Tuesday, April 10, 2012

Update on balance issues in ADOM II 0.2.3 (and much more)

Hi everyone! First of all a short explanation about the rather silent time after the release of ADOM II 0.2.3: My wife and I have been on holiday for 17 days (in South Africa) and leisure was a top priority during that time ;-) ). Nonetheless I managed to work quite a bit on ADOM II 0.2.4 and since yesterday also caught up on the discussion about various balancing issues.

First a short progress report:
  • The "King of the Orcs" story arc is 98% finished - some small details still are missing. There is a lot more detail I could add but I'll refrain from that (for now) in order to "release early, release often". My next major priority is implementing the missing subsystems as outlined in the original roadmap (which means: missile combat, arcane spell casting and clerical spell casting as major subsystems as well as making all skills useful and adding lots of monsters regarding various minor subsystems influenced by that). Over time the KotO story arc will get a lot more cool details though ;-)
  • Corruptions are back in the game (all from ADOM and it's trivial to add more over time - the idea here being that in future games you won't be guaranteed a fixed set of corruptions in random order but rather a fixed subset - to introduce even more chaos in that ;-) ). Corruption is (no yet) an automatic feature but can occur due to special monsters and might become a latent feature due to the the actions of the PC in existing story arcs.
  • All the army stuff now works... including being able to fight armies or experience "sieges". (Although the term right now is a bit grand... it's just armies storming settlements to slaughter everything).
Concerning the (in my opinion) most critical bug in ADOM II 0.2.3: The game hanging should be over. (see this issue and this one for the original). The various very helpful stack traces trivialized fixing it... it was a very stupid oversight on my part. BTW: Initially I assumed that the hanging game was caused by memory consumption... I have introduced a memory display in ADOM II 0.2.4 and it's surprising: Sometimes a single move causes 5-10 MB of memory to be shuffled around. This does not seem to have tangible performance effects (although there probably are some) and I'll investigate that at a later point. You will see for yourself once ADOM II 0.2.4 is released.

The biggest issue that concerns me right now are various balancing issues addressed in the forums in post like this one (dual-wielding staves for incredible DV bonuses), this (again double-wielding and unhittable DVs), this (critical hits from large monsters, stat drain and unreachable dungeons), this (resting taking too long, too many wilderness encounters and others), this (stat draining and attribute training), this (hunger and satiation as del as frequency of wilderness encounters) and most prominently this post (size issues related to damage and PV issues).

For a short moment I panicked a little bit as there are quite a lot of viable balance issues raised. But then I remembered that ADOM 0.6.x and ADOM 0.7.x (as well as earlier versions) had these problems, too, and they were fixed quite well (IMHO). So back to work and fixing that in ADOM II - the playable game still is young, the engine is new in many respects and balance issues were to be expected ;-) Here's my current line of thoughts:
  • I really need to do something about the immense amount of damage dealt by large, huge (and even larger) beings. This is a typically case of realism destroying game fun: Larger beings wield larger weapons which cause more base dice of damage (larger armor also grants more PV BTW). Now larger beings usually also are a lot stronger than smaller beings. If you take e.g. a fire giant you have a monster with a strength around 46 at least (+18 fixed damage bonus minimum) wielding e.g. a great sword (which should do around 12d15+3 points of base damage) - for a base total of 12d15+21 or so. The simple fixes are obvious: I need to tone down the increase in base damage for weapons - somewhat more is ok but these increases probably are way too much. Then we have the question of fixed damage bonuses - maybe they shouldn't be fixed anymore? (so that e.g. 12d15+21 actually becomes 12d15+d21)? I'm not too sure about that though. As for damage... critical multipliers probably really only should kick in after PV has been deducted to reduce the deadliness of hits from huge beings. I probably also will reduce critical hit probabilities for monsters in order to protect the PC a bit more. While this fudges the generic game rules in favor of the PC it should be good for game fun.
  • ADOM II 0.2.4 will finally introduce a reasonable system for fighting with more than one weapon (at least I hope the system will be reasonable). It should take away many of the overpowered effects.
  • As for stat draining: I'm considering an approach where drained stats have a certain chance for recovery after a longer amount of time (e.g. something like a (20 + level + willpower + toughness + concentration skill score)% chance to recover a drained stat point after 500+d500 turns or so). Only if that doesn't work the stat will be lost forever. Additionally I over time might introduce some items (and other means) to recover lost stat points (e.g. in temples).
  • Some other issues are much easier to fix (e.g. tuning down the likelihood of wilderness encounters).
  • One other major thing I reimplemented for ADOM II 0.2.4 is the whole hunger and satiation system... I was pretty unhappy with how it worked and now have desolved actual time passed from hunger effects. Hunger effects now are more related to the number of turns past than actual game time which again is not as realistic but seems to be a much better approach for game fun.
So there you go: I take these balance issues very seriously but I'm as seriously sure that we'll get them fixed pretty quickly. We are still in the early ADOM II release cycle so some experimentation with new approaches is to be expected ;-)

