Jump to content

Holy Flame damage bonus ...


  • Please log in to reply
50 replies to this topic

#21 fs_robotussin

fs_robotussin
  • Guests

Posted 17 February 2010 - 22:02

Everything in my experience and calculations. Has pointed exactly at #2 as well.

#22 fs_deljzc

fs_deljzc
  • Guests

Posted 18 February 2010 - 01:02

I am very confident in my opinion #3 is correct.

If any of the other formulas were true, there would be monsters I would NOT have one-hit and I would have gone through Imps every 500 stamina.

I know specifically when I was doing the Undead in the 390's somewhere I was using only DD and HF and the formula working was 6000 * 1.2 * 1.3 = 9360. No other "enhancements" were throwing off the numbers.

#23 BalianRW

BalianRW

    Veteran

  • New Members
  • PipPipPip
  • 755 posts

Posted 18 February 2010 - 05:48

Ok, so I gathered some data and unfortunately the results are inconclusive. I had calculated the "damage" for each of the 4 options giving both the Max and Min creature armor as listed on the Ultimate Fallen Sword Guide. This gave me 4 ranges (one for each option). My plan was to compare the damage as listed in the FSH combat log to those ranges to see where they fell. Unfortunately the randomness of the combat calculations spread the results all over the place, from below the minimum calculated for any of the four options (given max mob armor) to above the maximum of the four options (using the min mob armor).

I hunt ever other day so I'll have to gather some more data on Friday. I guess what I am going to have to do is check each creatures armor and see what kind of damage I do, for 10-15 creatures. That will hopefully give me a much tighter grouping.

If anyone has any other data that they have gathered I would be very interested in seeing it.

#24 fs_deljzc

fs_deljzc
  • Guests

Posted 18 February 2010 - 15:02

We should be able to compare #2 vs. #3 extremely quickly. The difference in those numbers is 3000 damage!!! I can't believe some of you actually use #2 for your calculations and waste all that stamina on CA and don't realize you're killing the undead monsters by +2000 damage every time.

I mean, you will know immediately on your first hit of an Undead Creature which formula is in the right ballpark.

#25 fs_robotussin

fs_robotussin
  • Guests

Posted 18 February 2010 - 15:06

i dont use CA, but i do use #2 and i rarely 2 hit

#26 fs_deljzc

fs_deljzc
  • Guests

Posted 18 February 2010 - 15:11

i dont use CA, but i do use #2 and i rarely 2 hit


Because you are overcalculation and use more base damage than you need. You should look and see how much overkill you are using in those instances (from combat, not helper).

Edit:

There is a big difference in the required damage (as listed in your Bio when you start hunting) between our formulas.

In your case robotussin, the last undead I found for you was level 618, Necral Horror. If we round to a required 1-hit damage of 15,500 damage to kill it (that's a different debate on how to calculate that number).

Formula #2 = 11,300 damage listed in bio before DD + HF level 150

Formula #3 = 10,350 damage listed in bio before DD + HF level 150

That's a pretty substantial difference in required damage.

#27 BalianRW

BalianRW

    Veteran

  • New Members
  • PipPipPip
  • 755 posts

Posted 18 February 2010 - 15:36

On my next hunt tomorrow night, I plan to do a more detailed analysis of each individual creature I combat, as opposed to analyze the type as a whole. I should get the chance to check both with and without HF (will be using CA when not HF).

#28 fs_robotussin

fs_robotussin
  • Guests

Posted 18 February 2010 - 16:04

im sitting in dagoresh right now surrounded by necro terrors... i will be back in a few with something solid for everyone to chew on

#29 fs_robotussin

fs_robotussin
  • Guests

Posted 18 February 2010 - 17:45

OK, HERE ARE SOME RESULTS:

https://spreadsheets... ... 0R0E&hl=en

AVERAGE RESULT WITH DD, CA175, HF150 WAS 23580 DAMAGE!

