Latest Diablo 3 News
DiabloWiki Updates
Support the site! Become a Diablo: IncGamers PAL - Remove ads and more!

Request for +elemental% and +skill% clarification.

Discussion in 'Diablo 3 General Discussion' started by Krond, Mar 5, 2014.

  1. yovargas

    yovargas IncGamers Member

    Joined:
    Oct 12, 2008
    Messages:
    3,672
    Likes Received:
    2
    Trophy Points:
    171
    Yeah, but without that +elemental mod it screws you out of using a bunch of skills you should be able to use it with. The Barb in particular has a lot of "as weapon dmg" ones instead of "as Physical".
     
  2. HardRock

    HardRock IncGamers Member

    Joined:
    May 4, 2009
    Messages:
    4,343
    Likes Received:
    4
    Trophy Points:
    356
    I'm 100% sure. Elemental damage has been working like this since the original beta, during which a dev or CM confirmed that this is the intended behavior. Skills that don't have an innate element should inherit our weapon's. What's probably not intended is that ever since the elemental themed title backgrounds were introduced in 2.0 skills without an innate element were classified as Physical, which is kinda wrong and definitely confusing as hell.

    If you absolutely want Physical damage you should use a true Physical skill. The weapon's element simply extends your options.

    Physical damage is just another type of elemental damage as far as the game is concerned. For Doombringer itself, it's pretty cool when it spawns with +physical damage (displayed as simply +damage), since it synergizes with its other bonus, meaning that you can use the weapon to boost true Physical skills, but also ones that don't have an innate element. As for when the elemental damages don't line up like that...

    Yep, that's true. When Doombringer's elemental bonuses are mismatched you can only boost true Physical skills with it.

    I hope I cleared up things for you. If not then ask away. :)

    Glad to be of service. :)
     
  3. SnickerSnack

    SnickerSnack IncGamers Member

    Joined:
    Dec 23, 2007
    Messages:
    6,165
    Likes Received:
    10
    Trophy Points:
    468
    So, I have been using a sword (edit: lol, now that I read the rest of the thread......) on my barbarian that has +lightning damage on it and also +20% physical skill damage. This means that unless I use skills that say "as physical" somewhere, the weapon is.....let's say.....poorly rolled for my purposes?

    Edit: The item is this one: http://us.battle.net/d3/en/item/doombringer, which *can* get physical damage instead of lightning. It would seem that sensible rolling would dictate that the % bonus would match the type of damage rolled so that it always buffed its own damage.

    @hardrock: Do you think that it would be appropriate to bring that up in your thread on the officials forums? If it is closely related enough maybe I should?

    Edit: @Ivan: Maybe they should just have another color and say 'inherited from weapon?' Or, should they have no color and no designation, but have that explained somewhere as meaning that the damage type is inherited?
     
    Last edited: Mar 10, 2014
  4. HardRock

    HardRock IncGamers Member

    Joined:
    May 4, 2009
    Messages:
    4,343
    Likes Received:
    4
    Trophy Points:
    356
    It certainly seems so. Those skills will of course convert all of your damage to Physical, making the 20% bonus useful, regardless of what type of elemental damage the weapon may have.

    I considered this as well, I think it would be a good idea. Legendaries have enough randomness as it is for my taste, without stuff like this messing up the theme of weapons.

    I'm not sure though if my thread is the appropriate place to post this. We should probably open a thread in the general discussion part of the official forums, since this isn't a bug at all and it may not even be viewed as a problem, depending on how much randomness people prefer in their items.

    For the record, I'd prefer if the Elemental bonus would be more diverse, instead of the Elemental damage becoming more limited and if they would always come in the same flavor. This would make weapons like Doombringer useful for much more builds, but in any specific case always just with one group of skills.
     
    Last edited: Mar 10, 2014
  5. Tahani

    Tahani IncGamers Member

    Joined:
    Nov 8, 2011
    Messages:
    48
    Likes Received:
    0
    Trophy Points:
    5
  6. yovargas

    yovargas IncGamers Member

    Joined:
    Oct 12, 2008
    Messages:
    3,672
    Likes Received:
    2
    Trophy Points:
    171
    Wow, that's a great post. And I think it's different than what we've been saying here. They really should change some of the wording around to clarify this stuff now that what element you have matters a lot.
     
  7. Tahani

    Tahani IncGamers Member

    Joined:
    Nov 8, 2011
    Messages:
    48
    Likes Received:
    0
    Trophy Points:
    5
    I have no idea if the post is correct or if its intended behaviour, just found it somewhere and it was very easy for me to understand so I linked it here.
    The guy who originally made that post seemed to have done at least some testing so its probably more than just theory crafting too.
     
  8. yovargas

    yovargas IncGamers Member

    Joined:
    Oct 12, 2008
    Messages:
    3,672
    Likes Received:
    2
    Trophy Points:
    171
    While he provides no specific evidence, the fact that he lists specific exemptions to the rule (eg. Chakram, Rend) would certainly imply it's not just random speculation. Unless it's totally made up, he must surely have tested Chakram and found that it worked differently.
     
