View Single Post
  #89  
Old 12-16-2025, 07:40 PM
DeathsSilkyMist DeathsSilkyMist is offline
Planar Protector

DeathsSilkyMist's Avatar

Join Date: Jan 2014
Posts: 8,207
Default

Quote:
Originally Posted by bcbrown [You must be logged in to view images. Log in or Register.]
This is much better work, good job! That's a good explanation and it matches my understanding of the EQEmu damage calculation. Unfortunately I think the p99 actual max damages are different enough that it's not especially useful for predicting p99 max damages. Here's a comparison for some of my parses:

Code:
target       strength  atk predicted actual
bloodmaw     239       40  69.49     74
bloodmaw     162       40  63.59     67
bloodmaw     196       40  66.20     71
bloodmaw     197       40  66.27     71
frost elite  184       95  75.4      72
frost elite  185       0   66        63
The 40atk values are from Greater Wolf Form, the 95 is that plus Strength of Nature and Swiftwind, using the wiki values. Used 238 archery skill for bloodmaw, 239 for the elites. Level 57 and 58 respectively. This is using a base weapon damage of 23 (huntsman + tolan) not 24, but that didn't make a big improvement. Here's a fully worked example for the first row in case I missed something:

1. Wrath = 238 Archery Skill + ((239 STR * 2) - 150) / 3 + 40 Total ATK = 387
2. RolledD20 = 20 / 10 = 2
3a. Base Bonus = (350 Wrath - 70 minusFactor) / 2 = 158.6
3b. Extra Percent = 100 + (158 Base Bonus * 1.0 Random Value) = 258
3c. DamageDone = (23 Weapon Damage * 2 RolledD20 * 258 ExtraPercent) / 100 = 118
118/2 + 10 = 69
The formula matches P99 DPS quite closely, which is the most important part for proving it out. All of the datasets from myself and others posted on these forums confirm this is the correct formula for damage.

The magic numbers probably need a bit of tweaking, but 4 damage off of the actual max isn't really an issue. The chances of getting a max hit is very low, so losing like 10 damage across a few max hits won't change anything DPS-wise.

Feel free to do some parsing to try and figure out where the magic numbers should be tweaked for P99. You should post your raw logs as well. Your messy data sets could be getting numbers from different tests, crits, etc.
Reply With Quote