Celtic Heroes

^Don't mind him Swan

I love this, the science side of Celtic. Very well done and a good post! Clears up alot of questions

Keep up the quality posts!

Well done!

Welcome back Swan

^Don't mind him Swan

I love this, the science side of Celtic. Very well done and a good post! Clears up alot of questions

Keep up the quality posts!

Don't mind who?


For pve, mobs can hit 50-100% of dmg on you rather than 60-100 and the curves are much different from what you would find on pvp (these pve values would apply to tanks). I wanted to do the opposite as well, and find a curve relating boss resist to percent dmg taken. Similar to the method you described, I wanted to lure a mob down to 0 and from there slowly decrease the value of my lure until I stopped doing max dmg. From there, I'd further decrease the lure and find max/min at each 50 increment until I came up with a curve. Finally, I'd have to repeat on several bosses to see whether they followed the same formula and use my existing data to determine approximate pre-lure elemental resists of every endgame boss

I used blackstorm because it doesn't hit high and has a lot of hp. I found that bosses have some sort of built in resist, which doesn't allow you to lure to 0. I decided at that point it wasn't worth my effort to continue this project.

With 0 lure- skills did a maximum of 75.55-75.61% of skill page dmg
With 1169 lure- skills did a maximum of 98.26-98.46% of skill page dmg
With 1335 lure- skills did a maximum of 97.88-98.41% of skill page dmg

Clearly, no matter how high your lure is, you can't hit for more than 98.4 ish percent of your max dmg. I assume it works similarly for other bosses

With 0 lure- skills did a maximum of 75.55-75.61% of skill page dmg
With 1169 lure- skills did a maximum of 98.26-98.46% of skill page dmg
With 1335 lure- skills did a maximum of 97.88-98.41% of skill page dmg

Clearly, no matter how high your lure is, you can't hit for more than 98.4 ish percent of your max dmg. I assume it works similarly for other bosses

This would have been very valuable in your original comment. I apologize, assumed you were just disagreeing with stuff.

This info is helpful as well, so kudos for figuring it all out

^Don't mind him Swan

I love this, the science side of Celtic. Very well done and a good post! Clears up alot of questions

Keep up the quality posts!

Don't mind who?


For pve, mobs can hit 50-100% of dmg on you rather than 60-100 and the curves are much different from what you would find on pvp (these pve values would apply to tanks). I wanted to do the opposite as well, and find a curve relating boss resist to percent dmg taken. Similar to the method you described, I wanted to lure a mob down to 0 and from there slowly decrease the value of my lure until I stopped doing max dmg. From there, I'd further decrease the lure and find max/min at each 50 increment until I came up with a curve. Finally, I'd have to repeat on several bosses to see whether they followed the same formula and use my existing data to determine approximate pre-lure elemental resists of every endgame boss

I used blackstorm because it doesn't hit high and has a lot of hp. I found that bosses have some sort of built in resist, which doesn't allow you to lure to 0. I decided at that point it wasn't worth my effort to continue this project.

With 0 lure- skills did a maximum of 75.55-75.61% of skill page dmg
With 1169 lure- skills did a maximum of 98.26-98.46% of skill page dmg
With 1335 lure- skills did a maximum of 97.88-98.41% of skill page dmg

Clearly, no matter how high your lure is, you can't hit for more than 98.4 ish percent of your max dmg. I assume it works similarly for other bosses

Thanks for putting your findings up—it seems like you did a lot more testing than I did!

I did a quick test and also found that NPC auto/skill hits on 0 armored players ranged from 50-100%, so you're right—it seems like PvP actually BUFFS auto attack damage by making the baseline 60% instead of 50%.

But then I went up to 500 armor and it seemed like NPC hits got halved. Which implies that the curve shape is the same in PvP and PvE, except the lower curve adjusted down by 10%. Tentative results I know… but I don't see why OTM would go through the effort of making two separate resist mechanics in and out of PvP. Much simpler to just adjust numbers, ie 50 to 60.

For bosses, once again I think the curves are the same. An annoying problem I ran into while testing was that I could never quite reach 100% even at 0 armor in PvP. Then I realized my skill hits were too high. Think about it—if you hit a 4k firebolt or whatever (in my case, shadowstrike), then the full range is 2000-4000 damage. Meaning the chance that you'll hit above 99% (> 3960 damage) is 40/2000 or 2%. With autos its easy to do hundreds of tests to ensure you get max (or close enough to max), however with skills the most spammable skill is firebolt. 6 sec CD, 4.2 sec with void hrung skull. Best case scenario, two or three times as slow as a decent dagger on haste. So it could be that you didn't do enough trials to reach max?

I'd edit the OP but I don't have a whole lot of free time at the moment… lol maybe later

I did hundreds of trials with every resist. Here's my method:

record highest and lowest hits after the first 5 or so. Take the low hit and divide by .6- this gives the "theoretical max" (as in if that low hit you accomplished was in fact the lowest possible, the "theoretical max" would then be the highest you could get). I compared the theoretical max to the observed max and when they were close enough I was satisfied. When I stated values like 98.26, that would have been the observed max. And 98.46 being the theoretical max. So clearly my results are quite precise, and mathematically it would have been impossible to hit over 98.46 (because it was computed using my observed low).


When you found that 500 resist gave about 50% reduction, was this for auto attacks? Pretty odd- with loredancers I got to 50% somewhere between 100-200 resist

I did hundreds of trials with every resist. Here's my method:

record highest and lowest hits after the first 5 or so. Take the low hit and divide by .6- this gives the "theoretical max" (as in if that low hit you accomplished was in fact the lowest possible, the "theoretical max" would then be the highest you could get). I compared the theoretical max to the observed max and when they were close enough I was satisfied. When I stated values like 98.26, that would have been the observed max. And 98.46 being the theoretical max. So clearly my results are quite precise, and mathematically it would have been impossible to hit over 98.46 (because it was computed using my observed low).


When you found that 500 resist gave about 50% reduction, was this for auto attacks? Pretty odd- with loredancers I got to 50% somewhere between 100-200 resist

I just did the same test on loredancers and found similar results. Then I had a crazy theory that level might play a role in affecting NPC damage on myself (just like if player level is below NPC enemy level, damage is reduced).

On my 160 rogue I aggroed a level 166 2* firbolg shaman which casts 2 spells: frozen bolt and savage flame. Recorded lowest/highest hits in about 40-50 trials in total (quick, so somewhat inaccurate). No armor: 490-890 frozen bolt damage, 682-1268 savage flame damage. Roughly 500 heat/cold resist: 235-456 frozen bolt damage, 343-625 savage flame damage. Note the approx 50% reduction at around 500 resist.

Those tests were done quickly (since it's really late and I should be sleeping lol) but I think they indicate some sort of damage penalty when lower level NPCs hit much higher level players. Would love to see others prove/disprove this result.

I can see why you gave up on this project despite your very comprehensive testing… it seems like the damage curve is an enigma with far too many factors involved.

The relative levels do matter quite a lot. If you are even one level under the npc your auto and skill damages will be less, however the npc will do only normal damage and it does not increase with level difference. The same is true in reverse, if you are even one level over the npc you will take less damage from skills and autos but you won't do more than your max damage.

The most dramatic differences are within 5 levels and once you are far away the difference per level goes to zero. It's not clear but it does look to be exponential of some type. It is very hard to test without altering several variables at the same time.

Not to be the negative one here, but didn't OTM say we aren't allowed to calculate formulas? Papi got in trouble for that didn't he? Did they only mean boss and mob mechanics?