Make all stats useful for all chars

[lessens]

The burden of proof is on the side who wants to change the status quo.

I don't see why that would be the case either from a logical or game design perspective.

[johlstei]

My point is purely that there is a built-in downside to having stats, as with any other mechanic. I'm not really trying to make a case either way - I'm only making a case that it is appropriate to question their existence. I disagree about the status quo - if no one (read: no developer) has a good reason why stats should stay around, they should be removed.

It seems like skill training and aptitudes are a pretty good way to differentiate characters and races already, and I'm okay with funneling other similar mechanics into that. A race currently is special for it's high int? Increase it's spell aptitudes. Likewise for strength? Increase its armour aptitude.

[dowan]

OK, I can play a DE with 27 conj and spellcasting, and put all my lvl up points into dex, and then play another DE with the same skills, but all lvl up points in int. These characters will be somewhat different. If you remove stats, now all DEs are the same. All that matters is skill level, where currently int can make up for a lot of missing skill levels.

I just don't see what this proposed change accomplishes, beside making all characters more similar.

This argument seems fully general. Those characters would not be very different. It'd be easy to make a proposal that adds more differentiation, it's just that the complexity involved in reasoning about them is unwarranted for what you gain.

Imagine there are 3 more stats now, wis, cha, and con. A point of end gives you roughly 1 hp, a point of wis makes your spellcasting a bit better, but not as much as int does, and makes your dodging a little better, but not as much as dex does. A point of cha gives you roughly 1% more piety with applicable gods, and a tiny bit of MR. Would you support adding this? It allows for more differentiation, but now you also have to reason about whether 2 wis is better than 1 dex and 1 int, which may be different in different contexts, and how important piety is to you, etc. It adds complex decisions for a very marginal effect. In my mind, stats as they exist may well be similar in that sense. Using your reasoning, we could never remove anything that technically added differentiation even if it was a large mental burden for a marginal benefit.

Well, the thing is, characters already are quite well differentiated with the existing 3 stats. Removing those 3 removes the differentiation completely. Adding 3 more stats maybe creates a little more differentiation, but at a significant complexity cost.

It's sort of like saying either we should only have 1 weapon type, because all these different types are needlessly complex and confusing, or we should have 5000 weapon types, because it's good to have different kinds of weapons for different play styles.

My reasoning isn't differentiation = good, so much as I think the complexity cost for the current stat system is worth it for the benefit it provides of making characters different from one another. I don't think it's worth it to lose that just so I don't have to press S, I, or D every 3 levels.

[Sprucery]

Removing stats from the game (while maintaining balance, replayability etc.) would be a lot of work. If it can't be established before-hand that removing stats would make the game better, the developers should do other things. I thought this is self-evident.

[Sandman25]

Imagine there are 3 more stats now, wis, cha, and con. A point of end gives you roughly 1 hp, a point of wis makes your spellcasting a bit better, but not as much as int does, and makes your dodging a little better, but not as much as dex does. A point of cha gives you roughly 1% more piety with applicable gods, and a tiny bit of MR. Would you support adding this? It allows for more differentiation, but now you also have to reason about whether 2 wis is better than 1 dex and 1 int, which may be different in different contexts, and how important piety is to you, etc. It adds complex decisions for a very marginal effect. In my mind, stats as they exist may well be similar in that sense. Using your reasoning, we could never remove anything that technically added differentiation even if it was a large mental burden for a marginal benefit.

I see now.
Str should be split into Melee damage and Heavy Armour (helps with EV and spells in heavy armour) wearing.
Int should be split into Spell Power and what else? Spell Hunger?
Dex should be split into EV bonus, Stabbing Bonus and Stealth Bonus?

Is this what you are suggesting?

[Sandman25]

The burden of proof is on the side who wants to change the status quo.

I don't see why that would be the case either from a logical or game design perspective.

You are trying to make devs spend their time on changing something.

[dpeg]

I don't think that stats are a particularly holy cow (for example, we did conflate the two weapon enchantments into one). On the other hand, I don't agree that stats are overly complicated (you can see the effects quite readily if you have some stat modifying items ready). And I do think they actually matter a bit.

Now it is a subjective matter whether what they gain is sufficient or not, and whether their cost (interface, cognitive) is too high. I think they're alright. This could certainly be dev inertia speaking, but there's so many things more urgently to address, in my opinion. Moreover, while removal is always an option, it is worth thinking about improvement. For stats, there are ideas.

johlstei: I think your statement is too purist. For example, removing stats *is* work: not only the actual effort of removing all odds and ends in the code, but you also have to come up with a solution for Chei, find something for !might, !brilliance, !agility. (And no, I don't want to discuss such replacements here or now... I am happy with stats.)

Pointless bottom line: Brogue is an excellent game with a single stat, Nethack is a pretty shoddy game with six stats, so I think three stats suit Crawl just fine.

Slightly less pointless afterthought: the recent change for more variable stat modifiers on artefacts is very good, in my opinion. I do think that stats are a bit too fine-grained, making them coarser (perhaps just at a 2:1 ratio) might improve them.