Another final thing that has bugged me: How many people do think that dungeons are too big? I thought that the new balance was much better than the old one but not everyone shares that opinion ;-) At some point I definitely am going to introduce a configurable setup for the game (size of the world, frequency of dungeons, size of dungeons, etc.) but that's still something quite a bit in the future.

Comments, ideas and discussions welcome. Currently I fiddle around with exactly these details. ADOM II 0.2.4 definitely will be released within the next 7 to 10 days to finally get more into "release early, release often".


  1. Well, roguelikes are about a single character killing hundreds and hundreds of monsters, so I think rules have to be fudged in favor of the PC in one way or another. Of course, the PC has intelligence which is very important, but in a world of troll brutes able to do mighty blows that may not be enough :)

    Glad to see that you are addressing all the main issues (crashes, unreachable dungeons and balance) raised on the forums, and looking forward to 0.2.4.

  2. I'd say that if the drained stat effect becomes temporary then the stat drain should be stronger, I mean, several points of the stat.

  3. A Large/Huge creature should do very high damage and be really tough. I think the damage they do makes sense however it does hurt gameplay as it stands. No one wants to be 1-hit'd.
    If you change the Crits to only apply once PV has reduced damage that gives armour more protection then it has in the past. I think it would help with the issue a little. Perhaps toughness can limit the amount of damage you can recieve in a single hit, the higher your toughness the less that cap could be?

    Having 12d15+d21 and similar is very messy and not an ellegant solution at all, I also very much dislike the idea of 'padding' the PC by artificially weakening the monsters.

    It would be interesting if you could collect stats from us during gameplay of damage delt by creatures/weapon criticles and so on. With a large dataset like that it would surely help pinpointing balance issues. Save it all to a compressed file and we can send it to a stats collecter.

  4. As for damage... critical multipliers probably really only should kick in after PV has been deducted to reduce the deadliness of hits from huge beings.

    Eh... I don't like that one. I'm not arguing that it isn't a legitimate problem, but in the game I just won (I think - game freezes when I enter Terinyo :() the only way I was able to harm Fire Giants at all was with critical hits.

    Now, I was wielding a long sword, which is clearly not the optimal choice for giant-slaying. But I could kill the fire giants eventually. If the crit mult didn't kick in until after PV reduction, I never would've killed the fire giants at all.

    How about this instead - logically, if the fire giant is wielding a sword as big as you are, how does it score critical hits at all? So make the critical hit math (%, multiplier, or both) size dependent. This will make giants somewhat less of a grind (because you'll do them lots of damage with the bonus critical hits you score) and also less likely to splat you in single hits, but still able to do heaps of damage such as will threaten the life of a high-level character.

    1. I'd never thought about what a critical hit represents, but if it's striking a major organ or some such, it ought to be much easier for a relatively smaller fighter.

