|
Post by Variaz on Mar 2, 2010 12:23:41 GMT -5
The latest characters builds that have been posted all have very high numbers. 1.7 billion damages, 80 million hp, and other things like that... Now, if those characters were level 100 or higher, it would have been fine by me. But level 40? This is too high.
Extra lives for monsters were added mostly to balance high damages and give some diminishing returns to focusing entirely on damages. This works, but I don't think it's a very fun solution. Having every monsters with extra lives isn't fun, as this should be something reserved for very powerful monsters, but if I don't put something like that, then every encounters becomes trivial because damages are just that high.
From a technical point of view, we're getting close to the 32 bits integer cap, which means that without moving to a 64 bits application, we're going to get(and already getting in some cases) numeric overflow problems. Rather than move toward 64 bits, I would prefer to reduce the numbers and make things a little more balanced.
I think the main things that causes this problem are:
- The multiplicative bonus from abilities. Putting points in two different abilities that multiply damages results in higher damages than putting more points in a single ability. Now, add a third ability, and even a fourth, and the numbers scales really, really high. There is one advantage, however, is that it encourages multiclassing, whereas making those passive bonus additive would make multiclassing less attractive. It would also make some passive abilities overall a lot better than others(why use Mastery of Elements? Why not just use Weapon Mastery/Sheer Power with a fire weapon?).
- The levelable items. With all items giving the possibility of getting stats and skills to very high numbers, this has caused the numbers to scale. One of these numbers is hp. Hp can now get really, really high, and this has forced me to raise the damages output of all enemies to higher levels. This is causing a balance problem, because Constitution has become a "god" stat that no characters can live long without.
- Base damages scale is too big. Even with hundreds of blows, daggers cannot really outdamage powerful two-handed weapons. For ranged weapons, this gets even more problematic, especially when combined with high hit rate, and for firearms, simply ridiculous.
I would like to hear some opinions about how we could rebalance the game a bit, and find alternative to really high numbers that would still make the monsters hard to kill, but maybe remove the need of putting extra lives on every monsters...and 500k damages outputs on depth 50 enemies!
|
|
khan
Veteran
Posts: 81
|
Post by khan on Mar 2, 2010 13:38:14 GMT -5
I agree - damages are way too high, but I don't know how you deal with it without significantly rewriting the game mechanics (very tedious) and then rebalancing all the monsters (very very tedious). I really enjoy this Angband variant (the character building options are by far the most varied and interesting and I actually find myself interested in the plots and sub plots and occasionally you actually have to think to get past monsters). However, currently when you build a character it either seems to be able to beat anything easily, or to be totally ineffective once you get past a certain depth, with very little inbetween ground.
As an example of the former I recently built an Elvish Archer with 20 levels in piercing shots and the rest in accurate shots. No need to worry about resistance, immunities, returning and accurate shots will get you past most counters. Combine this with 100s of milliions of damages per shot and 6 shots and there's not much that you can't kill from a distance (even managed to take out a death drake and an elemental elemental with him - PS the death drake was easier than normal as I turned of it's fire and dark spells since the game crashed every time he used them although this meant I had to put up with more mature prismatic dragons!)
|
|
|
Post by Variaz on Mar 2, 2010 13:48:52 GMT -5
Death Drake would have killed you. It's breaths would have most likely pierced your resistances(if any), easily bypass your magic resistances and just destroy you.
But he probably did over 2 billion damages, and therefore caused an overflow, which crashed the game. A perfect example of damages becoming too high.
|
|
kipar
Champion
Posts: 201
|
Post by kipar on Mar 2, 2010 14:35:26 GMT -5
As a partial fix for high damages problem I suggest changing way how armor class works. AFAIK right now it's pitch&toss - you dodge or take huge damage. Same with monsters - they dodge\block everything or gets killed with your overpowered weapon. If AC will decrease damages instead of just rolling hit\miss (something like fuzzy sets vs normal sets). So you can still have 50mil of damage but high armored monster will get only 500k. Same with you: you can choose between having high AC\magic defense and getting small hits or having high HP and getting bigger hits.