  9. Steven Hazani

    Steven Hazani IncGamers Member

    Joined:
    Jul 8, 2012
    Messages:
    2,311
    Likes Received:
    27
    Trophy Points:
    258
    Well, the game does display this stuff. Death animations with a fire weapon make enemies explode, cold makes them shatter, etc. If you are using an inherit skill, that skill will begin using the appropriate death animation.

    So really you can just put a weapon on and a skill on then use your eyes to determine what's happening.
     
  10. HardRock

    HardRock IncGamers Member

    Joined:
    May 4, 2009
    Messages:
    4,343
    Likes Received:
    4
    Trophy Points:
    356
    Retested this again, Druin is right. This means that Krond's and my original assumption were correct. In my previous test my mistake was using 35.45 as the Dexterity multiplier, instead of 36.45, which would have been correct of course with 3545 Dexterity. As a result of this miscalculation I got a lower range prediction than I should have and when in-game a damage number popped up that exceeded this range I assumed that the calculation wasn't correct.

    What did I learn from this? Never do research when I should be sleeping. :)

    This seems to be true as well. Previously I didn't do very precise tests about this with Multishot. I saw that my damage was higher with Elemental bonuses and assumed that it's because skills like this inherit our weapon's element completely. As it turns out the situation is more nuanced than that.

    So, there are two types of Physical skills:

    1) Converts all of our damage to Physical, boosted only by Physical bonuses.

    2) Keeps the original Elemental distribution of our weapons. The damage of these skills are increased by two Elemental bonuses when our weapon doesn't spawn with +Physical damage. The effectiveness of the increase will depend on the original Elemental distribution of the weapon. Skills like this can be easily identified, because our attacks with them will result in Elemental effects corresponding to our weapon's element. What complicates things even more is that if Druin is right (and I'm convinced that he is) then even skills that clearly state that they'll do damage "as Physical" can behave this way (see his Deadly Reach example).

    I think my bug report thread is still valid though, it's just that this issue is much more complicated than I thought. Blizzard better come up with a new and easy to understand description for Physical skills.
     
    Last edited: Mar 10, 2014
  11. Mage Slayer

    Mage Slayer IncGamers Member

    Joined:
    Jan 26, 2011
    Messages:
    541
    Likes Received:
    10
    Trophy Points:
    106
    So it actually does work the way I always thought it did, with the exception of the tooltips not always being worded accurately. They just need to:

    a) decide which skills are meant to be weapon damage and which purely physical and make sure the skill descriptions match.
    b) put in a new type of header for the weapon damage skills. (While a dynamically updating banner would be nice, I'm not gonna hold my breath waiting for that). Personally, I think "weapon damage" would be enough of a description - it does the damage type(s) of the weapon.
     
  12. Steven Hazani

    Steven Hazani IncGamers Member

    Joined:
    Jul 8, 2012
    Messages:
    2,311
    Likes Received:
    27
    Trophy Points:
    258
    I dunno, I kind of like a game that requires a little math and thought, without going full autist. Even if it is a bug, let it be a feature. It's not like old black damage where it clearly isn't working as it should. x% lightning damage does in fact boost lightning damage by x% from all sources.
     
  13. SnickerSnack

    SnickerSnack IncGamers Member

    Joined:
    Dec 23, 2007
    Messages:
    6,165
    Likes Received:
    10
    Trophy Points:
    468
    Thanks for the link. It doesn't really settle things for me since the post doesn't explain how it reaches its conclusion, so I can't verify them.



    @hardrock:

    I'm trying to test this out myself. I don't have screencap software, so I'm just whaling on enemies and watching for the biggest crit number that I see. Here is the calculation I'm using:

    [max damage]*[1+%elemental]*[1+%str]*[1+%chd]=[highest possible crit]

    I'm calculating physical damage separately from lightning damage and then adding them afterward. I'm testing this with a Doombringer with 20% extra damage for physical skills and some bracers with 12% extra damage for lightning skills (for the damage on DB). I'm using Frenzy: Berserker since it appears to be one of those skills that inherits the damage type of the weapon. I've tried several different assumptions on how the +%elemental damage is calculated/applied, and no matter how I try it, my calculated max crit damage is lower than what I see in-game.

    I'm critting for up to about 99.5k damage, but assuming that elemental bonuses are applied proportionally (as in the post linked above), the calculated highest crit damage is 78.3k. If I assume that all of the damage is getting a 20% (which is what you'd expect if it were all converted to physical), then the calculated max crit is 83k. Do you see anything I'm doing wrong?

    I've double-checked that the numbers I'm using for the calculation match those in the stat sheet in-game. Those could be wrong though...
     