      Would you consider adding standing/kneeling/prone stances to the game? DF has them. While they mostly come about due to being wounded, hitting the dirt gives significant protection from missile attacks (and perhaps magic as well, although that's not implemented in DF)

      So I could see a huge clumsy giant swing a sword so big the best option for killing your PC would be squishing him with the flat of the blade, while your PC darts around, slashing at his achilles, or striking his ankle bones if wielding blunt. Since the tendons and bones on the giant would be the size of small trees, it should be no problem to strike exactly where you want every single time.

      After enough damage, the giant would fall, which would give your PC a to-hit and damage bonus, due to less mobility and the fact that eyes, brain, heart, lungs etc are now strikeable.

      As long as the damage/to-hit chance of the giant aren't too high for an appropriately leveled character, it could make each battle an awesome, epic fight. And make attacking two or more giants at once a real tactical challenge.

    2. Just read down, noticed Qui suggested exactly what I just said in his first paragraph :)

      Just to be clear, the extra bit I'm suggesting, if it's reasonable, a standing/prone system, with certain DV penalties in melee and bonuses in missile, along with an implementation that makes large monsters fall after a certain threshold (25% damage) to represent all your strikes hitting their legs/knees/ankles/feet, depending on how relatively massive the monster is, which would make the fight high danger, high strategy (which can be fun, as long as it's winnable) for the early going, while lightening up after a while, but still giving you that satisfying feeling that your PC is mauling an actual Giant/Dragon/Demon whatever. Think Shadow of the Colossus.

      And while I'm making a wishlist of things for you to implement:

      Targeting your strikes would be incredible. The best (or most ridiculously elaborate) system is dwarf fortress, but I think there was a Roguelike way back called Saladir, which if I remember correctly you consulted on? Anyway, it might be nice to explicitly model the fight I described by allowing you to select a body part to attack (head, l/r arm or leg, torso) with different effects, (high crit chance, dropping weapon/shield, falling/slowing, high hit chance). Larger and larger monsters would only allow hits to the lower and lower parts of the body, at least until you chop them down.

      Missile attacks would theoretically be able to hit anywhere on any size monster, which would make it a very useful skill to have (currently in ADOM 1 I only train up so I can deal with clawbugs and use -slaying ammo properly)

      I guess the problem then would be allowing yourself to be damaged in the same way. Currently DF has a problem in which a successful leg strike is nearly always fatal to the PC, as you are much more vulnerable, but can't escape. It also expects you to heal in real time, and doesn't allow for the body to repair nerve damage.

      Not that I expect anything to come of this, but food for thought.

  5. @Sam: It's good to discuss that. Your comment just made me re-introduce armor-penetrating hits in ADOM II. So you can get the best of both worlds. They will be somewhat rarer than critical hits (for monsters; about the same chance like criticals for the PC) and only will allow you to bypass PV from protective gear (e.g. armor, helmets, shields, ...; not rings of protection, magical bracers, etc.) and thus help with the fire giant problem.

    Additionally they will have some potential to damage armor permanently. Which soon makes smithing worthwhile, etc. ;-)

    1. re-introduce armor-penetrating hits in ADOM II.


      Also, RFE: giant claw bugs and death! Wailing, despair, gnashing of teeth. Also, in the round when claw bugs kill you, the text should be in german regardless of language localization.

      Klauebugstarb und Jammerndeverzweiflungzähneknirschen?

      That's my one german-words-are-long joke for the month. I think some part of that compound is probably inflected wrong anyway.

  6. HOw about using speed as a factor? Make very large enemies slower and give the PC more speed relative to the biggie. I haven't played much of the new ADOM ( keep getting slaughtered), so I haven't seen any really big enemies yet.

  7. Critical hits should be dexterity dependent. Dexterity for most large NPC's should be low, making critical hits unlikely. The issue should be addressed at least to some extent, however, by the reintroduction of missile attacks. It is difficult to balance toe-to-toe combat between giants and non-demigod player characters.

    On a hopefully easy to implement ease of playing note, it would be nice to see text color linked to cursed / uncursed / blessed items in inventory lists once more.

