, the players
won't notice if one of the kinds is especially rare or if the number of
hits required is larger than 5-6 -- you stop counting then.
Introducing randomness allows you to have much smaller bonuses that still
have impact on the actual gameplay, but are harder to notice and can even
make the player believe in things that are not true (if he encounters a
monster that is especially weak for its kind, and he scores an especially
lucky blow, he will remember that and -- even if the majority of
encounters is not like that at all -- people tend to remember "special
cases", not the average ones).
> Second, there is more than one way to give +5% effectiveness. If the
> increased effectiveness is a change-of-kind rather than just
> incremental, it is a lot more noticeable. Learning a fireball spell
> is like this - you are X% more effective than when you only knew
> sticky flames and magic missile (the prereqs), but regardless of the
> magnitude of X, people notice it.
Change in quality instead of quantity is an excellent solution! Although
it requires the designer to work much harder, coming up with things that
actually differ. Obviously, interesting choices cannot be generated
automatically.
> One thing I was very impressed with Crawl was the "You head-butt the
> orc" type messages. Head butting could be hidden as an extra 5% bonus
> one would never notice to attack, but adding an explicit message means
> one will notice it.
Bringing the attention of players is another very good technique. A sword
named "flammingo flamer" will seem much more effective against zombie
flamingos, ironically, even when it has no real bonuses!
Gerry Quinn
danh...@inreach.com says...
> On Fri, 24 Aug 2007 13:10:04 +0100, Gerry Quinn <ger...@indigo.ie>
> wrote:
> >In article <b1krc355dt1ha3tno...@4ax.com>,
> >danh...@inreach.com says...
> >> On Wed, 22 Aug 2007 13:36:31 -0000, Jeff Lait
> >> <torespon...@hotmail.com> wrote:
> >
> >> >My goal is thus to save up the +1% increments until they can become
> >> >something substantial that *is* noticeable. One's pact with the
> >> >players should be: "If you don't notice a difference between A & B,
> >> >there isn't a difference between A & B" rather than "differentiating"
> >> >items by adding incredibly subtle effects. (Note I use the word
> >> >"goal" - I wouldn't claim I've been entirely successful :>)
> >>
> >> The problem is that until you about double effectiveness, a fairly
> >> random system, combined with selective memory, is going to prevent
> >> correctly noticing differences. If you're happy de-randomizing your
> >> combat system, sure, go ahead, but otherwise I think you are
> >> unnecessarily (but not necessarily harmfully) limiting the things you
> >> allow in your game.
> >
> >But is it a loss to leave out things that players won't notice?
>
> Yes, if you mean things players won't notice when left with nothing but
> their own subjective impressions.
I said things the player will not notice. If the sword is called
Orcslayer, or has a description like "this sword does +5% damage
against orcs", the player will notice this, even if he is not
collecting detailed battle statistics.
I agree that in such cases, you can use a finer-grained system, because
you are disclosing the information elsewhere.
> You eliminate a lot of tactical gaming
> choices.
If they are not noticed by the player, they cannot be choices.
> +5% is generally significant, but not enough for subjective
> notice. +25% could easily escape notice, with a fairly random combat
> system. I really don't think you can count on the player having a good
> idea of superiority from use until you have a doubling of effectiveness
> and that's a lot courser grain than I'd want for most purposes (now for
> something like a superhero game, a logarithmic scaling makes sense).
If you don't want a game in which items are neatly labelled with its
exact powers, then I still say there is no point in a fine-grained
system.
- Gerry Quinn
Ray Dillinger
> If you don't want a game in which items are neatly labelled with its
> exact powers, then I still say there is no point in a fine-grained
> system.
>
I dunno. I kind of like the model where there are things hidden
from the player except for small clues and sometimes it pays to
take notes. Or just to be suspicious and cautious about things
even if you don't know exactly why. Or to do things for completely
superstitious reasons because they seem correlated with effects
you don't understand how you produced the first time. Etc.
I think winning should involve a lot of guessing and hoping and
being relieved when one is shown to be right. Or alternatively,
starting a new character as a slightly wiser player when one is
shown to have been wrong.