Another idea. Just decrease attack power of monsters and players leaving hp rolls the same. Let's weapons skills raise damages not by 100%\150% but by 10%\15%, first digit in all weapons base damages will be approx. three times lesser then second digit, monsters won't get such high mind\strength too or bonus from their skills won't be so big... Damages are ~level^3, so at some level they _will_ become too high anyway, but if this happens only at level 200 it won't be so scary.
|
|
|
Post by zekk on Mar 2, 2010 17:37:25 GMT -5
On the player's side of things, I would think lowering the percentage bonuses from skills would help. I don't know if 1/10 the value is a good idea, leaving spell-craft at a measly 1% boost per point. It might need different changes depending on the skill. For the monsters, I'm not really sure how their damage works, so I can't really give any input.
I remember there was an old HP system that gave you less overall. I forget what was different, but the player usually started off with.... 10 hp? unless it was a race/class with the now non-existent CON penalty, which had 5. Anyway, if the damages get lower that hp system could be used again.
I think redoing the damages might fix summoning as well. If I remember correctly, the main problem with summoning was that monsters did too little damage to each other.
|
|
|
Post by Gando on Mar 3, 2010 0:24:29 GMT -5
It is my humble opinion that you Variaz started the numbers race with the insanely hard to kill monsters. Naturally everyone found ways to make the most of their abilities. That said I have always felt smaller numbers with more meaning to each makes a better game. Bigger numbers give you plenty of room for rebalancing and exercising various math games but in the end smaller numbers do a better job of being efficient. Also as you said it is a matter of the compiler not being able to give us truely large numbers without overflow problems. (Though you could just convert to long integers and deal with them that way.) What I would like to see is a good progression rate for leveling (not the stop start it is now) + better survival rate vs the nastier monsters. I understand that they wouldn't be nasty if we could just walk through them so I am not asking for that. I am asking to have tactical elements matter more. As it is right now, if a monster hits you at the higher levels its a one shot kill unless you have an ability to stop it. (Forget Great Guard...useless against anything larger than a kobold really, but spell absorption works sometimes vs the nastier spells.)
First thing set a cap on how high you want things like Hit points and Armor Class to go. Then convert the to hit system to a percentile chance to hit rather than a vs system (a vs b). So with 1200 ac yon monster has to get a 20% on a roll of a 1d100 to hit. You can make it open ended to allow for criticals and the like.
Then decide what each level of monster should be capable of for that level of player. Obviously a level 100 rat should have a 99% chance to hit a level 1 character. and such a character should have 99% chance to miss such a monster or to have the damage done be minimalized.
The exact process I don't have an answer to but it should be something you can make a formula for and then project that formula into numbers.
|
|
|
Post by sekira on Mar 15, 2010 13:09:45 GMT -5
I have been dreading this day. I knew it was coming, the weapons race between characters and monsters couldn't go on for ever. To be honest I like the super difficult bosses, they really give you a sense of accomplishment when you kill them, but the arms race means that you end up with a nuclear option: whoever hits first wins. Defense eventually becomes meaningless. It makes sense when you consider that the number of multipliers that characters have for offensive purposes far exceed the number of multipliers for defensive purposes. Same with monsters, although blocking seems to keep up pretty good (but then characters can get attacks that bypass blocking... so the same still applies)
I think that this is not going to be a quick fix solution. I do think it is doable though, but when the fix is done, the scale of the numbers will be NOTHING like the scale used to be. I'm not saying portralis will be totally different, I'm just saying there will be no comparison between Portralis 0.4 and Portralis 0.5 in terms of what "good damage" is or "good hps" are, etc.
First we absolutely have to keep the differences in portralis (vs other angband variants) that are good and in and of themselves don't automatically cause number scaling problems.
1. I know Gando doesn't agree, but I absolutely love the power vs power simple contested roll standard. 2. Skills, including feats must stay 3. Classes, Class Abilities, and unlimited multi-classing must stay 4. Multiple bonuses from items (ie, 1 str and 10 dex, not just 5 PVAL) must stay 5. Damage types and variable resists (-100 to 100%) must stay. In fact, with a lower damage scale, resists will mean something again, as things stand now the only resist that really matters is 100%
I don't like simply "patching" in a new game mechanic especially when it is simply another multiplier... that just adds another branch to this arms race of numbers scaling. I'm talking about simply letting AC give an additional effect of reducing damage. All that does is give characters incentive to find ways to scale up their AC with multiplicative bonuses just like they already do with Damage. And of course, AC reducing abilities will become overpowered (shattering blow anyone?). I don't think monsters have any way of reducing the character's AC.