[Wahaha]

It's probably wrong to say that stats are hard to understand and that they add too much mental burden. They're pretty simple. I also disagree that their effect is not significant. But if it was not significant enough, the effect should be increased instead of removed. Why? - Because stats are a part of character customization and create decisions, and I think stat decisions are in a good spot in the game. Stats are not only selected on level up, items have them too, and that does affect equipment decisions.
Making stats use a smaller scale might be nice.

[archaeo]

You are trying to make devs spend their time on changing something.

So? We're having a polite chat about mechanics on a website about elf games. I don't think there's any sense in talking about "burdens of proof," especially since I don't think quality of debate has ever been used as a criteria for changing the game.

Also, how are we going to "make" the devs do anything? They're not dummies, and we can't exactly hoodwink them.

Pointless bottom line: Brogue is an excellent game with a single stat

dpeg is pender paying for you to advertise Brogue?

It's probably wrong to say that stats are hard to understand and that they add too much mental burden. They're pretty simple.

I'm not saying that it's like doing differential calculus or something, but I see two things that add an unnecessary mental burden:

1) Stats affect a whole bunch of things. You can easily play and win the game only knowing the bare outlines of what all of those things are, but it doesn't change the fact that having a thorough understanding of how stats work requires either a bunch of tedious trial-and-error playing with stat+ items or reading spoilers.

2) It's often incredibly unclear what, exactly, will make my character better. I imagine that a healthy percentage of my stat selections are suboptimal because it's hard to know which stat is going to have the biggest effect. There are only three stats and they're all tangled up, so making the correct decision requires having all that spoilery knowledge from the first point.

Personally, I think it would be easier to remove stats and add more clear mechanical replacements than it would be to "clarify" stats in any meaningful way.

[dpeg]

Pointless bottom line: Brogue is an excellent game with a single stat

dpeg is pender paying for you to advertise Brogue?

No, and I wouldn't take his money! But having finally played Brogue after hearing so much about it (I rejected it initially because I think its visuals actually detract from the gameplay, Brogue is hard for me to read: too many needless colours), it was obvious how Brogue solves a number of issues that Crawl has. Also, Pender implemented some of my proposals. And despite all the differences between the games, there's some obvious overlap: we don't have to discuss about removing no-brainers or increasing choices, that's a given.

I'm not saying that it's like doing differential calculus or something, but I see two things that add an unnecessary mental burden:

1) Stats affect a whole bunch of things. You can easily play and win the game only knowing the bare outlines of what all of those things are, but it doesn't change the fact that having a thorough understanding of how stats work requires either a bunch of tedious trial-and-error playing with stat+ items or reading spoilers.

2) It's often incredibly unclear what, exactly, will make my character better. I imagine that a healthy percentage of my stat selections are suboptimal because it's hard to know which stat is going to have the biggest effect. There are only three stats and they're all tangled up, so making the correct decision requires having all that spoilery knowledge from the first point.

Personally, I think it would be easier to remove stats and add more clear mechanical replacements than it would be to "clarify" stats in any meaningful way.

I agree with 1) and 2). In a way, it seems to hard to conceive a scenario where mis-stating killed you. But the same could be said about the difference of 2 AC, or +2 on a skill. (These numbers may be all too fine grained, but obviously they are meaningful.)

Personally, I think it would be easier to remove stats and add more clear mechanical replacements than it would be to "clarify" stats in any meaningful way.

Well, you're asked for it: hereby, I challenge you to come up with more a clear mechanic!

[XuaXua]

So far the most helpful (to the unspoiled) thing in this thread that anyone has come up with is when XuaXua recommended increasing the accuracy of the game manual documentation.

[Sandman25]

You are trying to make devs spend their time on changing something.

So? We're having a polite chat about mechanics on a website about elf games. I don't think there's any sense in talking about "burdens of proof," especially since I don't think quality of debate has ever been used as a criteria for changing the game.

Also, how are we going to "make" the devs do anything? They're not dummies, and we can't exactly hoodwink them.

I am glad you realize it. Some players want devs to change something. Other players reply that there should be a reason for the change. I am surprised it is not obvious to some.

[all before]

I think plenty of reasons have been given in this thread for changing/removing stats. That said, I understand dpeg's point that the ratio of work required to make the change vs. how much it improves the game makes this a lower priority than other changes.

[chequers]

Well, you're asked for it: hereby, I challenge you to come up with more a clear mechanic!

I think the clearer mechanics already exist:

Strength:
* Fighting skill
* Weapon skills
* Armour skill
* Shield skill
* Spellcasting skill
* Spell school skills

Dexterity:
* Fighting skill
* Weapon skills
* Armour skill
* Shield skill
* Dodging skill
* Stealth stkill

Intelligence
* Spellcasting skill
* Spell school skills

Here are the effects of attributes I can think of that aren't trivially covered by skills, and possible solutions:

All attributes:
* Artifact properties
** Can be replaced with other properties, eg Str -> one of Slay, or +Armour skill
* Racial/role differentiation
** Can be replaced with aptitudes or various other things mentioned by Archaeo's earlier post
* Chei bonus
** Replace with skill bonuses
* Str + Dex > random2(50) to proc auxiliary attacks
* Str + Dex for auxiliary attack damage

Strength:
* Escape constrictions vs enemy HD (entire formula could be redone easily enough)
* Monsters escaping your constriction
* Player constriction damage
* Player freeing from web

Dexterity:
* Player stab bonus damage
* Player water attack fumble chance
* Player avoiding blade/net traps
* Chance to break nets that hold the player when escaping them

Intelligence:
* ???

