Tweet
v4 is about six weeks old now and a few people have given some detailed feedback. It's clear that it already plays very differently from V because of the new affix system for ego items, but also that there is a lot of fine-tuning still to do on things like item naming, squelching/ID and damage bonuses. This latter issue made me want to start this thread: I'm not intending to "fix" the issue of damage bonuses properly until quite a lot of other related things have changed (though of course they'll be tweaked in the meantime), and the same applies to a number of other affix-related wrinkles. So I thought it might be helpful to set out my thoughts about the longer-term development of v4, so people can tell me "look, X is really important, please prioritise this more" and so on.
Disclaimer: I cannot speak for other members of the devteam, though I hope they will contribute their own thoughts/intentions to this thread. Each of us has RL to deal with and will contribute stuff as and when we can - when we're not getting 3.4 ready for release!
So, while other devs are more likely to make changes to dungeons, monsters, food, detection, classes, races, etc., my focus is now on combat. The importance of different values of +dam will change, hence my postponing the final fixing of item generation until those changes are known.
TL;DR: I want to move towards O-combat, where dice are more important than plusses, but with intuitive (i.e. linear) scales (and other changes).
The long version: before making major changes to combat, I'm planning to rewrite the way effects are handled by the game. This will have a number of other impacts (e.g. on spells and devices and message handling), but it will also enable far greater flexibility in the handling of combat. To take one example, at the moment monsters have a lot of "on-hit" effects (which are called procs in some circles): when they hit you in melee, things happen - you get confused, or stunned, or poisoned, etc. By contrast, only two procs are available to players (earthquakes and monster confusion), and they're both handled by special-case code. The new effects framework will allow easy definition of procs in text files, and they can be made available to monsters and/or the player (via items, spells, race or class abilities, etc.).
So, once effects are done, what's actually wrong with combat? Well, threads passim, including very recently, show a consensus opinion that the current model where warriors use daggers because of the number of blows is flawed. There have also been many prior threads on other issues: ranged combat multipliers, asymmetric LOS, hockey-sticks, pillar dancing and so on. Trying to break these things down gives a starting list of:
- the effects of STR and DEX on combat are identical for every weapon, race and class
- plusses to damage are inconsistent in their effect: more powerful on light weapons (more blows) but heavy launchers (higher mult)
- there's no distinction between on-weapon and off-weapon damage boosts
- there's no distinction between weapon weight and weapon speed
- there's no distinction between avoiding damage and withstanding it
- to-hit has too little impact on the number of hits, and none on their quality
- armour penalties are pretty meaningless, and the same for all races and classes
- the critical hit system uses only one factor (weapon weight) and makes little difference to combat
- extra whole blows are very powerful on high-damage weapons
- thrown weapons aren't anywhere near viable
- line of sight is asymmetric
- ranges are too large (especially when we consider Shockbolt's tiles and handhelds)
- spellcasting monsters can cast indefinitely
- immunities negate damage from some physical attacks
There are probably other issues too, which we'll remember or discover as we go. My intention, subject to the views of other devs, is to start at the 'centre' and work outwards. So starting with melee, to re-examine how often you swing, how often hits land and how much damage they do. This means considering all of
skills (i.e. race/class/level)
stats (i.e. STR and DEX)
energy/speed
weapon weight and/or weapon speed
plusses to-hit
armour inhibiting swings
evasion/dodging (also affected by armour)
armour absorbing or deflecting damage
base damage (dice and sides)
plusses to-dam
damage mods (brands and slays)
criticals
That covers about two-thirds of the above list. The rest are specific to ranged combat, or to effects (spells/resists), and can be dealt with separately. I guess the next step is to figure out exactly what, if anything, should be changed for each of the above aspects of melee (for example, do we want a dodging skill which is affected by race and class, or should evasion be purely a function of what you're wearing?). You can see that a lot of those aspects are or could be affected by items - so hopefully it's a little clearer why I'm not focusing obsessively on balancing affixes until this is done.
(But I'm happy to merge a pull request to sort out item naming though ;-))
Disclaimer: I cannot speak for other members of the devteam, though I hope they will contribute their own thoughts/intentions to this thread. Each of us has RL to deal with and will contribute stuff as and when we can - when we're not getting 3.4 ready for release!
So, while other devs are more likely to make changes to dungeons, monsters, food, detection, classes, races, etc., my focus is now on combat. The importance of different values of +dam will change, hence my postponing the final fixing of item generation until those changes are known.
TL;DR: I want to move towards O-combat, where dice are more important than plusses, but with intuitive (i.e. linear) scales (and other changes).
The long version: before making major changes to combat, I'm planning to rewrite the way effects are handled by the game. This will have a number of other impacts (e.g. on spells and devices and message handling), but it will also enable far greater flexibility in the handling of combat. To take one example, at the moment monsters have a lot of "on-hit" effects (which are called procs in some circles): when they hit you in melee, things happen - you get confused, or stunned, or poisoned, etc. By contrast, only two procs are available to players (earthquakes and monster confusion), and they're both handled by special-case code. The new effects framework will allow easy definition of procs in text files, and they can be made available to monsters and/or the player (via items, spells, race or class abilities, etc.).
So, once effects are done, what's actually wrong with combat? Well, threads passim, including very recently, show a consensus opinion that the current model where warriors use daggers because of the number of blows is flawed. There have also been many prior threads on other issues: ranged combat multipliers, asymmetric LOS, hockey-sticks, pillar dancing and so on. Trying to break these things down gives a starting list of:
- the effects of STR and DEX on combat are identical for every weapon, race and class
- plusses to damage are inconsistent in their effect: more powerful on light weapons (more blows) but heavy launchers (higher mult)
- there's no distinction between on-weapon and off-weapon damage boosts
- there's no distinction between weapon weight and weapon speed
- there's no distinction between avoiding damage and withstanding it
- to-hit has too little impact on the number of hits, and none on their quality
- armour penalties are pretty meaningless, and the same for all races and classes
- the critical hit system uses only one factor (weapon weight) and makes little difference to combat
- extra whole blows are very powerful on high-damage weapons
- thrown weapons aren't anywhere near viable
- line of sight is asymmetric
- ranges are too large (especially when we consider Shockbolt's tiles and handhelds)
- spellcasting monsters can cast indefinitely
- immunities negate damage from some physical attacks
There are probably other issues too, which we'll remember or discover as we go. My intention, subject to the views of other devs, is to start at the 'centre' and work outwards. So starting with melee, to re-examine how often you swing, how often hits land and how much damage they do. This means considering all of
skills (i.e. race/class/level)
stats (i.e. STR and DEX)
energy/speed
weapon weight and/or weapon speed
plusses to-hit
armour inhibiting swings
evasion/dodging (also affected by armour)
armour absorbing or deflecting damage
base damage (dice and sides)
plusses to-dam
damage mods (brands and slays)
criticals
That covers about two-thirds of the above list. The rest are specific to ranged combat, or to effects (spells/resists), and can be dealt with separately. I guess the next step is to figure out exactly what, if anything, should be changed for each of the above aspects of melee (for example, do we want a dodging skill which is affected by race and class, or should evasion be purely a function of what you're wearing?). You can see that a lot of those aspects are or could be affected by items - so hopefully it's a little clearer why I'm not focusing obsessively on balancing affixes until this is done.
(But I'm happy to merge a pull request to sort out item naming though ;-))
Comment