I think that "borowing" from other games when it comes to scaling will be very helpful.
I propose that for damage there should be a pre set number of multipliers for each type of damage, and all racial, skill, class, ability, and item bonuses should ADD an ammount to a predefined multiplier, not create a whole new multiplier.
for weapons: Base roll (dice and sides, etc) * skill multiplier * stat multiplier * Ability multiplier.
To keep things managable, the multipliers should not be huge like a whole multiple per skill point... but smaller, like 5% per skill point, 10% per stat point, 10% per ability point
Keep in mind AC and to hit could have a set number of multipliers too, just like damage.
AC: Base AC * skill multiplier * stat multiplier * abilities multiplier
again, multipliers are NEVER huge, they should be about 5% for skills, and 10% for abilities or stats per point, note this is a perfect example of combining two multipliers into a single multiplier... Defense skill and Agility skill... No longer would they be seperate multipliers, now they each provide a bonus per point, and those bonuses are added together to get the single skill multiplier.
To Hit: Base to hit (not sure how to determine base on this one) * skill multiplier * stat multiplier * abilities multiplier
Now having the 4 multipliers on each of these values will still allow for multi-classing to be vital, but keeping each multiplier reasonable will keep the scale from becoming insane.
Of course blocking will have its own set of multipliers, spell damage will be different too. Perhaps spells will have their own Magic to hit and Magic AC values, so it isn't ALL dependant on the one skill magic defense. And of course, hp calculation will have to be rethought
I like something like this to keep things standard:
base hp (based mostly on level, like 50 per level I think) * skill multiplier * stat multiplier * abilities multiplier
Now, I'm not sure what skill would help with hp (defense?) maybe none, but it should still be there, to keep things standard.
Now, how does this work out with boss abilities... Haven't figured that one out yet.
Well, this is getting too huge already, so I'll leave it there for now, more to come later if V or the forums community shows interest.
Now, abilities will do more than just add 10% per AP to one of these various multipliers, they will do the many varied things they always did... except there should be a push to avoid multiplicative creeping... try not to give 10% per ap bonus to STR AND 10% per ap bonus to the abilities multiplier to melee damage. The same goes for skill feats... I'm sure feats that let a skill give a 5% bonus per skill point to an additional multiplier will not be uncommon. IE... Defense will give 5% bonus to the skill multiplier to AC per skill point, but maybe there is a defense skill feat that lets it also give 5% bonus to the skill multiplier to hp or maybe blocking, etc.
of course this doesn't even address how the monster scaling will have to be fixed as well. I think similar mechanics should be used.
|
|
|
Post by Gando on Mar 15, 2010 13:55:23 GMT -5
It seems to me the heavier the armor the less evasion it should provide and the more defense vs damage it should provide. I mean we are essentially tool users. Very broadly speaking sentience is defined scientifically by the ability to adapt and use tools in an environment that would not normally provide them. So to my mind armor is a tool used to reduce damage. The heavier it is the more protection you gain. Also the less agility you have to move around in (One of the problems of having 'Agility' as a skill name when it really describes an attribute in English.)
One rpg system that seemed to reflect this fairly well was rolemaster and another runequest. Both those had one system for hitting and an entirely separate for indicating damage/wounds/death.
Also it is often the case in rpgs that the heros do not need to KILL their opponents to win. They merely need to cause them to flee or surrender. This happens because of moral issues. Something the game also does not pretend to get involved with. I think a moral system might make combat more interesting.
Simple skill vs skill contesting is ok but I dislike the disproportion between skills and stats in this game. Skill vs stat seems wrong to me since the numbers are usually way different in scale.
|
|
|
Post by sekira on Mar 15, 2010 14:04:08 GMT -5
hmm, I just reread my previous post, and I really think it is a good idea, but I wanted to give a concrete example so everyone knows what I'm talking about.