(will edit with corrections if people have them)

[Siegurt]

Here are the effects of attributes I can think of that aren't trivially covered by skills, and possible solutions:

All attributes:
* Artifact properties
** Can be replaced with other properties, eg Str -> one of Slay, or +Armour skill
* Racial differentiation
** Can be replaced with aptitudes or various other things mentioned by Archaeo's earlier post

As earlier discussed, aptitudes don't make adequate replacement for stats in terms of race differentation, as for "things mentioned by Archaeo's earlier post" << file not found >> Can you direct me to what you might possibly mean here?

* Role differentiation
** Replace with further itemisation differences, or aptitude modifiers(??), or other things mentioned by Archaeo's earlier post

Again "things mentioned by Archaeo's earlier post" << file not found >> Can you direct me to what you might possibly mean here?

[savageorange]

I don't object to the skillification idea, it seems good at a casual inspection (except for the problem of racial differentiation). But it seems hard to imagine that this has not been considered in the past, especially the concept of artefacts that boost skills. Can any devs comment on this?

[all before]

Why do some of these more obscure values for stats even need to exist? After a year of playing, I had no idea that str impacted constriction until chequers' above post.

[chequers]

I went through the codebase to try and find all other uses of stats, it's a bit difficult because str and int are both part of common C++ data types but I think I got most of them. Hands up who knew dexterity affected your chance to avoid two specific traps only?

[bcadren]

If you remove stats, you have to improve Aptitudes to compensate. If you do that; Demigods and Humans would literally be identical except the no gods thing.

Also it means all human size races get the same max EV. Makes everything more identical.

Make the game more boring, tbh.

It'd make more sense to make INT affect something other than spells...

[dpeg]

chequers: Thanks for the long list.

However, I am still not convinced. Trying to get out of my comfort zone, I attempt to imagine Crawl with these changes. Here are some observations (some already mentioned, sorry I'm so slow):

1. Skills max out, stats don't.
In other words, characters with 27 Spellcasting (say) would be equivalent in a stat-less system. Currently, 30 Int vs 10 Int would still matter. (It is debatable whether is matters enough, I'd hope so.)

2. Cheibriados vs Ashenzari
One gives stat boosts, the other boosts skill. I don't think that these have the same gameplay effect. Perhaps it is possible to emulate the stat boosts quite accurately with skills, but I am not convinced.

3. Race/role diversification without stats
Sure, we are using aptitudes as one of the main tools for race diversification. Using aptitudes for backgrounds seems out of the question.
Perhaps one way to look at it that two species with some identical aptitudes can still differ a bit through stats. In other words, stats as a secondary trait.


Str & constriction: this is on purpose. The idea was twofold: a naga/octopode should noticeably constrict differently whether it has Str 6 or Str 26. (This formula may have been nerfed in the mean time, but the idea is that the player has some impact on constriction usefulness.)

Regarding Int: I think it is actually very good that Int does nothing but help with casting spells. Yes, if you will never cast at all, that decision is made and you only got to choose between Str and Dex. On the other hand, most non-Troggies will want to cast at some point, and then we will have characters casting spells with wildly differing intelligence -- as it should be.

[Sandman25]

DEWz who puts all points into Int:

  Code:

Attack: Deep Elf Wizard vs. yak (4000 rounds) (2015/05/15/ 7:27:57)
Deep Elf Wizard: XL 1   Str 4   Int 22   Dex 13
Wielding: +0 triple sword, Skill: Long Blades
yak: HD 7   AC 4   EV 7

