Hardiness and Armor

[EvilEye]

I've been running some tests to see if its worth it to increase hardiness, luck and use some of the damage decreasing traits.

 

All tests done with the human race and on normal mode, using the goblin in the beginning cellar as the damage doer. Tests done over many times ( over 50 each ). The goblin has a chance of 50% to hit.

 

------------------------------------------------

 

Normal ( no traits, no armor, no hardiness, no spells ): goblin does 5 - 11 damage per hit

 

Good Constitution at level 1: goblin does 4 - 11 damage per hit

 

Good Constitution at level 30: goblin does 4 - 11 damage per hit

 

Thick Skin at level 1: goblin does 4 - 11 damage per hit

 

Thick Skin at level 30: goblin does 4 - 11 damage per hit

 

Thick Skin AND Hardiness at level 30: goblin does 3 - 11 damage per hit

 

Wearing Leather armor ( 10 defense ): goblin does 4 - 11 damage per hit

 

20 Hardiness: goblin does 1 - 9 damage per hit

 

Protection spell ( skill 20 ): goblin does 4 - 8 damage per hit ( did not decrease chance to hit at all, the description is WRONG! )

 

------------------------------------------------

 

So in conclusion both thick skin and good constitution seem to have the effect of wearing around 10-15% armor for each and do NOT increase with level. Combined it seems they add 22-30% armor or 2 seperate 10-15% armor checks. Much more testing with higher numbers is needed to determine exactly.

 

Also armor does NOT decrease damage by a % per piece, instead it appears to have a chance to decrease damage %. What this chance is exactly only Jeff knows.

 

Hardiness seems much more impressive then armor, apparently sometimes decreasing damage by 4/5 at level 20!

 

The Protection spell did not decrease the goblins chance to hit at all as it stated, it was still 50%.

[Ephesos]

Impressive work. I never really liked Hardiness as an ability, though... needing 20 to make a significant difference seems overly steep.

 

What about different levels of Hardiness, say 5/10/15?

[Alorael at Large]

20 Hardiness is expensive. Great armor is free after time spent searching and mutilating.

 

—Alorael, who in a thoroughly qualitative test didn't notice any appreciable difference between having and not having defensive traits when wearing moderately good armor. The defensive bonuses get swallowed. Conclusion: those are two really unhelpful traits.

[EvilEye]

Quote:

Originally written by Ephesos:
Impressive work. I never really liked Hardiness as an ability, though... needing 20 to make a significant difference seems overly steep.

What about different levels of Hardiness, say 5/10/15?

Haven't bothered testing that yet.

Since hardiness starts at 1 cost it only takes roughly 110 skill points to get to 20.

Thats a lot but I am thinking you could neglect all other defensive skills except maybe luck so you might actually come out ahead in points.

Kinda why I was testing it in the first place.

[Randomizer]

Hardiness increases armor at 1 point equals 1% reduction. So Hardiness 20 means 20% less damage which is what you got. It works better versus fire, energy, cold where it does 2% per point of hardiness.

 

It's hard but you can raise enough skills: dexterity, endurance, and hardiness to 8 to get resistance which does 4%. Helps your fighters but not spellcasters.

[Delicious Vlish]

Good constitution and toughness, I complained that those were useless in the beta.

 

I was right.

[Slawbug]

Armor (from Hardiness or equipped armor) DOES decrease by a % of the damage. It's just that this % is often small, and pretty much all forms of damage in A4 have a very wide variance for how much damage they can cause before armor reduction is applied.

[EvilEye]

Quote:

Originally written by Randomizer:
Hardiness increases armor at 1 point equals 1% reduction. So Hardiness 20 means 20% less damage which is what you got.

No, its not what I got, read it again.

20 points in hardiness sometimes decreased 5 damage to 1 damage.

Quote:

Originally written by Slartucker:
Armor (from Hardiness or equipped armor) DOES decrease by a % of the damage.

I never said it didn't.

BUT.

Its clearly random as to what % it decreases it by. I am guessing its from 0% - 10% for leather armor, instead of a straight 10%. So sometimes it will not decrease the damage by any, which is rather lame.

I am guessing hardiness works in the same way except I suspect the % for hardiness is around 4 a level instead of 1 a level based on my results.

[Slawbug]

Quote:

Originally written by EvilEye:
20 points in hardiness sometimes decreased 5 damage to 1 damage.

You're making a false assumption here. You cannot directly see the amount of decrease in damage hardiness causes. Why? Damage is rolled randomly EACH TIME an attack hits. So, unless you have more information than you are giving (for example, if the attack had a damage range of 5-10 against 0 hardiness and 1-2 against 20 hardiness, over many trials of each), you can't draw that conclusion.

I've never seen anything to discount the conclusion the damage is random, but armor % reduction is consistent. If you have new data I'd love to see it, though.

[EvilEye]

Quote:

Originally written by Dame Annals:

Quote:

Originally written by EvilEye:
20 points in hardiness sometimes decreased 5 damage to 1 damage.

