🛈 Announcement
Greetings! Assassin's Creed forums are now archived and accessible in read-only mode, please go to the new platform to discuss the game.
  1. #1

    The truth about elemental damage calculation

    I have made another test session about elemental damage for warrior last night, because it bothered me that there was no final agreed conclusion on that topic.

    I found out that the elemental calculation is using elemental base as impact damage as I predicted, but since damage over time 4600 (high poison base) and 4900 (fire base) are very close it is not certain to predict whether it is the same or a different base for the impact damage

    The calculation for elemental impact damage for warriors (fire) goes as follows:

    [ ((2490+WeaponDPS) * (1+Warrior)) + (4900 * (0+Elemental) ] * 1+Damage per weapon

    You notice that elemental impact damage uses elemental base damage, it looks to be the same value as for damage over time. But elemental impact damage does not start at 100%, it starts at 0% (0+Elemental) And instead of what everybody thought, it gets into the damage per weapon multiplier, which emphasizes it so that it might look like elemental impact also starts at 100% and does not use the weapon multiplier, with certain combinations of weapon damage and elemental damage. But it starts at 0% explaining zero damage increase at 0% elemental and gets all the multipliers aswell.

    It predicted the damage increase correct in multiple tests with fire and poison.

    I will also make a video about it on my channel very soon.
     1 people found this helpful
    Share this post

  2. #2
    I recently ran a test in an attempt to answer this exact question. Testing conditions were 230% warrior damage, 230.2% poison damage, 0% weapon damage, 525% crit damage, 100% crit chance, and weapon equipped in the right hand i.e. base of 2470. All hits were recorded with a spear using a light attack from the front on lions while at full health. Here are the observed data values with and without poison damage applied as well as the average damage of each:

    Normal Poisoned
    150,250 234,131
    144,218 246,231
    147,281 246,118
    146,662 254,556
    147,531 248,550
    140,018 234,725
    150,725 244,887
    144,075 253,606
    137,850 243,818
    146,887 238,850
    137,768 238,643
    141,800 248,037
    144,137 238,143
    139,500 242,950
    147,956 238,543
    140,906 241,825
    149,812 254,568
    145,318 236,712
    142,843 247,243
    149,137 243,725
    147,906 249,431
    138,293 243,687
    145,700 242,281
    144,275 247,481
    141,106 239,000
    148,750 249,581
    147,881 245,212
    140,418 257,156
    143,875 250,950
    143,806 253,650
    145,462 252,556
    145,281 254,225
    141,468 246,543
    146,418 250,893
    148,287 246,362
    139,875 243,550
    143,093 239,212
    138,118 239,025
    138,293 241,862
    143,093 246,931
    141,331 238,581
    147,393 257,537
    145,387 253,843
    143,481 257,693
    147,825 256,506
    138,300 244,950
    144,356 251,193
    139,606 257,968
    148,331 258,306
    147,231 239,550
    150,481 239,187
    150,475 254,968
    149,593 255,193
    145,806 236,275
    141,243 243,268
    150,612
    148,500
    146,806
    140,118
    =====================
    144,626 246,554 Average

    Given the hypothesized formula above we would expect a percentage damage increase of:
    Code:
    (4600*2.302) / [(2470+7829) * (1+2.30%)] = 31.16%
    The observed percentage increase in damage on average is
    Code:
    (246,554 - 144,626) / 144,626 = 70.48%
    At over double the predicted increase we can safely reject the hypothesis that elemental base is 4600 for poison.

    What should it be then? Well we see an increase of 70.48%/ 31.16% = 226.2% more than predicted so if we take the poison base predicted and multiply by that factor we get 4600 * 226.2% = 10,405 Remarkably that's ~1% off from the warrior base which is within the margin for error on a sample of this size. It certainly looks like, from this data set at least, that elemental base is the same as warrior base and it can't be damage per weapon that's causing the discrepancy because the test was done with 0% damage with weapon type.

