Improving the Absroption Formula

[Bluddy]

I think it may be important to state that my suggestions don't change the current formula that much. Even my favorite option right now, which is the linear formula (post #2), keeps most of the values just about the same, especially around the lower levels/armor values. It's higher levels/armor values which are affected.

Speaking of which, does anyone have a plate/mail wielding character to donate? I really would like to get some samples with high armor values.

[Crisses]

Originally Posted by Bluddy

I'm pretty sure the current formula only allows mages to reach 10-25% absorption at most. I think mages are expected to use teleport, shield spells and potions, and to die a lot.

Actually, it would be nice if Cloth (and maybe Leather) armor -- especially body armor -- got a bonus that no other type of armor gets. For example, how about up to an extra 2 enchantments? Or stronger enchantments.

The D&D reasoning for mages "not being able" to wear metal armor is that it somehow blocked or reflected the magic of the environment that the mages needed to gather to power their spells. After D&D made this choice, all role playing games since have basically followed suit. Utter fabricated nonsense, but if we go with that reasoning, metal armor types wouldn't be able to have as many enchantments -- some bonuses could reflect superior craftsmanship, species workmanship, etc. but you would find less actual enchantments on metal armor/objects. Ok, we're not going there sooooooo----

What if mages, instead of avoiding metal armor because it somehow blocked their magic (it doesn't seem to in Din's Curse) preferred cloth armor because of it's special powers or bonuses. The mage schools do not encourage or train metal armor use BECAUSE it's worthless in some way compared to using cloth armor. i.e. what if Gandalf or Merlin wore robes BECAUSE they were robes, not because they were the last resort resembling armor that they had to settle for. Is that silver thread in the lining? Is that a symbol stitched on the inside of the sleeve? Was that hood woven with unicorn hair? How much EXTRA enchantment could you give to such an item? What is a perk that cloth armor could have that would make even a hybrid say "Hrm, maybe I should wear that quilted vest."

Another thought is more set armors for cloth (and leather) than for mail & plate. i.e. incentives to collect and use cloth, and possibly leather.

Right now there's basically only 1 reason to have a pure mage. The ability to specialize in fire or ice AND be able to teleport at the same time. You can't pour enough points into being offensive AND defensive. I wanted a fire mage with ice protections -- simply isn't happening. I had to pull all my points out of the ice chain because I simply am not hurting critters fast enough -- with only 2 offensive fire spells! So basically the character has 4 skills: Fireball, Fire Storm (is that what it's called?), Fire Mastery & a point in Teleport. I'd frankly prefer to have a hybrid with either a healer, a fighter or even possibly hunter or rogue to get the better armor, weapons and actually be able to defend myself because all my points in area effect fire attacks isn't enough to keep critters from beating me up.

I'll keep up with playing the class but it's already seemingly pointless to have chosen the pure-class -- maybe I'm wrong and it evens out quickly.

[Bluddy]

Originally Posted by Crisses

Actually, it would be nice if Cloth (and maybe Leather) armor -- especially body armor -- got a bonus that no other type of armor gets. For example, how about up to an extra 2 enchantments? Or stronger enchantments.

Not a bad idea. But it can't be a big enough advantage to offset the armor disadvantage, or else leather characters will wear only cloth. It's a little screwy that the warrior gets the best armor AND the best weapons, no? Oh well.

Right now there's basically only 1 reason to have a pure mage. The ability to specialize in fire or ice AND be able to teleport at the same time. You can't pour enough points into being offensive AND defensive. I wanted a fire mage with ice protections -- simply isn't happening. I had to pull all my points out of the ice chain because I simply am not hurting critters fast enough -- with only 2 offensive fire spells! So basically the character has 4 skills: Fireball, Fire Storm (is that what it's called?), Fire Mastery & a point in Teleport. I'd frankly prefer to have a hybrid with either a healer, a fighter or even possibly hunter or rogue to get the better armor, weapons and actually be able to defend myself because all my points in area effect fire attacks isn't enough to keep critters from beating me up.

Yeah it seems like spells are still weak. It's been mentioned elsewhere. The spell skills need to increase naturally with level, or perhaps with intelligence. Priest spells should increase with spirit. But that's a discussion for another thread.

BTW the other advantage wizards are supposed to have is very high resistances, but right now I don't think those work too well with the current function -- they don't go high enough.