(ANY COMBATS WHERE PS/CH ACTIVATED WERE TOSSED FROM RESULTS)

REMEMBER THAT THIS CRITTER AVERAGES 1700 ARMOR AND THAT IS THE NUMBER I APPLY TO ALL FORMULAS. I WILL PLUG THESE NUMBERS INTO EACH OF THE FORMULAS THAT BalianRW POSTED ON PAGE 2 OF THIS THREAD (I WILL MAKE ADJUSTMENTS FOR MY BUFF LEVELS). MY DAMAGE IN BIO WAS 11231. WHEN YOU TAKE AWAY 1700, YOU ARE LEFT WITH 9531, WHICH DEVIATES BY JUST OVER 2% OF THE 9792 RESULT FROM MY TESTING (RIGHT IN LINE WITH THE COMBAT RNG).

HERE WE GO:

1) DD, CA, and HF %bonuses added then multiplied by damage
(11231)* (1 + 0.2 + 0.44 + 0.3) -1700= 20088

2) DD and CA % bonuses added then multiplied by damage then apply HF after armor
([11231 * (1 + 0.2 + 0.44)] - 1700) * 1.3 = 21735

3) DD, CA, and HF % bonuses multiplied sequentially
(11231 * 1.2 * 1.44 * 1.3)-1700 = 23530

4) DD and CA %bonuses multiplied sequentially then HF bonus applied after armor.
[(11231 * 1.2 * 1.44) - 1700] * 1.3 = 23020



WELL, MY RESULTS ARE INCONCLUSIVE... ALL OF THESE FALL IN RANGE OF THE RNG :(

#3 IS CLOSEST, BUT I THINK WE NEED A LARGER SAMPLE FROM A CRITTER WITH A MUCH HIGHER ARMOR QUOTIENT

PLEASE LET ME KNOW IF YOU FIND ANY DISCREPANCIES IN MY POST, I WILL EDIT!!!

#30 fs_tangtop

fs_tangtop
  • Guests

Posted 18 February 2010 - 18:28

Some options for you to test ...

Jiang Shi - level 553 - 3500 armor
Harbinger of Storms - level 538 - 3400 armor
Elven Shade (Champion) - level 428 - 5190 armor
Eternal Tormented - level 397 - 3530 armor
Dar GromSol City Guard - level 391 - 3380 armor
Semerkhet (Elite) - level 370 - 10150 armor
Dagoresh (Elite) - level 350 - 9670 armor
Mortugus (Elite) - level 185 - 4000 armor

#31 fs_deljzc

fs_deljzc
  • Guests

Posted 18 February 2010 - 19:12

OK, HERE ARE SOME RESULTS:

https://spreadsheets... ... 0R0E&hl=en

AVERAGE RESULT WITH DD, CA175, HF150 WAS 23580 DAMAGE!

(ANY COMBATS WHERE PS/CH ACTIVATED WERE TOSSED FROM RESULTS)

REMEMBER THAT THIS CRITTER AVERAGES 1700 ARMOR AND THAT IS THE NUMBER I APPLY TO ALL FORMULAS. I WILL PLUG THESE NUMBERS INTO EACH OF THE FORMULAS THAT BalianRW POSTED ON PAGE 2 OF THIS THREAD (I WILL MAKE ADJUSTMENTS FOR MY BUFF LEVELS). MY DAMAGE IN BIO WAS 11231. WHEN YOU TAKE AWAY 1700, YOU ARE LEFT WITH 9531, WHICH DEVIATES BY JUST OVER 2% OF THE 9792 RESULT FROM MY TESTING (RIGHT IN LINE WITH THE COMBAT RNG).