Lets "build" a level 10 character in my vision of portralis 0.5
at level 10 that would mean 10 AP, 20 Stat points, and 50 Skill points to spend.
say we want a swordsman, so we spent 20 skill points in sword, 10 in defense, 10 in agility, and 10 in combat feats
we spent 10 stat points in str and 10 in dex
we spent 2 AP in Accurate Strike, 4 AP in Strength, and 4 AP in battle skill
For simplicity, I'll ignore race and class modifiers and item bonuses.
with a steel longsword (2d6) how much damage would this guy do? Well, lets first calculate the stats given the abilities...
STR 15 base + 40% (from Strength ability) = 21 DEX 15 base other stats don't influence melee damage
because of battle skill, we will calculate to hit first
For our purposes lets say base to hit is 10 per dex, and because battle skill adds to base to hit (this should be ok within reason, since it doesn't create a new multiplier) INSTEAD of adding to the abilities to hit multiplier, then it should be ok, but we have to be careful, for now lets calculate as it is written.
Base to hit = 15 Dex * 10 + 4 AP in Battle Skill * 20 = 230 skill multiplier = 100% +5% per skill point in sword = 200% stat multiplier = 100% + 10% per dex = 250% Abilities multiplier = 100% + 10% per AP (if using ability) = 120% with accurate strike or 100% with normal attacks
230 * 200% * 250% * 120% = 1380 total to hit with accurate strike, but since accurate strike is an active ability, the passive to hit would be: 230 * 200% * 250% * 100% = 1150
Now, damage calculation
base damage roll is 2 minimum 12 maximum skill multiplier = 100% + 5% per skill point in sword = 200% stat multiplier = 100% + 10% per str = 310% abilities multiplier = 100% + (3% of to hit per ap in Battle skill)% + 10% per ap in accurate strike = 100% + (12% of 1150)% + 20% = 100% + 138% + 20% = 258%
minimum 2 * 200% * 310% * 258% = 31.992 (round down) = 31 maximum 12 * 200% * 310% * 258% = 191.952 (round down) = 191
personally I don't think battle skill as it is written would work well in portralis 0.5 because it adds too much to base to hit (20 per ap should be reduced to 10 per ap) and it is adding an already multiplied value (to hit) as another multiplier for damage, which is multiplicative creep. it should add only a % of the base to hit not the total to hit. Even this might be problematic, but has much less chance to get way out of hand.
See, the only way this will work all the way to level 200 is if we are EXTREMELY careful to avoid multiplicative creep at every little step along the way. Otherwise, we will simply be trading in one set of overly multiplied abilities and skills for another set.
I hope that explains things a little better
|
|
|
Post by Variaz on Mar 17, 2010 16:01:30 GMT -5
One of the biggest changes to be made for the next release will be to get rid of the multiplicative bonus altogether, and instead make the abilities more varied instead.
Right now, you take Weapon Mastery + Element Mastery with a Fire elemental sword for maximum damages. What would be ideal is instead get the same total bonus for every points spent(+10%) and let them be additive. Instead of relying on multipliers, each abilities would be given an extra benefit that has yet to be determined, but that would make the two abilities different, while providing the same amount of damages enhancement. So you'll have the choice to focus on either of them, or have a mix of both, but all choices would be viable since the main thing you want is damages boost and you'll get it, the rest would be extra.
Sekira is right that with damages rebalance in 0.5, the game might become very different, but in a good way. Right now, I feel the "epic" feeling of the game is lost, since a boss with millions of hit points isn't scary anymore since you do damages three times it's health per blows, or more. This makes combat reliant on counters and extra lives, which ends up normalizing all three main combat styles, but it's not necessarely a fun way to do things. A monster with extra lives should be something scary that will make you think twice about facing that monster. Right now, it's a needed mechanic to prevent the game from being laughably easy.
|
|
|
Post by Gando on Mar 17, 2010 23:02:23 GMT -5
One of the biggest changes to be made for the next release will be to get rid of the multiplicative bonus altogether, and instead make the abilities more varied instead. Right now, you take Weapon Mastery + Element Mastery with a Fire elemental sword for maximum damages. What would be ideal is instead get the same total bonus for every points spent(+10%) and let them be additive. Instead of relying on multipliers, each abilities would be given an extra benefit that has yet to be determined, but that would make the two abilities different, while providing the same amount of damages enhancement. So you'll have the choice to focus on either of them, or have a mix of both, but all choices would be viable since the main thing you want is damages boost and you'll get it, the rest would be extra. Sekira is right that with damages rebalance in 0.5, the game might become very different, but in a good way. Right now, I feel the "epic" feeling of the game is lost, since a boss with millions of hit points isn't scary anymore since you do damages three times it's health per blows, or more. This makes combat reliant on counters and extra lives, which ends up normalizing all three main combat styles, but it's not necessarely a fun way to do things. A monster with extra lives should be something scary that will make you think twice about facing that monster. Right now, it's a needed mechanic to prevent the game from being laughably easy. It is also a mechanic that makes the game laughably hard. Too hard for some class combinations.