Long Blade | AvHitDam | MaxDam | Accuracy | AvDam | AvTime | AvSpeed | AvEffDam
         0 |      7.1 |     19 |      55% |   3.9 |   190  |  0.53 |      2.1
         1 |      6.8 |     19 |      56% |   3.8 |   185  |  0.54 |      2.1
         2 |      7.0 |     20 |      57% |   4.0 |   180  |  0.56 |      2.2
         3 |      7.2 |     20 |      60% |   4.3 |   175  |  0.57 |      2.5
         4 |      7.3 |     21 |      59% |   4.3 |   170  |  0.59 |      2.6
         5 |      7.4 |     22 |      60% |   4.5 |   165  |  0.61 |      2.7
         6 |      7.6 |     23 |      63% |   4.8 |   160  |  0.62 |      3.0
         7 |      7.7 |     24 |      61% |   4.8 |   155  |  0.65 |      3.1
         8 |      7.8 |     25 |      64% |   5.1 |   150  |  0.67 |      3.4
         9 |      8.0 |     24 |      65% |   5.3 |   145  |  0.69 |      3.6
        10 |      8.1 |     25 |      64% |   5.2 |   140  |  0.71 |      3.7
        11 |      8.2 |     26 |      65% |   5.4 |   135  |  0.74 |      4.0
        12 |      8.5 |     27 |      67% |   5.7 |   130  |  0.77 |      4.4
        13 |      8.8 |     28 |      68% |   6.0 |   125  |  0.80 |      4.8
        14 |      8.6 |     29 |      66% |   5.8 |   120  |  0.83 |      4.8
        15 |      8.9 |     29 |      68% |   6.1 |   115  |  0.87 |      5.3
        16 |      9.3 |     30 |      67% |   6.3 |   110  |  0.91 |      5.7
        17 |      9.4 |     31 |      69% |   6.5 |   105  |  0.95 |      6.2
        18 |      9.7 |     31 |      69% |   6.7 |   100  |  1.00 |      6.7
        19 |      9.6 |     31 |      69% |   6.6 |    95  |  1.05 |      7.0
        20 |     10.0 |     32 |      69% |   7.0 |    90  |  1.11 |      7.7
        21 |     10.2 |     33 |      70% |   7.2 |    85  |  1.18 |      8.5
        22 |     10.3 |     34 |      70% |   7.2 |    80  |  1.25 |      9.0
        23 |     10.5 |     36 |      70% |   7.4 |    75  |  1.33 |      9.9
        24 |     10.6 |     35 |      72% |   7.7 |    70  |  1.43 |     11.0
        25 |     10.8 |     37 |      71% |   7.6 |    70  |  1.43 |     10.9
        26 |     11.1 |     38 |      73% |   8.2 |    70  |  1.43 |     11.6
        27 |     11.3 |     38 |      74% |   8.4 |    70  |  1.43 |     12.0

DEBe who puts all points into Str

  Code:

Attack: Deep Elf Berserker vs. yak (4000 rounds) (2015/05/15/ 7:29:20)
Deep Elf Berserker: XL 1   Str 23   Int 11   Dex 14
Wielding: +0 triple sword, Skill: Long Blades
yak: HD 7   AC 4   EV 7

Long Blade | AvHitDam | MaxDam | Accuracy | AvDam | AvTime | AvSpeed | AvEffDam
         0 |      9.7 |     28 |      63% |   6.2 |   190  |  0.53 |      3.2
         1 |      9.7 |     28 |      65% |   6.3 |   185  |  0.54 |      3.4
         2 |      9.8 |     28 |      66% |   6.6 |   180  |  0.56 |      3.7
         3 |     10.3 |     31 |      65% |   6.7 |   175  |  0.57 |      3.8
         4 |      9.9 |     30 |      67% |   6.7 |   170  |  0.59 |      3.9
         5 |     10.1 |     31 |      68% |   6.9 |   165  |  0.61 |      4.2
         6 |     10.5 |     31 |      68% |   7.2 |   160  |  0.62 |      4.5
         7 |     10.6 |     35 |      71% |   7.5 |   155  |  0.64 |      4.9
         8 |     11.1 |     36 |      69% |   7.7 |   150  |  0.67 |      5.2
         9 |     11.5 |     34 |      71% |   8.3 |   145  |  0.69 |      5.7
        10 |     11.2 |     36 |      70% |   7.9 |   140  |  0.71 |      5.6
        11 |     11.8 |     36 |      70% |   8.3 |   135  |  0.74 |      6.2
        12 |     12.1 |     38 |      72% |   8.7 |   130  |  0.77 |      6.7
        13 |     11.9 |     40 |      71% |   8.5 |   125  |  0.80 |      6.8
        14 |     12.7 |     41 |      72% |   9.3 |   120  |  0.83 |      7.7
        15 |     12.7 |     43 |      72% |   9.2 |   115  |  0.87 |      8.0
        16 |     12.9 |     43 |      71% |   9.3 |   110  |  0.91 |      8.4
        17 |     13.3 |     43 |      74% |  10.0 |   105  |  0.95 |      9.5
        18 |     13.2 |     45 |      73% |   9.7 |   100  |  1.00 |      9.7
        19 |     13.1 |     46 |      74% |   9.8 |    95  |  1.05 |     10.3
        20 |     13.4 |     46 |      74% |   9.9 |    90  |  1.11 |     11.0
        21 |     14.0 |     47 |      75% |  10.6 |    85  |  1.18 |     12.4
        22 |     14.1 |     52 |      75% |  10.6 |    80  |  1.25 |     13.2
        23 |     14.6 |     49 |      74% |  10.9 |    75  |  1.34 |     14.6
        24 |     14.6 |     53 |      75% |  11.1 |    70  |  1.43 |     15.8
        25 |     15.1 |     52 |      76% |  11.6 |    70  |  1.43 |     16.5
        26 |     15.4 |     56 |      76% |  11.8 |    70  |  1.43 |     16.9
        27 |     15.5 |     53 |      76% |  11.9 |    70  |  1.43 |     17.0

[dpeg]

Sandman: Numbers are good, but you should evaluate at the end! Do you think the difference is marginal or relevant?