[Crisses]

Originally Posted by Bluddy

Not a bad idea. But it can't be a big enough advantage to offset the armor disadvantage, or else leather characters will wear only cloth. It's a little screwy that the warrior gets the best armor AND the best weapons, no? Oh well.

I don't think of this as an "oh well" thing. A big strength of Dins is mixing hybrids but two or thee classes seem like a waste of time. If you Have To hybrid all the Mage or priest classes to make a viable character the game is out of balance. Over 10 years later people are STILL playing Diablo2 and the classes there are very well balanced.

Most of a fighters power comes from their equipment which is independent of skill point allocations. Stat pts only act as gatekeepers and when the stat drops the weapon or armor remains functional. That is broken in terms of game balance between classes. A goon pumped on potion steroids with a big sword hits hard no matter where he put his skill points. Priests and mages MUST rely on skill allocations for their effectiveness in game play. But what if a cloth robe gave you an extra 5 skill pts when you level? Or experience bonuses? Etc. Something serious without it being a direct boon in battle.

[Maledictus]

I'm having a hard time reading that sheet. The anchor value for any calculation is armor, so as far as I'm concerned the need for min/max determination is not relevant (although I get why you would want to do that, given the spikes at the high end of the linear sheet values). The current (old) function produces consistent (and correctly related) values across the spectrum, it strikes me as being of a simple elegant beauty. The linear function spikes weirdly at high armor values. Also, the current function has it's primary effect in the area of enemylevel-10 to enemylevel+10 (and it is consistent in this), which is as it should be in my opinion, while the linear function can produce wild fluctuations within that same area. This means that the current function will always produce a workable value, no matter what the entered armor value is.
Finally, in it's current incarnation (with the current constants) the absorption % is lower for any location in the 'linear' sheet than it is for that same location in the 'old' sheet (as far as I can tell from sampling a number of armor/level relations), but this can be attributed to the constants (what am I missing here?).
From this I'd almost conclude that it's not the current function that is wrong, it's the limited availability of armor value at higher char/enemy levels.
Again, this is my interpretation and I could be completely off my rocker. Blame that on ignorance.

On a sidenote: a player's experience (hard/easy), and therefor his/her frustration/elation level, is directly connected to the choices made at town creation; enemy level is x levels above current char level, etc. This taints the observations, in my opinion. As I said elsewhere, the game should be balanced for fun factor and difficulty at a given level "enemy level=char level+x", where I would think x should be one value for all test situations (0, I suppose). Meaning, if someone says "I get killed all the time by enemies 10 levels higher than me", I'd say "and so you should be".

Good grief, past midnight here... Spent another evening on studying instead of playing... Must say though this is very entertaining, and educational.

[Bluddy]

Originally Posted by Maledictus

I'm having a hard time reading that sheet. The anchor value for any calculation is armor, so as far as I'm concerned the need for min/max determination is not relevant (although I get why you would want to do that, given the spikes at the high end of the linear sheet values). The current (old) function produces consistent (and correctly related) values across the spectrum, it strikes me as being of a simple elegant beauty.

I'm glad you're looking at the sheet. First, look at the old function at level 1. You see that after 300 it passes 80%. By 700 it's at 90%. After that you get less than 0.7% per 100 points! This is as 'spiky' as the linear function, since you have to cap the maximum somewhere anyway, and 100s of points get you almost nowhere.

Now let's look at level 50. Notice that the constant (2nd column) is 1300. That means that around 1300 the function, which is very flat as it is because the constant is so big, becomes very slow. At 1300, an extra 100 points of armor gives you only 2% more! At 1600, an extra 100 points gives you only 1.5% more. 200 points can be the difference between mail and plate, or leather and mail! You're getting virtually no return for getting much better armor.

In contrast, the linear function splits things evenly. You get over 5% between each 100 point gap, and 200 points gives you a 10% advantage over the previous armor. I still need to tweak the linear function to realistic values found in the game, but the 'spiking' really doesn't matter since there's a maximum % value. The nonlinear function may be more 'elegant', but it kills the higher levels unless armor starts giving massively bigger numbers at the high end.

[Baki]

Originally Posted by Bluddy

Now let's look at level 50. Notice that the constant (2nd column) is 1300. That means that around 1300 the function, which is very flat as it is because the constant is so big, becomes very slow. At 1300, an extra 100 points of armor gives you only 2% more! At 1600, an extra 100 points gives you only 1.5% more. 200 points can be the difference between mail and plate, or leather and mail! You're getting virtually no return for getting much better armor.