|
|
|
Post by random on Mar 18, 2010 11:34:36 GMT -5
It is also a mechanic that makes the game laughably hard. Too hard for some class combinations. I was thinking that. I'm not sure it's so much an arms race as it is an attempt to compensate for fairly specific character builds. I've never even bothered with a gunner after reading about how powerful they get, but I've had non-ultra build level 40 characters that were instakilled by bosses on level 40 dungeons. I know bosses are supposed to be hard, but shouldn't a non-ultra build be able to survive at least one blow from a monster on a non-out of depth dungeon level, or inflict enough damage to at least have a sporting chance? I don't pretend to be intimately familiar with the numbers and formulas being talked about on this thread, but it does seem to me that jacking up monster stats isn't the only way to deal with certain builds...tinkering with the classes and abilities involved in the builds themselves is also a viable approach.
|
|
|
Post by Gando on Mar 18, 2010 18:21:42 GMT -5
Well the classes are not and should not be on the same playing field. The races still need a ton of work imho ....(cf my old thread concerning new potential races from last year) but even so they will not be balanced against each other. The variation between class and race choice is what gives some people the ability to replay the game with various different strategies. Balance it all too much and things go to heck. On the other hand different approaches should be valid for defeating monsters at any given level. Honestly though without redesigning everything from scratch I don't have a good solution and I am inclined to agree that patching in fixes will only exacerbate matters in the long run as players end up needing to find new ways to level faster (leveling slowly is a serious issue in the game at certain experience breaks), get better items (another power creep issue) and thus seek higher power levels.
|
|
|
Post by random on Mar 18, 2010 19:50:40 GMT -5
Which is pretty much what I wanted to say. Right now, the issue isn't precisely overpowered characters -- it's specific build strategies that lead to overpowered characters. Your experiences with rogues and bards are a good case in point -- it's not really feasible to beat the game by employing the natural strategies involving these classes. The really powerful builds, gunners and elemental lords and the like, are at the root of the original premise of this thread. The strategies with those are fairly straightforward and basically involve powering your way to the top. A variety in this regard would help alleviate this issue, or at least alter the nature of the problem...if a rogue can feasibly win the game, but not by the old "staggering amounts of damage" approach, then it's less of an arms race and more of issue of tailoring the game to the various approaches. In the case of the rogue, for instance, if certain quests involved deathly powerful monsters that needed to be bypassed but not necessarily killed, a sneaky sort would be at an advantage over a gunner, especially if the monsters were, say, uniformly immune to physical, mana and missile damages...but vulnerable to poison.
|
|
|
Post by Gando on Mar 18, 2010 22:41:02 GMT -5
Yeah stealth + poison is generally a bad strategy unless you are somehow abusing the One With Shadow thing. (Not really sure how that helps but apparently it does.)
Actually the most broken builds involve multiclassing
Defender + Warrior + Paladin + Elemental Lord is one route Defender + Soul Guardian + Bard is another and Crafter + Gunner or Crafter + Kensei is a third. There are a couple more but those are the ones with the staggeringly high numbers that I know of. Mages tend to be really horrible. Priests less so because with wisdom casting you never run out of mana. Diviners are another potentially very strong class but some of their major abilities are broken in bad ways. Overall I think the situation as it stands now none of the classes not mentioned here in combination with the others would be able to beat any of the higher end bosses (Simon included) much less be able to handle the current top tier of monsters (level 50s+).
I think the solution has to be in revamping combat and how it works and probably bringing magic into line with melee powerwise. Or at least give mana restoration a boost.
Really though we could discuss what needs to be fixed for ever and ever but it all lies in Variaz's hands.
|
|