  8. @Aral: Colored items are on the TODO list ;-)

  9. Agree with Cheryl on trying to nerf large beings by reducing their speed (and their DV, I suggested this in the forums) so that the damage doesn't need to be reduced (or not so much) and the diversity of sizes stays. I think the diversity of sizes is a great addition to ADOM II and, while I'm sure it needs balancing, it would be a pity to dilute it so much that sizes don't really matter much.

  10. The game is at an early stage, so definitely no need to panic about balance issues. :-) I very much like the general idea of balancing size through dex and / or speed, as suggested above.

    Disarming could perhaps be tweaked a bit as well? For example, making success (more?) dependent one dex, and maybe also on (monk) level?

  11. @blurry: On the contrary, I believe that it's important to get good, balanced gameplay as early as possible. Things are much more difficult to fix at a later stage, when all the various features affect each other so much.

  12. I think having a tendency to reduce speed (and maybe to-hit, a little bit) for large monsters is a decent solution. There was nothing wrong with how (greater) molochs operated in ADOM I, for example. They're dangerous, but not terribly so and can be avoided fairly easily if the PC isn't strong enough to deal with them. Having a similar strength monster that is as fast or faster than a typical PC is more of a problem. Sam's idea of making critical hit probabilities size-dependent might not be a bad idea as well.

  13. heres an idea, most of the really big enemies have low inteligence. exagerate that and make them easily vulnerable to confusion, if a hurthling ever killed a humongous troll, he would have to outwit him, if large enemies are stupid and slow we could make it easier for wizards and smarter characters to get behind them and take them out.

  14. I maintain that the current stat drain system just takes away the fun of playing the game, since there seems to be almost no way other than athletics of increasing stats. One of the better features of ADOM I was the rational way it handled stats. I'd like to see something that resembles it. Your proposed solution seems to make it all but certain that ever character will still suffer massively from stat drains with only a small hope of recovery, which doesn't seem like much of an improvement.

    1. I don't think what he's saying is as bad as it seems. The probability formula being suggested means that stats will probably recover about 40-50% of the time, typically, but could easily be as high as 100%. 500d500 is WAY too long--that's 125250 turns per stat point, pretty much exactly*, and a win currently takes (based on only one data point...) on the order of 100000 turns. I'd think about tweaking it down to more like 5-10000 turns and it shouldn't be too bad.

      What I'd like to see is a system where there are lots of things that you can do passively to slowly train (or hurt)your stats--like being burdened trains St, using missile weapons or dual wielding trains Dx, buying and selling goods trains Ch, finding secret doors trains Pe, etc. I think stat training in ADOM I is a little too easy, honestly, particularly with herbs and ring of weakness-related exploits.

  15. I think a recurring quest or some lengthy work on the players part should be able to restore partial or all stats. Finding the fountain of renewal deep in a cave that requires a valuable sacrifice in order to use, the power of the restoration is correlated with the value of the item. Full stat regeneration for instance would require an artifact level item.

  16. Regular hobgobs and orcs with great weapons seems too much.

  17. +1 to the dex-based crit system concept (with larger creatures having lower dexterity). Seems intuitive - larger creatures have bigger swings with more momentum, so you can see a blow coming much easier. Could affect to-hit probability in general too.

  18. Big creatures should hit harder than small creatures, that's sort of the point of them. Lowering to-hit chance and crit chance based on size difference seems like the sensible approach. It would end up something like: small creature has no problem hitting big creature, doesn't do all that much damage but has a decent chance of a critical hit. Big creature finds it difficult to hit a small creature, and even more to inflict a critical on it, but every hit causes a lot of damage.

    Also, as the game is being developed, the player will have more options to deal with threats: missiles, spells, wands, etc. And there wll be more monster variety, which should decrease amount of giants ;).