[Sandman25]

Sandman: Numbers are good, but you should evaluate at the end! Do you think the difference is marginal or relevant?

Sorry, I thought they are obvious to everyone.
41% difference in damage at Long Blades 27 (c.f. haste gives 50%). Another way to compare is DEBe with LBl 20 deals the same damage as DEWz at min delay (LBl 24), that's a lot of saved XP.

Also DEWz has double Int (22 vs 11). It means double spell power (before step down). With bonus points the difference becomes 20.

So XL 27 DEBe deals 40+% more damage with melee and has Int 17 instead of 37 (only 45% spell power before step down). (DE gets 1 Int naturally at every 4th level).

I wouldn't like stat removal, I enjoy playing every race more than once.

Edit. Also I tested the same DE with CPA and Armour 27/Dodging 27 to see EV penalty. DEBe has EV 18, DEWz has EV 0.

[dowan]

FR: No stats RC option. When turned on, stats aren't shown, and levelup stats are chosen at random and hidden. Rings of str, dex, and int are just called plain rings.

No actual change to the mechanics, but people who really hate stats don't have to see them.

I personally find the tension between choosing str (I can wear heavier armor and retain decent spell success and EV), dex (more EV, slight bonus damage/acc), and int (More powerful spells, better spell success) a very interesting choice, and I often sit for a minute or two trying to decide on most characters. And they definitely make a big difference, try a character with 10 str in plate vs a character with 25 str.

Furthermore, I just don't see what stats take away from the game. One keypress every 3 levels isn't really that hard, and if you don't want to think about it, just press D, it'll probably be fine.

[all before]

The only sensible argument I've seen for stats value is that they help differentiate races. If that's the case, why not remove stat rings and chosen stat increase on level up? Both of those make races less distinct and are basically redundant w/other, better aspects of the game (ac/ev/slay rings and skilling).

[Sandman25]

The only sensible argument I've seen for stats value is that they help differentiate races. If that's the case, why not remove stat rings and chosen stat increase on level up? Both of those make races less distinct and are basically redundant w/other, better aspects of the game (ac/ev/slay rings and skilling).

Both rings and stats on levelups can increase OR decrease difference between races and even different games with the same combo/god, it's all up to player and is very good for replayability/strategy. For example, I have already mentioned Int vs Dex choice for Enchanter.

[dowan]

Play a DgGl and choose STR for each levelup. You will end up with a hard hitting character who can wear heavy armor well and retain good EV, and decent spell success
Next, play a DgGl and choose int for each levelup. You will end up with a lighter armored character who can cast powerful spells with a fairly low XP investment.

You will end up with very different characters, even though you have the same race and background. Why do we want to lose this?

[all before]

The choice of stats made when leveling up are spoiler-y (it is not at all intuitive that increasing str can be good for casters or dex for melee brutes). Moreover, they replicate the choices already being made with skilling. Given two mechanics that basically do the same thing (skilling and choosing stat gains), why does the less intuitive one need to exist?

[Sandman25]

The choice of stats made when leveling up are spoiler-y (it is not at all intuitive that increasing str can be good for casters or dex for melee brutes). Moreover, they replicate the choices already being made with skilling. Given two mechanics that basically do the same thing (skilling and choosing stat gains), why does the less intuitive one need to exist?

Well, crawl is full of hidden/unintuitive mechanics but effect of str for casters and dex for melee brutes is not that spoilery because you can see the effect when you equip a ring of strength/dexterity.
Skilling is not equivalent to stat increase and this is good IMHO, we don't want characters with maxed skills be the same.

[dowan]

But... they don't do the same thing, at least, no more than firestorm does the same thing as a dagger (They both kill stuff).

I guess you could consider them somewhat spoilery, in the same way that it's spoilery that it's a good idea for a caster to train a weapon skill, or for a minotaur fighter to learn blink and repel missile.

What's really spoilery is the stepdown function on dex!

[all before]

we don't want characters with maxed skills be the same.

What happens to characters with maxed skills is pretty unimportant: at that point, they've won. For the part of the game that matters, both skilling and stat growth are about choosing how to allocate a limited resource. I think skilling does this in an admirably clear manner (with some exceptions like how dodging works with heavy armor), while stat growth is needlessly complicated (I have to go look info up and/or go into wizmode to know what I'm really getting with +1 str vs. dex vs. int).

[Sandman25]

After editing the post several times, here is final version of the post:

I will be happy with stats removal if it allows to get everything we need to play optimally without spoilers or using copy of current character in wizmode.

[bananaken]

One thing regarding stats that I really like:

I don't know about you guys, but it feels good to see FIFTY on your INT or STR or something absurd like that. It feels good to achieve big numbers as a result of good planning or even good chance. This is aided by the fact that these numbers are uncapped. Big numbers are exciting, specially when you know how disproportionate these numbers are in relation to "average" stats. Not to mention stuff like fifty INT can be actually useful to help you reach max spell power breakpoints with less spell enhancers.