Bear
R. Dan Henry
wrote:
Well, since the whole issue was whether fine-grained systems are
suitable, then we agree. Although I will say that I don't like
fine-grained systems because I like giving out +1% bonuses (or -1%
penalties), but because I like being able to differentiate more and less
adept types by given +5% or +3% per level bonuses to those type rather
than +4% for normal types. That sort of thing.
>> You eliminate a lot of tactical gaming
>> choices.
>
>If they are not noticed by the player, they cannot be choices.
Which is why I don't like hidden mechanics.
>If you don't want a game in which items are neatly labelled with its
>exact powers, then I still say there is no point in a fine-grained
>system.
And I think that you shouldn't hide mechanics unless there's some
gameplay in discovering them.
R. Dan Henry
<torespon...@hotmail.com> wrote:
>On Aug 24, 2:18 pm, R. Dan Henry <danhe...@inreach.com> wrote:
>> On Fri, 24 Aug 2007 13:10:04 +0100, Gerry Quinn <ger...@indigo.ie>
>> wrote:
>> >But is it a loss to leave out things that players won't notice?
>>
>> Yes, if you mean things players won't notice when left with nothing but
>> their own subjective impressions. You eliminate a lot of tactical gaming
>> choices. +5% is generally significant, but not enough for subjective
>> notice. +25% could easily escape notice, with a fairly random combat
>> system. I really don't think you can count on the player having a good
>> idea of superiority from use until you have a doubling of effectiveness
>> and that's a lot courser grain than I'd want for most purposes (now for
>> something like a superhero game, a logarithmic scaling makes sense).
>
>First, I am less pessimistic than you about people's ability to notice
>bonuses. I think +5% to hit is likely on the boundary of threshold,
>while I think you would put it even higher.
Without meticulous note-taking and facing enemies with a constant
difficulty (if different monsters have different evasive ability, it all
becomes much more complex), a 5% difference will be swamped by selective
memory. With a system simply as random as mimicking D&D style straight
die rolls, you'll be taking notes for quite awhile before getting enough
data to judge. With a system like Crawl's, even if you understood all
the calculation, if the game hid the numbers from you, you'd have little
hope of working out values for a given weapon.
>Second, there is more than one way to give +5% effectiveness. If the
>increased effectiveness is a change-of-kind rather than just
>incremental, it is a lot more noticeable. Learning a fireball spell
>is like this - you are X% more effective than when you only knew
>sticky flames and magic missile (the prereqs), but regardless of the
>magnitude of X, people notice it.
Yes, indeed, they'll notice this even if you've screwed up and made
fireball no more useful than prereqs. I delight in seeing my sword go
from +1 to +2, not because I actually notice any increase in
effectiveness, but because I know +2 is supposed to be better and trust
that there isn't a bug preventing that from being so.
>The parry skill in POWDER provides about a 10% dodge bonus. However,
>it is not required for the player to notice the slight change in hit
>ratios - successful parries are described as such.
What about Dodge? I haven't actually noticed any effects from it. How
much is it helping?
>I also disagree about the importance over all of 5% in a roguelike.
>They aren't, generally, games of attrition. Encounters tend to be
>binary: Survive or Die.
And which you do can depend on a very slight edge at times. The more
fights you get into (and roguelikes tend to involve a lot of fights),
the more likely that 5% edge will make a difference in one of them. I'll
buy a scroll to get that +1 damage on my 1d20 damage pike, thanks. If I
can secure a permanent 5% price advantage when buying scrolls by
quaffing a Potion of the Bargaining Tongue, that's good, too.
I certainly am not saying everything needs to be fine-grained or that
qualitative or large increases in ability aren't more exciting than
small increases in general. On the other hand, I'd just as soon have a
+1 bonus every level, rather than +5 every five levels. If I can
increase my spell-casting success by 1% for every sparkling pixie mage I
eat, I'm going to consider it worth eating every sparkling pixie mage I
can find (and using my Cleaver of +5% Chance of Getting a Corpse to kill
them with). Qualitative differences should be more common when it comes
to items, but even POWDER doesn't abandon quantitative differences
entirely (+/- to items).