HERE WE GO:

1) DD, CA, and HF %bonuses added then multiplied by damage
(11231)* (1 + 0.2 + 0.44 + 0.3) -1700= 20088

2) DD and CA % bonuses added then multiplied by damage then apply HF after armor
([11231 * (1 + 0.2 + 0.44)] - 1700) * 1.3 = 21735

3) DD, CA, and HF % bonuses multiplied sequentially
(11231 * 1.2 * 1.44 * 1.3)-1700 = 23530

4) DD and CA %bonuses multiplied sequentially then HF bonus applied after armor.
[(11231 * 1.2 * 1.44) - 1700] * 1.3 = 23020



WELL, MY RESULTS ARE INCONCLUSIVE... ALL OF THESE FALL IN RANGE OF THE RNG :(

#3 IS CLOSEST, BUT I THINK WE NEED A LARGER SAMPLE FROM A CRITTER WITH A MUCH HIGHER ARMOR QUOTIENT

PLEASE LET ME KNOW IF YOU FIND ANY DISCREPANCIES IN MY POST, I WILL EDIT!!!



I think it actually confirmed #3 pretty strongly.

It certainly isn't #1 or #2. The numbers are too far away for that.

Thank you for the data though and taking the time to do it.

#32 fs_robotussin

fs_robotussin
  • Guests

Posted 18 February 2010 - 19:31

it doesnt confirm any of them, although i DO expect that #3 is correct...

i will repeat with a much higher armor creature when i have the time, and hopefully something will emerge as a firm conclusion

#33 BalianRW

BalianRW

    Veteran

  • New Members
  • PipPipPip
  • 755 posts

Posted 18 February 2010 - 20:19

Actually I believe that it at least confirms that the correct calculation is either #3 or #4. I have taken your numbers and expanded on them a bit. What I have added is a % Deviation calculations sections for each of the 4 Options (you will have to scroll down to see them all). Along with mean and standard deviation calculations. You can download my spreadsheet here.

http://balianrw.99k.... ... alysis.xls

First we will use Column C (damage with no buffs) as the control. From that we can see that the damage varies from the calculated expected by between -1.3% to +9.11% (In reality this top number is an outlier as it is almost 3% higher than the next highest value, but we will keep it anyway). Also, you will notice that the standard deviations for each column and each option type are all between 2.5% and 4%.

Of particular interest is Column G. That is the only one where both DD and CA was present. If we look at the results in that column we notice that for options 1 and 2 (where the % bonus for DD and CA are added together before being multiplied by the damage) every single damage was above (and in most cases well above) the calculation expectation. For Option 1 (column G) every single data point is outside the -1.3% to +9.11% Deviation range established by the control (column C). Additionally slightly over 1/2 of the results from Option 2 comparison are also outside that range. This seems to suggest that applying HF AFTER armor may be the correct calculation. But we will discuss that more below.

If we continue on down the spreadsheet to Options 3 and 4, we notice that for both options the Deviations are for the most part with in the range established by the control. We also, notice that for Option 3, several of the Deviations extend below the minimum established by the control, especially in the cases where HF was used. Similarly with options 1 and 2 above, the option where HF is applied after armor is subtracted (i.e. Option 4) seems to behave more like the control. Option 4 has less than 19% of it's entries outside the controls range (compared to 39%, 27% and 32% for Options 1, 2, & 3 respectively), and all of those are by a much smaller amount than many of those from the other 3 options.

From all of this, I have come to the conclusion that Option #4 is actually the correct calculation to use. Additionally it appears that the damage (due to random effects of the game) can range from between -4% (or maybe even -5%) to +10%.

A set of data taken against one of the Undead that has high armor could help further distinguish which of Option #3 and Option #4 are truly correct, but based on this data I am absolutely convinced that the effects of DD and CA are sequentially multiplied rather than having the bonus % added together first.

