No-Brainer Attributes / Int for Combat; Str for Spellcasting

[XuaXua]

I think Int should be made a viable choice to enhance for non-spellcasters, outside of being able to defend from stat loss, and strength / dexterity viable choices for spellcasters outside of encumbrance limitations.

While there may be a lot of reliance on dexterity and strength, I feel Intelligence should play some factor in combat.

One idea might be to have INT enhance the ability to achieve a critical hit (to know WHERE the critical locations are to hit). Unfortunately, with the way the game is designed currently, this would greatly enhance stabbers who are, more often than not, spellcasters with high Int.

An idea with Dexterity enhancing spellcasting would be to alter the rate of speed with which a spell can be cast. Likely for flavor purposes, this is a set value, but I don't know for sure.

An idea with Strength enhancing spellcasting would be for it to enhance the resistance to miscasts. Unfortunately, miscasts are infrequent on a normal basis, so this isn't much of an alteration.

Are there other ideas to give the attributes added utility in areas where they currently aren't used?

[Wahaha]

I think int should help you cast spells.

[1010011010]

Regardless of whatever perks for int, str & dex for spell casting, skirmishing & stealth you may put in place; you will always have one stat trump the rest for specialised classes. if you made dex decrease the delay in spell casting to a noticeable degree then everyone would dump in dex. The not-so no-brainers with stat distribution arise in hybridisation, where strength may or may not be more important than intelligence.

[and into]

Actually, taking into consideration some of the recent changes, I don't think the stats are *that* unbalanced any more.

For certain character types your optimal choice may be pretty clear and unidirectional, but some degree of that is unavoidable unless you add a bunch of (further) obnoxious side effects for each stat that matter for all characters. I don't think that would be an improvement. It is fine that each stat should provide clear and meaningful (but not overpowering) bonuses in particular areas. Many characters will have at least two stats they care about substantially, even if one is usually the better choice, and often they still won't be completely uncaring about the third.

Strength could matter a bit more to damage in melee, probably with a step down after a certain point (like Dex with EV) so you can't get insane 300% multipliers on TrMo of Chei with +strength gear or whatever. Strength already has a pretty good effect on ranged weapon damage though actually, for the most part. Tweaking the numbers a bit with melee for strength and also, I guess, the bonus "dexterity weighted" weapons get from high dex, is probably warranted but beyond that I don't think much needs to change to be honest.

[drywall]

Having to juggle multiple stats to fine-tweak every action sounds awful if required and pointless if it uses an obtuse ranged weapon-like formula (it will just make stats feel irrelevant). Stat breakpoints and magic numbers are also pretty bad (if I can just get 1 more int I can cast fire storm at min delay when hasted...)

Also "no-brainer attributes" is kind of dishonest because with almost all characters you will want to consider str or dex. I almost never take int unless I'm using hexes or high level conjurations, or if my int is really bad (< 12ish) and I am not with trog.

[Amnesiac]

I like the way stats work now. It's as straightforward as those backgrounds that don't hesitate for what stat they need the most. Int for mages and not much point in str(at least for me) and fighters usually don't need Int and if it's Be then they don't have any use for Int at all. But for hybrids, stabbers and nonstandard fighters(not MiFi/Be and such) it's not so simple. Should I raise dex to increase my dodging and stabbing(and a bit of dmg for low str weight weapons, I guess?) or shoud I increase Int for some blasty spells backup or more spell power on enchantments. Should I raise dex to increase EV of my ogre or should I raise Str to get the maximum damage from slings, rocks and maces and even a bit of shields increase and evp decrease? I'm not sure that my points are valid, but I get such ideas about stats.

If my points about choosing str over dex are not justified as much as I think, that only means that str needs to be further boosted, because otherwise it's kinda useless.

[galehar]

why? You keep missing this step.

It's good that each stat has a well defined range of effect and that each effect is affected by a well defined and restricted number of factors. You keep suggesting that "this should be affected monster size and strength and weapon weight" or that "this skill should affect speed and range and fumble chance". This isn't a simulator. Multiplying the number of factors in a formula just make each one less important and everything more complicated.

[Daggtex]

Well... I /do/ have an excel spreadsheet that I've been tweaking around for a while to balance str, dex, and int for a roguelike that I hope to make as my comp sci project this year. My goal was to make it so that each of the 3 classes, ranger, warrior, and mage, have a main stat, but still make it always useful to invest in other stats, and I've been pretty successful so far (meaning I'm pretty sure I balanced range and melee, and have a good idea of how to balance magic). However, it does require a little smidgen of an inventory/item change along with the things the stats affect. If that's not too out there, I would be glad to explain some of my methods, maybe even post my current formulas.

[jejorda2]

Having the stats' effects not overlap means that the decision between the stats is also a decision between better spell casting (INT), better stealth/evasion (DEX) and better armour (STR). You have to decide which of those to give up when you make the stat decision. When the effects do overlap, like melee weapon damage, it means that the decision is less clear: you don't give up meaningful damage increase by choosing DEX over STR or STR over DEX, but you do by choosing INT. Making the stats overlap makes it clearer what a player gains and gives up by choosing a stat. Making the stats overlap more makes the decisions matter less, which makes them less fun.

We ought to remove the str or the dex component to damage from each weapon, or make it so only the higher one matters.

[Galefury]

Well... I /do/ have an excel spreadsheet that I've been tweaking around for a while to balance str, dex, and int for a roguelike that I hope to make as my comp sci project this year. My goal was to make it so that each of the 3 classes, ranger, warrior, and mage, have a main stat, but still make it always useful to invest in other stats, and I've been pretty successful so far (meaning I'm pretty sure I balanced range and melee, and have a good idea of how to balance magic). However, it does require a little smidgen of an inventory/item change along with the things the stats affect. If that's not too out there, I would be glad to explain some of my methods, maybe even post my current formulas.

It's probably not relevant to Crawl, I think Crawl's stat system is currently pretty okay. But it sounds very interesting and I'd like to read about it.

[and into]

Well... I /do/ have an excel spreadsheet that I've been tweaking around for a while to balance str, dex, and int for a roguelike that I hope to make as my comp sci project this year. My goal was to make it so that each of the 3 classes, ranger, warrior, and mage, have a main stat, but still make it always useful to invest in other stats, and I've been pretty successful so far (meaning I'm pretty sure I balanced range and melee, and have a good idea of how to balance magic). However, it does require a little smidgen of an inventory/item change along with the things the stats affect. If that's not too out there, I would be glad to explain some of my methods, maybe even post my current formulas.

Seconding Galefury. I'm very interested in game design, personally (as you might imagine) and I'd be interested to see your take on these things. But I agree with him that it is probably not not appropriate for this forum unfortunately (since the game in "Game Design Discussion" is limited specifically to Crawl). If you have an idea for a very specific reform to stats, for Crawl, that would be appropriate to discuss in this thread. But a major overhaul that requires changes in multiple areas simultaneously is the kind of thing that will only be entertained if there is a very major problem it is addressing. I think the consensus is that the effect of stats in Crawl could perhaps use a bit more tinkering, but don't need drastic changes. (It is part of the intentional design goal of Crawl that stats have a meaningful impact, yet not so huge an impact as in many other games, because Crawl tries to make it possible to branch out into multiple different directions over the course of a game. So it is a rather delicate balancing act.)

For discussing your work in progress, CYC might not quite be what you'd want, though I bet if you post a thread requesting serious discussion only, that would be *mostly* adhered to I think. A mod could give better advice here.

[Galefury]

I'm pretty sure talking about it here would be fine.

[galehar]

Yes, GDD isn't crawl only, it's fine to discuss other games or general game design. Although do open a new thread.

[Siegurt]

We ought to remove the str or the dex component to damage from each weapon, or make it so only the higher one matters.

Just a side note here: The way the stat bonus stuff for melee is calculated, raising str is ALWAYS better than raising dex for increased melee damage, and raising dex is ALWAYS better than raising str for increased melee to-hit chance. The weapon strength weight thing merely determines by how much a str is better than dex for damage, and by how much dex is better than str for to-hit.

That seems like a clear cut division and doesn't seem to merit a change "Raise strength for more melee damage per hit" and "Raise dex for better melee to-hit chances" seems like a clear enough choice even inside the realm of melee weapons.

[dpeg]

Siegurt: I agree. It's not Dex's fault that accuracy matters little.

[Sandman25]

Just a side note here: The way the stat bonus stuff for melee is calculated, raising str is ALWAYS better than raising dex for increased melee damage, and raising dex is ALWAYS better than raising str for increased melee to-hit chance. The weapon strength weight thing merely determines by how much a str is better than dex for damage, and by how much dex is better than str for to-hit.
.

I couldn't believe it and looked into the code
int melee_attack::calc_stat_to_dam_base()
{
const int weight = weapon ? 10 - weapon_str_weight(*weapon) : 6;
return you.strength() + (you.dex() - you.strength()) * weight / 20;
}

Dagger has weapon_str_weight = 1 so it becomes
str + (dex - str) * 9/20=str + 0.45dex - 0.45str=0.55 str + 0.45 dex

Executioner's Axe has weapon_str_weight = 9 so it becomes
str + (dex - str) * 1/20=str + 0.05dex - 0.05str=0.95 str + 0.05 dex

Yes, you are completely correct.

[Siegurt]

Yes another way to look at it is that damage is modified by the average of Str&dex, plus an amount of Str based on the str weight of the weapon.

So it's always .5 dex + .5 Str + ( 0.? * Str) where ? is determined by str weight of the weapon. So Str is always *at least slightly better than* dex for weapon damage, theoretically if there was a 0 Str weight weapon, Str would be equal to dex, but there is no such weapon.

[jejorda2]

Oh, that's good then. I mistakenly thought STR and DEX has roughly equal contributions to melee damage, varying only by the strength weight.

Never mind.

Spoiler: show

Mod edit: Not in GDD you don't

[dck]

I think currently stats are pretty good.
Int isn't something you normally want that much, but it's good if you want spells that do important things better with more spellpower, like killing people or going through MR. Dex is borderline overpowered because of EV being borderline overpowered, but it's still mostly fine. Str is still by far the weakest of all and almost never worth rising but at least it does something now, sometimes, which is kinda alright I guess.

[Implojin]

raising str is ALWAYS better than raising dex for increased melee damage

Melee combat in DCSS doesn't really work like this. Raising dex on levelups and getting your damage from skill contribution gives you EV and damage. Raising str on levelups only gives you (a minor contribution to) damage. EV is *very* strong in this game, to the point where even after the EVP->encumbrance rating change I can only think of a few edge cases where I would ever want to level strength, and most of those involve wanting carry weight on a particularly low str combo to mitigate interface annoyance.

[Mankeli]

raising str is ALWAYS better than raising dex for increased melee damage

Melee combat in DCSS doesn't really work like this. Raising dex on levelups and getting your damage from skill contribution gives you EV and damage. Raising str on levelups only gives you (a minor contribution to) damage. EV is *very* strong in this game, to the point where even after the EVP->encumbrance rating change I can only think of a few edge cases where I would ever want to level strength, and most of those involve wanting carry weight on a particularly low str combo to mitigate interface annoyance.

You should really learn how to properly quote others. Did you even read the whole thread? Siegurt definitely did not say that you should always increase strenght because it increases the damage in melee. The whole point of his post was to compare to-hit and to-damage boni gained from dex and str not to write an universal guide on what stats to pick. Obviously the stats, especially dex, have other uses too.

[Implojin]

Did you even read the whole post? Siegurt definitely did not say that you should always increase strenght because it increases the damage in melee. The whole point of his post was to compare to-hit and to-damage boni gained from dex and str not to write an universal guide on what stats to pick. Obviously the stats, especially dex, have other uses too.

Of course I did. Did you? I never claimed that the intention of his post was to write any kind of universal guide. I am attempting to point out the inadequacy of comparing the two stats based on their contribution to raw melee damage, as EV is the more important factor to gameplay by a large enough margin as to make stat choice on levelup often an uninteresting choice. I don't see that your hostility is warranted.

[Siegurt]

raising str is ALWAYS better than raising dex for increased melee damage

Melee combat in DCSS doesn't really work like this. Raising dex on levelups and getting your damage from skill contribution gives you EV and damage. Raising str on levelups only gives you (a minor contribution to) damage. EV is *very* strong in this game, to the point where even after the EVP->encumbrance rating change I can only think of a few edge cases where I would ever want to level strength, and most of those involve wanting carry weight on a particularly low str combo to mitigate interface annoyance.

I didn't say that raising strength to increase your melee damage was useful or effective, only that Str was more effective *at doing that one thing* than Dex was, because someone had earlier implied that they were roughly equivalent, at doing that one particular thing.

Raising strength to inflict more melee damage is a bit like taking Oka for the weapon gifts, you can do so, but it's not really the *point*. Each point of Str is worth between 1.45% and 2.5% more damage (Depending on weapon used) Which is way less than a point of weapon skill or even fighting.

[siprus]

I thought stat point main function is to differentiate between, stealthy, casting and heavy armor using builds. I'm not sure if any additional function is needed.

[dck]

Eh, honestly I thought about making a similar remark to Implojin's about this chatter about str and dex effect on melee damage because (even after being buffed recently) it's still tremendously irrelevant compared to anything else stats do and I think it's required to make clear you were theorycrafting about minimal percentages of extra damage, going back to reality and making a small footnote about how it all doesn't in any way apply to the way stat training should ever be done in a real game you're playing to win. In fact I find it a disservice to other forum users not to do so, and a pretty serious one at that.
Why? Because there have been 592 views of this thread at the time posting and there'll be a couple more before I'm done writing this up. Not every single of those readers is going to be experienced enough to understand that this sudden torrent of formulas and numbers and talk about how str gives you the MAX damage is pointless and should be ignored if you play with the intention of winning. New players are often misled by trying to see how the game works "under the hood" and stop paying attention to much more crucial aspects of the game in order to learn about these little things that to their understanding gives them the edge they need to win.
So I think posting things that can directly impact their development as players that badly is pretty serious and should be frowned upon.

[Kate]

Uh, talking about how the game works "under the hood" in the game design forum is not something that should be "frowned upon". Certainly it would be bad if someone were to post a bunch of unhelpful formulae in an advice thread but I'm pretty sure this is not one of those.

(Not that it sounds like anything particularly useful is coming from talking about the formulae here either, in this case).

[Daggtex]

For those interested, I started a new thread with the balancing I was doing.