You're making a false assumption here. You cannot directly see the amount of decrease in damage hardiness causes. Why? Damage is rolled randomly EACH TIME an attack hits. So, unless you have more information than you are giving (for example, if the attack had a damage range of 5-10 against 0 hardiness and 1-2 against 20 hardiness, over many trials of each), you can't draw that conclusion.

I've never seen anything to discount the conclusion the damage is random, but armor % reduction is consistent. If you have new data I'd love to see it, though.

I already stated in my first post that the damage range for the first goblin is 5 - 11 on normal difficulty.

I tested this hundreds of times ( no armor, no traits ). Never got anything below 5, or above 11.

So what is unclear about that?

[Slawbug]

You're right. Err, not sure how I missed that part. Sorry!

 

So, you're right, this is pretty weird. Now I am thinking these results may be affected by some kind of minimum value requirement imposed at some point in the calculation, or by a weird way of dealing with remainders. The data would probably be easier to interpret if it dealt with a much stronger attack (say, ranging up to 50 damage). Obviously, such data is also more of a hassle to collect.

[EvilEye]

Well, here is what I think is happening ( could be wrong of course ):

 

For every piece of armor damage is reduced by 0 - % listed on armor.

 

So if you had a 10% leather armor it would roll a number from 0 - 10. The number rolled would be the percent the damage is reduced by.

 

So if it rolled a 5 then the damage would be reduced by 5%. So 10 becomes 9.5, which should become 9 due to integers truncating the decimal value.

 

The reason I think it does this is because the previous avernum games reduced damage by a random number for every piece of armor. If you had a piece of armor that listed 1-16 then it would reduce damage by 1-16 every time you were hit by a physical attack.

 

For hardiness I am thinking it just adds 4% armor per level to its own specific armor check, which is why 20 hardiness could reduce 5 damage to 1 damage. Of course, the exact number is hard to tell without bigger numbers.

[Radix Malorum Est Cupiditas]

No, Jeff specifically said how the armor percentages work, 10% means 10% off the total, 10 gets turned into a 9.

[Slawbug]

Yes, Jeff was pretty clear about that. Also, if armor effect really was random as you suggest, then that means we would expect to see HUGE damage variances when a character wearing lots of armor is hit, much huger than the variances when an unarmored character is hit. And that's not the case at all.

 

Also, about Hardiness: Hardiness does not give 1% armor per point. It gives 2% armor per point, BUT this bonus is not displayed, AND it is subject to the 10-cap many skills have, so 10-11 hardiness gives 20% armor, 12-13 gives 22%, 20-22 gives 30%, 23-25 gives 32%, and so on. That means that your test at 20 Hardiness is really providing 30% damage reduction. So reducing the range max from 11 to 8 makes perfect sense. Reducing the range min from 5 to 1 makes less sense, but because the numbers are so tiny, I really think it's a min value or rounding issue. To clarify this issue, tests with bigger damage values are needed.

 

Also note that Luck displays a 1% armor bonus per point, but Luck does not actually provide any armor at all.

 

Protection reduces all damage by 25%. Steel Skin does as well. They are cumulative. Prismatic Shield reduces elemental damage by 0-50% determined randomly for each attack.

[EvilEye]

Why don't you guys just test it yourselves instead of telling me I didn't get the results I got.

 

If armor really did reduce damage by the listed percent EVERY TIME then I would never have got hit for 11 damage when I was wearing leather armor. NEVER. Its simple math.

 

I don't see why you persist on your beliefs that "Jeff said this so it must be the way it is". Based on my results, which I trust, he was either wrong or you misunderstood what he said.

[Dikiyoba]

Dikiyoba is reasonably sure that Slartucker and other people have tested damage, defense, hardiness, and every other thing in the game. Differing results and conclusions are to be expected. Studies that don't match up with other studies are everywhere.

[Fort]

The numbers are too small in my opinion. In some obscure way perhaps each die used in damage calculations is affected individually by percentages. I don't know, but I think it would useful to do these tests again with much higher damage numbers and something like an iron breastplate.

[Kelandon]

Spiderweb games are notorious for behaving differently on different machines that appear to have identical specs. Talk to anyone who was a part of the BoE Arena for an earful about this.

[Deckard]

I've done some testing by myself, here are the results:

 

All tests done with the human race and on normal mode, using the goblin in the beginning cellar as the damage doer. Tests done over many times. The goblin has a chance of 50% to hit. Character has no traits/skills.

 

No armor (0): goblin does 5 - 11 per hit

 

Begginer's armor (19): goblin does 3 - 10 per hit

 

Just my two cents...

 

Note: Begginer armor means all you can find in the first room that provides some armor, it consists of: Leather Armor (10 armor bonus), Wooden Shield (6 armor), Sandals (2 armor) & Poor Leather Helmet (1 armor). Total armor bonus = 19.