    Prediction: the planned video that demonstrates the accuracy of the formula will be done under conditions such as the following:
    • ~127% damage with weapon type using poison damage and 0% armor penetration epic weapon ~133% if legendary
    • ~110% damage with weapon type using fire damage and 0% armor penetration epic weapon ~120% if legendary
    • ~134%.damage with weapon type using fire damage and 10% armor penetration on a target with approximately 10% damage reduction from armor 143% if legendary

    These are the damage with weapon type values that would yield roughly the correct results using the formula hypothesized above in spite of the underestimated poison and fire base damage.

    An alternative formula I would like to propose:
    [(Base Damage + Weapon DPS) * (attacks per second) * (1+Warrior Damage) * (1+Damage with weapon) * (1 - (enemy resistance - armor penetration))
    +(Base Damage + Weapon DPS) * (attacks per second) * (Poison/Fire % damage) * (1 - enemy resistance) ]
    * (1 + crit damage)

    Using this formula and with the stats from my test above we could have:
    [(2470 + 7,829) / (4/3 APS with spear) * (1+230%) * (1+0%) * (1- (10% - 0%))
    +(2470 + 7,829) / (4/3 APS with spear) * (230%) * (1 - 10%) ]
    * (1 + 525%)
    Yielding a predicted average damage of 243,313.88 vs the 246,554 observed a difference of ~1% with a standard deviation of 2.78%. For non poison the results are similar predicted 143,381.39 vs 144,626 observed a difference of ~1% with a standard deviation of 2.67%.
     2 people found this helpful
    Share this post

  3. #3
    Just wanted to say, PDM, kudos for making an actual post this time rather than a thread to promote your youtube channel. See you also added time stamps to the last video you posted too which is thoughtful.
    Share this post

  4. #4
    @AnimusLover
    Thx

    @Tetracycloide.
    I post a formula, and you do not use it correct. You use it on poison when I posted it for fire. And say its wrong. That is not how you can convince people of anything. I could simply say the formula is correct for fire, would that prove yours wrong? It does not.

    In fact we agree already on the fact that elemental starts at 0%.

    Here is a simplified example between your calculation and my recent proposal:

    (120% Fire and 80% Weapon)

    My formula:
    (10500 + 4900 * 1,2) * 1,8 = 29,5k

    Your formula:
    10500 * 1,8 +10500 * 1,2 = 31,5k


    Another example:

    (40% Fire and 30% Weapon)

    My formula:
    (10500 + 4900 * 0,4) * 1,3 = 16,2k

    Your formula:
    10500 *1,3 + 10500 * 0,4 = 17,8k


    Its actually so close that it is inside the damage spread and need further testing. Both of them could be correct.

    So if you would use other peoples suggestions correctly and pay attention, we might get into a discussion here. But you are already being arrogant and insulting.

    You are hacking your game to prove something.

    Even if you might be correct, it certainly does not help your credibility.
     1 people found this helpful
    Share this post

  5. #5
    Where did I use your formula incorrectly? If I made a mistake please point it out. I reviewed my post and your formula again and I cannot find any mistakes. You say I used poison when you posted fire but I didn't use the fire base I used the poison base so I'm confused what I got wrong there.

    I completely agree that for the base stats used in your examples it would be hard to tell which formula is better predicting the results but there are base stat values that yield very different outcomes like the one I used in my testing (230.2% poison 230% warrior 0% weapon):

    Your formula:
    (10500 * 3.3 + 4600 * 2.3) * 1 = 45.2k
    Proposed alternative:
    10500 * 3.3 + 10500 * 2.3 = 60.9k

    I am sorry I came off as arrogant and insulting as that was not my intention. I do not feel like anything in my post insulted or attacked you in any way and I am very open to being wrong about something in fact I'd be surprised if I have everything exactly right even after doing a fair amount of testing. I have not hacked anything to gather this data, what would be the point as the tests would not be reproduceable in that case. All tests were done with ordinary vanilla items found in game with no modifications. I'm just trying to improve my own understanding of the game's mechanics and pass the knowledge along to others. If you would like to do that same then please review the data I gathered and let me know where I made any mistakes so I can correct them.
     1 people found this helpful
    Share this post

  6. #6
    @PDM
    @tetracycloide

    First of all, sorry for resurrecting this thread, but I figure this quote should be heeded:

    The aim of argument, or of discussion, should not be victory, but progress.
    ―Joseph Joubert (1754–1824), Pensées de J. Joubert (1848)
    Share this post