    When it comes to stat drain, there are two possible approaches: easy to gain and easy to lose (like in ADoM 1) or hard to gain and hard to lose. Personally, I enjoy the stat system of ADoM, as opposed to some other games where gaining stats is very limited (crawl, tome). The problem now is that there just aren't many options for avoiding stat-draining creatures, as well as regaining stats, but this will likely improve as new features get implemented.

    When it comes to balance, there is one thing that I find very annoying: danger level based on character level. When I find a dungeon too difficult, the usual response is to leave, get to an easier one, level up, get better equipment, go back to the difficult one. Right now, if I meet something I can't beat, there's a pretty good chance I meet one in the next dungeon too. And finally, I think leveling up should be always desirable. There were exceptions to that in ADoM (SMC, level-dependent Kranach and pyramid quests), but exceptions are fine. The rule isn't. When gaining a level means that each and every dungeon level everywhere just got more dangerous, something is wrong ;).

    So, what's the alternative to that? I'd say a fixed danger level for each generated dungeon, and increasing with depth. If a character can't beat trolls, just let him hunt goblins for a bit ;).

    PS. My comments have a tendency to disappear for some reason.

  19. +1000 to everything qui said, he managed to write a quite long comment where I agree with every word!

  20. Maybe after stat X is drained, you can get immunity from draining of stat X for a certain number of turns? I ran into a tension-room full of strength stat drainers and had my strength reduced to 2 before it was over.

  21. Just a quick note, where I wanted to say: Don't worry too much about balance until Missile weapons, Wizard magic and clerics/gods are added to the game! You might have it perfectly balanced in the current iteration (melee only), only to have it become massively unbalanced when ranged attacks are added. Or it might seem unbalanced now, but will suddenly become more balanced when ranged is added!

    I like the minor chance of recovering a drained stat, particularly because it uses willpower/concentration, and makes these stats/skills more useful to a non-magic PC.

  22. Without wizardry and ranged weapons implemented, it's not really worth getting to brass tacks with melee damage balance. In ADOM, all but the most powerful melee-centered characters will die if they try to fight every single creature in melee right away. Sometimes you either run, or use missiles, spells, wands or other items to get an edge. If you go up against a sufficiently huge troll, it SHOULD be a bad idea to melee unless you're equally brutal. Your main advantage as an adventurer is a varied strategic and tactical approach.

    Incidentally, a combat AI that uses potions of blindness and wands of paralyzation AGAINST me if they pick them up or are generated with having either in the inventory would kind of make me happy in the pants.

  23. To put it in other words, it's not the big hulking creatures with the big hulking weapons that need nerfing right now; it's the PC who needs to get access to approximately two thirds of his potential offensive arsenal!

  24. So much yes on monsters not hard scaling with character level, by the way. It's been tried in many CRPGs of the last years and from what I hear it's always been terrible. If you worry about places being unfairly hard, add a skill, or give the Survival skill a potential benefit, in detecting danger. Eerie sounds, eerie lack of sound terrifying wails in the dark, decidedly too large prints of something that could be foot, but more likely the limb of a dreadful creature whose name you will not speak... Scaling all the way from "Pawprints and the remains of a goblin meal: It appears there are some predators on the loose." to...

    "You feel a sudden urge to run... right now."

  25. "@Aral: Colored items are on the TODO list ;-)"

    I hope this includes colored words of 'hit' and 'critical hit' during combat.

    1. Those are in 0.2.3... or do these not show up for you?

  26. Yes, please no level scaling! I've hated level scaling ever since I played Oblivion and completely lost the game because I focused on non-combat skills, and was not able to progress any further in the game because all easy monsters in the game had been replaced with high level enemies.

    I feel that dungeon danger levels should be somewhat intuitive; partly based by location. If I travel to the furtherest corners of the map, if I travel to the middle of a mountain range, if I travel to an untouched island, I would expect high level dungeons and fights. But if I travel to rat dungeon which is half a days travel from the starting town, I'm expecting an easy fight, even if I'm level 100. Things like that would make the world much more interesting, and rewards player logic a bit too.