#34 fs_robotussin

fs_robotussin
  • Guests

Posted 18 February 2010 - 20:46

Wow, excellent use of my data. Another thing that i should have pointed out was that i have enhancement "holy" active at 50% from my guild structure. A guildmate ran tests for greenskin slayer (vs goblin in cave) and the % difference between GSS vs no GSS was ~4.85%-5.15%, so lets just assume that holy is the same (adding 2.5% damage). This explains the almost exactly 2.5% positive difference in my assumed pre-buff damage vs. my actual damage in combat. Hmmmm... interesting....

i wonder when that bonus gets applied... maybe someone without the holy shrine can tell us?


EDIT: the guildmate i mentioned posted right below me

#35 dowuones

dowuones

    Veteran

  • Members
  • PipPipPip
  • 835 posts

Posted 18 February 2010 - 21:37

i made the same test for HF & Holy (less accurated than for Greenskin), but all the results showed that Holy is applied when the combat starts (so 100% Holy + DD = profile damage * 1.25) and HF after armor check.

#36 fs_robotussin

fs_robotussin
  • Guests

Posted 18 February 2010 - 21:52

i made the same test for HF & Holy (less accurated than for Greenskin), but all the results showed that Holy is applied when the combat starts (so 100% Holy + DD = profile damage * 1.25) and HF after armor check.



that is exactly how i thought it worked (HF after armor check), but it seems to be applying the bonus to the entire damage stat, not just after armor... hmmmm

#37 BalianRW

BalianRW

    Veteran

  • New Members
  • PipPipPip
  • 755 posts

Posted 18 February 2010 - 22:03

Wow, I adjusted the calculations applying the 2.5% Holy bonus directly to the Damage first. Updated spreadsheet

http://balianrw.99k.... ... alysis.xls

That that fixed the skew in the Control so that it is now roughly +/- 5%. It's hard to tell with the bonus being so small in comparison with the DD bonus, but I think that it is sequentially multiplied with DD (i.e. Dmg * (1+Holy) * (1+DD)).

I also ran a test to see what if CA was applied after armor and while a case with DD and CA only would be helpful in determing this, it appears that CA is applied before armor just like DD.

Also, redoing these calculations with the Holy Enhancement factored in, I am convinced of the following:

1)Buff effects are sequentially multiplied rather than having their % bonuses added together (i.e. option 4).
2) Holy (and other Enhancements), DD, and CA are applied before armor
3) Holy Flame is applied after armor.
4) There is approximately a +/- 5% random variance from the calculation.

#38 fs_coyotik

fs_coyotik
  • Guests

Posted 18 February 2010 - 22:11

I'm quite sure that HF is applied to whatever is left from the total damage AFTER monster armor is substracted from that. It most certainly worked that way a long time ago, before DD and CA existed - and from programmers' perspective, it is very unlikely that Hoof would have changed that just because he added damage-increasing buffs.

#39 fs_tangtop

fs_tangtop
  • Guests

Posted 18 February 2010 - 22:16

I'm quite sure that HF is applied to whatever is left from the total damage AFTER monster armor is substracted from that. It most certainly worked that way a long time ago, before DD and CA existed - and from programmers' perspective, it is very unlikely that Hoof would have changed that just because he added damage-increasing buffs.

But remember there was a code change a while back (can't exactly remember the reason for it), but a side effect of the change (before it was fixed), is that DD and CA were not applied to the stat bonuses form berserk, rage, fury, etc. This has since been fixed, but given that something changed in the code then, we cannot guarantee that things still work as they used to. I like the testing that people are doing and would like to see more. I would love to have time to put into gathering stats too, but things are far too busy unfortunately.

#40 fs_coyotik

fs_coyotik
  • Guests

Posted 18 February 2010 - 22:23

This has since been fixed


And you wouldn't believe how much persistent nagging of support it cost me to get it fixed...

but given that something changed in the code then, we cannot guarantee that things still work as they used to.


That's true... the research above, however, doesn't seem to indicate any change (#3 and #4 were quite close to the measured number).


0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users

Font:
Arial | Calibri | Lucida Console | Verdana
 
Font Size:
9px | 10px | 11px | 12px | 10pt | 12pt
 
Color: