Player Stats, Skills, and Weapon Effectiveness

[dolphin]

Player stats barely affect weapon effectiveness in Crawl, which certainly non-obvious when I began Crawling, and still does seem non-obvious to me.

I guess I'm thinking of 3 related problems.

• The dominant boost to Weapon Effectiveness (W.E.) is non-obvious (Weapon Skill -> Delay reduction).
• There are many things that affect W.E. negligibly (Fighting Skill).
• STR and DEX affect W.E. negligibly, which is also not intuitive.

Therefore, I suggest that STR and DEX become greater factors in W.E., and reducing the dominance of the weapon skill. The easiest way to do that might be to give Weapon skill diminishing returns in delay reduction, and make STR and DEX factors in weapon delay (not independent of weapon skill, but rather enhancers of weapon skill, so that a skill-less player can't use short swords at min delay just because of high DEX).

Another way to make STR and DEX important to W.E. might be to give weapons STR and DEX requirements. Not binary you can wield it or you can't, but instead like armor has a STR requirement for being without additional penalty.

Whatever negligible bonuses are left over after Weapon Effectiveness has been changed should be removed or buffed (Fighting, Weapons skill (past min. delay), Str/Dex Weight, EW1, etc). Players making choices that significantly affect the game is one of the appeals of a game. Negligible bonuses, particularly where powerfully significant bonuses are available, do not involve player choice, but player whimsy or coincidence.

STR in Crawl is currently most like Constitution in other games; it lets you wear heavy armor and carry more stuff without getting tired. If STR is unchanged or the negligible damage component removed, then I would rename Strength to Constitution.

[galehar]

There have been some discussion about that on the wiki. I think as a general rule, we should aim to have fewer parameters with stronger effect for damage and accuracy. Weapon delay is ok IMO.
I like the idea of stats having a strong effect on accuracy and none on damage (and the opposite for skill). It could help us differentiate hybrids and melee. To use the better weapons effectively in each category, you would need to invest in dex and/or str. If you put all in int, then you'd have to use smaller weapons.
We can also change the strength weighting of some weapons to get at least one dex and one str based in each category, to avoid the reduced variety minmay has talked about.

Also, I think we currently lack the tools to balance fighting. We'd better start by improving fsim and/or allowing player in the arena. This would make it easier to balance any change to the fighting code.

[RFHolloway]

It may be helpful to consider the statistic

Players killed by monster/monsters killed by player

from a design point. This gives a measure of the relative threat of the monster. You have the stats from the morgue files, I don't know how hard parsing it all would be though.

[galehar]

It may be helpful to consider the statistic

Players killed by monster/monsters killed by player

We already use those stats to balance monsters and uniques. I don't see how they can be of any help for balancing changes to the fighting code.

[RFHolloway]

If you are already monitoring that, then unexpected changes to the profile in beta/arena would be your indication - perhaps viewed by player combo.