Whatever you do, please preserve this effect, it's cool!

[duvessa]

...Those numbers ARE capped.
There'd still be AC/EV/SH.

[archaeo]

First, thanks chequers, for answering dpeg's question way better than I could've done.

Again "things mentioned by Archaeo's earlier post" << file not found >> Can you direct me to what you might possibly mean here?

The first one you ask about is racial differentiation, for which I suggested species mutations and other mechanics. I think most species would survive on their existing mechanical differentiation, so VS and Gh and the like would probably be fine. The only species that sticks out to me is Dg, as bcadren notes, and I figure you could solve that by giving them a neat suite of mutations that level up over time. Heck, you could even just make them more like a "choose your own mutation" DS-lite, where at the level-ups where you'd normally choose stats, you could choose between three mutation tracks. Maybe some kind of AC/Slay bonus, an Archmage/Wiz bonus, and an EV/Archery bonus?

Likewise you ask about role differentiation, and I think chequers was referring to my briefly mentioned idea that you could always give bonus aptitudes depending on the background. Maybe that DEWz gets a +1 to spellcasting, and the DEFi gets a +1 to fighting, etc.

@dpeg: the only observation of yours that wasn't responded to was the issue of Chei vs. Ash, but why would we muck that up? Can't Chei just give you pseudo-mutations that stack up the Dg bonuses I talked about above? Couldn't you just treat worshiping Chei as always being under the effect of Might/Brilliance/Agility, or some fractional level thereof depending on piety?

[dowan]

So, stats are too complicated, so we should fix it with a brand new set of special mutations, which essentially do what the current stats already do, and then make more special mutations for DGs that basically do what the current stats already do. Also, chei needs to use these special mutations too.

Now you've just renamed 'stats' into 'special mutations'. It seems like you're just set on removing stats for some reason, even if you have to invent an even more complicated system to replace them, all in the name of simplicity.

9 stat points (or 18 for dgs) allows you to have another layer of customization, which allows for character diversity, even within the same races and backgrounds. Racial stat bonuses allow certain races to peform better or worse at certain tasks, even at the same skill level.

And most importantly, I haven't seen this question answered: How is no stats crawl better than crawl with stats? I think a lot of examples have been given why crawl with stats is a good thing. Changes to a game shouldn't be made just to make the game different, they should make the game better.

[Sandman25]

Can you please describe how those pseudo mutations would work?
Also instead of "pick one of three stats to increase" player would get "pick one of three mutations to activate", right? Or are you going to remove that choice completely and make it automatic depending on species, background and probably god?

[archaeo]

So, stats are too complicated, so we should fix it with a brand new set of special mutations, which essentially do what the current stats already do, and then make more special mutations for DGs that basically do what the current stats already do. Also, chei needs to use these special mutations too.

Now you've just renamed 'stats' into 'special mutations'. It seems like you're just set on removing stats for some reason, even if you have to invent an even more complicated system to replace them, all in the name of simplicity.

I was actually going to bring this up in my response to dpeg that never happened (chequers did a better job than I could've): the issue with stat clarity is that, if you're going to make the mechanics more clear, you end up having to have a lot more individual mechanics. That's really the only benefit to having stats: that they're three little numbers that can sit in your HUD and do a bunch of things.

I think the benefit of the mutations is that they're not numbers, but easily recognizable things.

And most importantly, I haven't seen this question answered: How is no stats crawl better than crawl with stats? I think a lot of examples have been given why crawl with stats is a good thing. Changes to a game shouldn't be made just to make the game different, they should make the game better.

I feel like those of us who feel the game would be better without stats have answered this question several times in this thread, but to give it another go: stats are mysterious numbers that have a prominent place on the screen. Even experienced players don't usually know the full extent of how stats work. However, we're often asked to make choices about stats, and their mysteriousness often leads people to make suboptimal choices. It doesn't help that practically every species gains stats in a different way, so, for example, I might mistakenly choose Str when I'm offered a decision despite the fact that the species' natural stat growth means I'll have more than enough Str and should've picked Dex.

That's why the game would be better: if you did it right, all those choices would be clear, and it would be easier to understand the underlying mechanics of the game for people who don't bother with code diving or fsim.

Can you please describe how those pseudo mutations would work?
Also instead of "pick one of three stats to increase" player would get "pick one of three mutations to activate", right? Or are you going to remove that choice completely and make it automatic depending on species, background and probably god?

I openly and cheerfully admit that I'm way too much of a crawl layperson to ever pretend like I could actually describe a formula to do what these mutations could do. But yes, ideally, Dg would get to pick one of those three mutations to "level up" whenever they get a choice.

e: also, "are you going to" sounds like I would actually have something to do with making the change, but since I'm basically just a webtiles gadfly, all I can do is say "this seems like a good idea" in the hopes that a dev will agree.

[all before]

9 stat points (or 18 for dgs) allows you to have another layer of customization, which allows for character diversity, even within the same races and backgrounds. Racial stat bonuses allow certain races to peform better or worse at certain tasks, even at the same skill level.