Please do not forget while trying to balance things that raising the absorption from 40% to 50% your damage mitigation isnt raised by 10% but by 10 percent points. The number that matters is the dmg taken in %. by raising the absorption from 40% to 50% you lower your dmg taken by 16,67% (1/6). To give an extreme example: if you have 99% absorption and raise it to 99,5% you cut the damage taken in half.

a linear armor -> aborption formula is nearly never a good idea. Gaining the same dmg taken mitigation in percent per additional point of armor should be the target.

Originally Posted by Maledictus

On a sidenote: a player's experience (hard/easy), and therefor his/her frustration/elation level, is directly connected to the choices made at town creation; enemy level is x levels above current char level, etc. This taints the observations, in my opinion. As I said elsewhere, the game should be balanced for fun factor and difficulty at a given level "enemy level=char level+x", where I would think x should be one value for all test situations (0, I suppose). Meaning, if someone says "I get killed all the time by enemies 10 levels higher than me", I'd say "and so you should be".

you would think so but if the player can 1 hit level+10 monsters and still dies while he tries to be no glass cannon he will probably also vs monsters level+0 monsters.

i hope i got the correct data here:

MinDamage 1.0
MaxDamage 2.5

MinDamagePerLevel 0.5
MaxDamagePerLevel 1.2

BaseStrength 3
StrengthPerLevel(warrior) 4.0

assuming a monster gets 1% dmg per str a level 50 monster would have an avg dmg of 44,25*3,03~134

level 60 would be 52,75*3,43~181

thats an increase of 35%. Yes, you also get a bit less mitigation and the attack rating is higher but it won't matter that much. If you die within a second vs 60 monsters you would probably die within 1,4 seconds vs level 50 monsters.

I am not saying Din's Curse should be made easier. I only say 1 hitting and getting 1 hitted isnt fun. Let me be able to take a bit more damage and do the same for the monsters.

[Bluddy]

Originally Posted by Baki

Please do not forget while trying to balance things that raising the absorption from 40% to 50% your damage mitigation isnt raised by 10% but by 10 percent points. The number that matters is the dmg taken in %. by raising the absorption from 40% to 50% you lower your dmg taken by 16,67% (1/6). To give an extreme example: if you have 99% absorption and raise it to 99,5% you cut the damage taken in half.

a linear armor -> aborption formula is nearly never a good idea. Gaining the same dmg taken mitigation in percent per additional point of armor should be the target.

It's true that the damage percentage change is not the same as the armor absorption change -- I'm not sure how that changes things, though. If a monster hits me for 200 points and my armor absorbs 40% damage, I get hit with 120 points of damage. If that changes to absorbing 50% damage once I put on plate, I get hit with 100 points of damage. I don't want a big armor upgrade to move me from from 40% to 41%, which will reduce 200 points of damage to 118 instead of 120. That's pretty much useless, and that's what the current formula does.

Not in response to you, but to restate the problem: as character level goes up, the armor function needs to cover a bigger scope of possible armor numbers. This stretches the function out (like a string), making each increment smaller, and nullifying armor increases. The problem is much worse because the function is non-linear, making the big increases at the bottom (where they matter least) and leaving even smaller increments at the top, (where you really need them at higher levels) . I propose 2 possible solutions: a linear function that evens out all the increments, or modifying the use of the non-linear function so that the 'meatier', more linear part is used mostly, making the increments bigger that way.

[Bluddy]

OK Baki, after seeing those character stats you posted in the repository thread, I realize there's just no way a linear curve could do the trick. If armor can get to 4000... 7000 - forget it, it's completely non-linear. I'll work on the non-linear function and see if I can get it to something that constitutes an improvement.

PS. I really should look through the "show your character" thread for stats.

[Baki]

Originally Posted by Bluddy

It's true that the damage percentage change is not the same as the armor absorption change -- I'm not sure how that changes things, though.

It shows the flaw of linear armor -> absorption formulas

WoW has a good armor -> mitigation formula and dins curse currently uses a similar formula with different values iirc.

this is a good read: http://www.wowwiki.com/Armor

i think we just have to tune the numbers a bit. it has been done alrdy with the last patch but it wasnt enough.