    Last edited: Mar 11, 2014
  14. Steven Hazani

    Steven Hazani IncGamers Member

    Joined:
    Jul 8, 2012
    Messages:
    2,311
    Likes Received:
    27
    Trophy Points:
    258
    What happens when you assume all damage is being magnified by 1.2 x 1.12?
     
  15. SnickerSnack

    SnickerSnack IncGamers Member

    Joined:
    Dec 23, 2007
    Messages:
    6,165
    Likes Received:
    10
    Trophy Points:
    468
    The calculated max is still only 92,991. I've had one crit just over 100k after more testing.

    Here's a google docs spread sheet with my calculations. Editting is on for anyone. https://docs.google.com/spreadsheet/ccc?key=0AkM3PaSJmUGFdGV4cXVHb0UtYnpuNWp1RldTQkdEanc&usp=sharing No griefing please. :) Also, please don't edit the first two sets of cells. If you want to try some new calculation, please make a copy lower down and edit it there.

    Edit: Face, meet Palm. I forgot that frenzy does more than 100% damage.....

    More edit:

    Okay, so I've updated the google doc, but the numbers still don't look right. Some of them are high enough now, but they're too high since I haven't seen any crits that high after around 30 minutes of swinging only Frenzy.

    Assuming that all the damage is converted to physical gets pretty close. In that case, the model predicts a max crit of 99,634, but I've observed two crits that were higher (one 99,9xx and one 100,xxx). It's possible that the higher crit was on an elite and my 3% bonus damage to elites caused that. I may take that item off for future testing.
     
    Last edited: Mar 11, 2014
  16. SnickerSnack

    SnickerSnack IncGamers Member

    Joined:
    Dec 23, 2007
    Messages:
    6,165
    Likes Received:
    10
    Trophy Points:
    468
    Now that I've fixed a glaring oversight, making that assumption puts the max crit way too high.

    Funnily enough, assuming that the physical damage gets magnified by 1.2*1.12=1.344 and the lightning damage is only magnified by 1.2 gives a max that seems about right, but that really doesn't make any sense since then the lightning damage bonus would apply to the physical damage but not the lightning damage. Also, the physical damage bonus would apply to both in that case.





    Edit: More testing. I used a different setup with only physical damage on the weapon (echoing fury) and 10% bonus physical damage on the bracers. The result was the model gives a max crit of 145,808 but then I observed several crits higher than that and the highest was about 153k.

    Edit: More testing. I used another setup with no %elemental damage bonus, and I still got higher crits than predicted by the model. So, unless there are display errors somewhere, the formula

    highest crit possible = [max weapon damage]*[1+%strBonus]*[1+%CHD]

    is incorrect. I may be overlooking something (again), but I don't see what that might be.
     
    Last edited: Mar 11, 2014
  17. Steven Hazani

    Steven Hazani IncGamers Member

    Joined:
    Jul 8, 2012
    Messages:
    2,311
    Likes Received:
    27
    Trophy Points:
    258
    Really stupid question: Average damage jewelry?
     
  18. SnickerSnack

    SnickerSnack IncGamers Member

    Joined:
    Dec 23, 2007
    Messages:
    6,165
    Likes Received:
    10
    Trophy Points:
    468
    That's not so stupid. I did have one ring with damage on it. I'll recalculate. I assume that the new formula should be

    highest crit possible = [max weapon damage + max damage bonus]*[1+%strBonus]*[1+%CHD]

    Edit: Ah, that's better.

    case 1: calculated max: 100,349; observed max: ~100k
    case 2: calculated max: 154,521; observed max: ~153k
    case 3: calculated max: 140,393: observed max: ~140k

    Thanks for finding my mistakes.
     
  19. HardRock

    HardRock IncGamers Member

    Joined:
    May 4, 2009
    Messages:
    4,343
    Likes Received:
    4
    Trophy Points:
    356
    Watching for crits isn't really a good idea, since if you don't have at least 50% crit chance you are reducing the relevant data points. Still, in the end everything worked out OK. What I (and I assume Druin) did was exactly the same thing. Use calculations to predict the result and then observe the in-game damage numbers. It's the scientific method and if it worked for our modern world it should work for us. :)

    To make my job easier I did simplify things by ignoring crits and reducing my damage range as much as I can. However, I did verify minimum damage numbers as well. Depending on your gear it's possible that one end of the range will line up with your predictions, but the other not quite.

    I think when I verified Druin's results I used a 7-23 damage bow, which had 3-18 Physical and 4-5 Fire damage. I couldn't find a weapon with Elemental damage that had a lower damage and I don't think they exist. Having a 2-3 damage range like in my other test would have helped a lot.
     
    Last edited: Mar 11, 2014
  20. SnickerSnack

    SnickerSnack IncGamers Member

    Joined:
    Dec 23, 2007
    Messages:
    6,165
    Likes Received:
    10
    Trophy Points:
    468
    Oh, I was using crit chance over 50%. Doing otherwise would have been pretty inefficient. I also considered using not crits instead, but I found that watching white numbers was a bit harder since there were so many.
     

Share This Page