This is the rub: does having two "layers" of customization really improve the game, or just add complexity for its own sake? I mean I suppose you could invest all your MiGl's stat gains in INT while training nothing but melee and defensive skills, and it would be "distinct" from a MiGl with the same skills, but why would you ever do that?

[Sandman25]

I openly and cheerfully admit that I'm way too much of a crawl layperson to ever pretend like I could actually describe a formula to do what these mutations could do. But yes, ideally, Dg would get to pick one of those three mutations to "level up" whenever they get a choice.

Oh, I expected something like perks from other games.
1) +10% damage for Fire Spells
2) +3 EV
3) CPA gives extra 5 AC
4) +100 MR
5) +50 Stealth

[archaeo]

Sandman, this is starting to get offtopic, but I figure that if you remove stats from Dg, you would want to retain the idea that they're just better, not a collection of neat mutations. Otherwise, you start to remove the distinction between Dg and DS, and I think that'd be a bad thing.

[Sar]

I wanted to say how much Crawl would benefit from Jinxed perk but then I realised that Xom exists!

[dowan]

9 stat points (or 18 for dgs) allows you to have another layer of customization, which allows for character diversity, even within the same races and backgrounds. Racial stat bonuses allow certain races to peform better or worse at certain tasks, even at the same skill level.

This is the rub: does having two "layers" of customization really improve the game, or just add complexity for its own sake? I mean I suppose you could invest all your MiGl's stat gains in INT while training nothing but melee and defensive skills, and it would be "distinct" from a MiGl with the same skills, but why would you ever do that?

Well, since there's no such thing as an int apt, my MiGl gets just as much benefit from 27 levels worth of XP put into the int stat as anyone else, even though all my spell schools have a -3 or -4 apt. On the other hand, I will get the same EV from adding all that to dex as anyone else, even though I have a +1 dodging apt. So maybe if I wanted to cast a smattering of low level utility spells, the XP I save on individual spell schools is worth the loss of the EV I would have gotten for putting those points in dex.

Now, of course it would be idiotic to pump int if I had no intention of ever casting spells, just like it would be pretty dumb to pump str on a DEFE who never planned to wear anything heavier than a robe or swing a weapon. But if I want my DEFE to wear that nice randart Ice dragon armour I might actually put some points into STR.

[dpeg]

I don't see how why we, the stats people, should prove that "stats are good". Nor do I think the anti-stats people have to prove "stats are useless". They aren't Crawl's best mechanic, and they do something in the game. We differ in how much stats bring to the table, and about the cost. I don't really like the "stats are spoilery" argument. (Of course, if someone helps improve the manual in this regard, I'd be very grateful.)

Altogether, I really see no reason to remove stats -- I am unable to see how stats are bad in a way that divinations, Hive, the dwarves, amulet of resist slowing, goliath beetles, or anything else we've cut over the years. I would like to improve them while keeping them. Currently I can think of two simple steps: Make stats coarser by divding them by two (except for the stat choices at levels 3,6,etc.); replace the random stat increases by either deterministic ones (fixed per species) or by player-chosen ones.

[bananaken]

...Those numbers ARE capped.
There'd still be AC/EV/SH.

If I get +30 strength from equipment on my 25 base strength character, I end up with 55 strength. As far as I can tell, this is not capped (maybe 99 or something). This is what I refer to as uncapped, is that the wrong way to describe it?

In any case, this aspect of raising stats is visually appealing to me, and I suspect most other casual players at least.. that's all I was trying to say.

[archaeo]

I don't really like the "stats are spoilery" argument. (Of course, if someone helps improve the manual in this regard, I'd be very grateful.)

I mean, for one thing, expecting players to read the manual is real 1980s game design, in my view. There's no reason the game shouldn't be able to teach you what it's doing, right there and then. At an absolute minimum, there should be a "Hit ? for help" option on stat level up decision prompts.

I'm not sure why "stats are spoilery" doesn't work as an argument, either. Sure, "bigger numbers are better" isn't hard to understand, but the game makes the player choose stats at relatively frequent intervals, and those choices are extremely murky to make optimally without spending a ton of time either a) code diving or b) sifting through inconsistent information on Tavern/IRC. In the most extreme cases, you're something like an MfIE or VSSk, where you have to balance AC, EV, and spellcasting to be successful. When you level up, what helps you most: Str, Int, or Dex? I don't really know, and I expect I'm not alone.

Altogether, I really see no reason to remove stats -- I am unable to see how stats are bad in a way that divinations, Hive, the dwarves, amulet of resist slowing, goliath beetles, or anything else we've cut over the years.

To put it succinctly, the fact that devs removed really obviously bad parts of the game doesn't mean they shouldn't remove the subtly bad parts of the game.

Currently I can think of two simple steps: Make stats coarser by divding them by two (except for the stat choices at levels 3,6,etc.); replace the random stat increases by either deterministic ones (fixed per species) or by player-chosen ones.

I think coarser stats are a good idea, but changing the randomness either makes stats pointless (since they grow deterministically and therefore could be replaced with things that aren't opaque numbers) or even more problematic (since now every level-up comes with this confusing choice).

[archaeo]

If I get +30 strength from equipment on my 25 base strength character, I end up with 55 strength. As far as I can tell, this is not capped (maybe 99 or something). This is what I refer to as uncapped, is that the wrong way to describe it?

IIRC, it's capped at 75. edited to add that it's 72 and I'm an idiot

Wait a minute, I thought you were just supposed to hit D every time.

Sure, that's how I usually play these characters. But if I was an MfIE, should I have picked Int instead, to boost my Cond. Shield and Ozo's Armour? As a VSSk, should I have gone for Str so I could upgrade my armour while still being able to cast my spells?

I imagine that in the vast majority of cases, there's a "correct" stat to upgrade to get the biggest benefit. I think crawl makes it hard to tell which one is right.

[Arrhythmia]

If I get +30 strength from equipment on my 25 base strength character, I end up with 55 strength. As far as I can tell, this is not capped (maybe 99 or something). This is what I refer to as uncapped, is that the wrong way to describe it?

IIRC, it's capped at 75.

72 (this number is the reverse of 27 you will notice)

[dpeg]

I don't really like the "stats are spoilery" argument. (Of course, if someone helps improve the manual in this regard, I'd be very grateful.)

I mean, for one thing, expecting players to read the manual is real 1980s game design, in my view.

Whatever floats your boat. Meanwhile, I am playing this game where I push buttons to move an @ across a 80x24 console screen, trying to turn alphanumeric symbols into full stops.

There's no reason the game shouldn't be able to teach you what it's doing, right there and then. At an absolute minimum, there should be a "Hit ? for help" option on stat level up decision prompts.

What in this world makes you so entitled? This is a free, open source game sitting in a tiny niche -- it's not even part of any video game market --, and there is no absolute minimum at all. If you feel so strongly about Str/Dex/Int, send us a patch or devise a help screen. (That's exactly what I did when I wanted to help Crawl and couldn't code.)

I'm not sure why "stats are spoilery" doesn't work as an argument, either. Sure, "bigger numbers are better" isn't hard to understand, but the game makes the player choose stats at relatively frequent intervals, and those choices are extremely murky to make optimally without spending a ton of time either a) code diving or b) sifting through inconsistent information on Tavern/IRC. In the most extreme cases, you're something like an MfIE or VSSk, where you have to balance AC, EV, and spellcasting to be successful. When you level up, what helps you most: Str, Int, or Dex? I don't really know, and I expect I'm not alone.

What I really don't like about your argument is how you construe this idea that people have to "sift through inconsistent information" or "code dive" in order to play the game. Yes, Crawl's information politics is medieval, there are good reasons for this, and if you want to play optimally, you have to do something like this (perhaps relying on learndb suffices, I wouldn't know). If you just want to play the game, including winning it, you don't have to. The game makes it roughly clear what Str/Dex/Int will do, and an observant player can actually figure out on her own. (Note: I expect players to be either literate or observant, if you're neither, perhaps play Angry Birds instead?)

To put it succinctly, the fact that devs removed really obviously bad parts of the game doesn't mean they shouldn't remove the subtly bad parts of the game.

You are telling me it's bad. I don't follow. Would you like to lead us to enlightenment?

PS: I am getting too angry over this, so I'll just make a note about coarser stats and call it a day.

[chequers]

Just to answer "why are stats bad?" myself:
Stats are a coarser, less flexible duplication of the skill system. (Most of the other functionality they have is exceptionally spoilery and should be clarified/removed regardless of this discussion.)

So, "duplication" of game systems is a value judgement. As someone upthread implied, weapons and (offensive) spells are a duplication too. There's a value judgement for each case as to whether the degree of duplication is problematic. In the case of daggers vs firestorm these are clearly qualitatively divergent. One is melee range, low skill, quiet while the other... isn't. So they occupy polar opposite niches in the killdudes system.

A recent example of problematic duplication was the removal of flight spell. There were too many other flight sources and there was no need for "cheap spell version". Flight spell was probably the first source of flight in game (I guess), but newer additions turned it vestigial.

So where do stats fit? I believe complexity is self-justifying in roguelikes (well moreso for nethack than dcss!) so multiple character skill systems is a good thing. If you want to skill a character in two directions though is training axes & pumping int any different than training axes & spellcasting? I think not.

Skill system is very clear. There are problems, like too many skills, weapon mindelay being spoilery and HP/XP apts being hidden, but it's overall a clear, simple design.

Compare to stats. Even the skill-duplicated effects of stats -- how many newbies would discover dex affects shields? Or that it affects bucklers more than large shields? Or that str makes you less vulnerable to constriction but not elephant knockback?

If stats had unique functionality, I would like them a lot more. How about armour skill only adds 1/22 bonus AC, and only str is considered for aevp mitigation? Or int controls spell power exclusively, and skill XP is for success? Or dex increases EV and we remove dodging skill? Right now they aren't unique, so I think they should die.