• Announcements

    • IMPORTANT - REACH US IN THE NEW FORUM   05/04/2017

      Ladies and gentlemen ATTENTION please:
      It's time to move into a new house!
        As previously announced, from now on IT WON'T BE POSSIBLE TO CREATE THREADS OR REPLY in the old forums. From now on the old forums will be readable only. If you need to move/copy/migrate any post/material from here, feel free to contact the staff in the new home. We’ll be waiting for you in the NEW Forums!

      https://community.blackdesertonline.com/index.php

      *New features and amazing tools are waiting for you, even more is yet to come in the future.. just like world exploration in BDO leads to new possibilities.
      So don't be afraid about changes, click the link above and follow us!
      Enjoy and see you on the other side!  
    • WICHTIG: Das Forum ist umgezogen!   05/04/2017

      Damen und Herren, wir bitten um Eure Aufmerksamkeit, es ist an der Zeit umzuziehen!
        Wie wir bereits angekündigt hatten, ist es ab sofort nicht mehr möglich, neue Diskussionen in diesem Forum zu starten. Um Euch Zeit zu geben, laufende Diskussionen abzuschließen, könnt Ihr noch für zwei Wochen in offenen Diskussionen antworten. Danach geht dieses Forum hier in den Ruhestand und das NEUE FORUM übernimmt vollständig.
      Das Forum hier bleibt allerdings erhalten und lesbar.   Neue und verbesserte Funktionen warten auf Euch im neuen Forum und wir arbeiten bereits an weiteren Erweiterungen.
      Wir sehen uns auf der anderen Seite!

      https://community.blackdesertonline.com/index.php Update:
      Wie angekündigt könen ab sofort in diesem Forum auch keine neuen Beiträge mehr veröffentlicht werden.
    • IMPORTANT: Le nouveau forum   05/04/2017

      Aventurières, aventuriers, votre attention s'il vous plaît, il est grand temps de déménager!
      Comme nous vous l'avons déjà annoncé précédemment, il n'est désormais plus possible de créer de nouveau sujet ni de répondre aux anciens sur ce bon vieux forum.
      Venez visiter le nouveau forum!
      https://community.blackdesertonline.com
      De nouvelles fonctionnalités ainsi que de nouveaux outils vous attendent dès à présent et d'autres arriveront prochainement! N'ayez pas peur du changement et rejoignez-nous! Amusez-vous bien et a bientôt dans notre nouveau chez nous
Sign in to follow this  
Followers 0

PVE Damage tests

24 posts in this topic

Posted (edited)

Did another damage test, very similar to the previous one I did: http://forum.blackdesertonline.com/index.php?/topic/123284-kzarka-vs-liverto-damage-test/

Note that there might be minor mistakes in the data, some of the sample sizes are small, and it's extremely hard to list exact character stats (I did my best to add up all the various bonuses from gear and guild).

Kzarka vs Liverto test 1: (same data as initial test in above link)

  • Character stats: Lvl57, ~136 AP, ~13 accuracy
  • Enemy type: Senior Soldier (lvl49)

TRI Kzarka - Lighting Strike
  Health bar counts: 80
  Health bar average: 39.9%
  Health bar lowest: 10.5%
  Health bar highest: 48.6%
  Complete misses: 0
  Total unique health bars: 23

TRI Liverto - Lighting Strike
  Health bar counts: 74
  Health bar average: 42.9%
  Health bar lowest: 15.7%
  Health bar highest: 50.8%
  Complete misses: 0
  Total unique health bars: 20

TRI Kzarka - Staff Attack
  Health bar counts: 41
  Health bar average: 89.9%
  Health bar lowest: 81.5%
  Health bar highest: 91.6%
  Complete misses: 0
  Total unique health bars: 5

TRI Liverto - Staff Attack
  Health bar counts: 48
  Health bar average: 90.7%
  Health bar lowest: 82.4%
  Health bar highest: 92.1%
  Complete misses: 0
  Total unique health bars: 5

Kzarka vs Liverto test 2:

  • Character stats: Lvl57, ~203 AP, ~13 accuracy
  • Enemy type: Senior Soldier (lvl49)

TRI Kzarka - Magic Arrow
  Health bar counts: 70
  Health bar average: 81.5%
  Health bar lowest: 73.2%
  Health bar highest: 88.1%
  Complete misses: 0
  Total unique health bars: 6

TRI Liverto - Magic Arrow
  Health bar counts: 55
  Health bar average: 81.5%
  Health bar lowest: 74.1%
  Health bar highest: 88.5%
  Complete misses: 0
  Total unique health bars: 6

Accuracy test 1:

  • Character stats: Lvl57, ~123 AP, ~5 accuracy (+ accuracy from TRI Kzarka)
  • Enemy type: Senior Soldier (lvl49)

+0 accuracy - Lightning Strike
  Health bar counts: 50
  Health bar average: 45.6%
  Health bar lowest: 22.8%
  Health bar highest: 51.3%
  Complete misses: 0
  Total unique health bars: 16

+10 accuracy - Lightning Strike
  Health bar counts: 49
  Health bar average: 44.7%
  Health bar lowest: 30.2%
  Health bar highest: 50.8%
  Complete misses: 0
  Total unique health bars: 16

+31 accuracy - Lightning Strike
  Health bar counts: 58
  Health bar average: 45.6%
  Health bar lowest: 17.5%
  Health bar highest: 51.3%
  Complete misses: 0
  Total unique health bars: 16

Accuracy test 2:

  • Character stats: Lvl57, ~181 AP, ~5 accuracy (+ accuracy from TRI Kzarka)
  • Enemy type: Basilisk Ambusher (lvl60)

+0 accuracy - Lightning Strike
  Health bar counts: 34
  Health bar average: 97.8%
  Health bar lowest: 95.1%
  Health bar highest: 96.9%
  Complete misses: 15
  Total unique health bars: 5

+10 accuracy - Lightning Strike
  Health bar counts: 23
  Health bar average: 98.2%
  Health bar lowest: 95.6%
  Health bar highest: 96.9%
  Complete misses: 13
  Total unique health bars: 3

+18 accuracy - Lightning Strike
  Health bar counts: 24
  Health bar average: 97.3%
  Health bar lowest: 94.2%
  Health bar highest: 96.9%
  Complete misses: 8
  Total unique health bars: 6

Bashim test:

  • Character stats: Lvl57, ~193 AP, ~13 accuracy
  • Enemy type: Kurd Reinforced Infantryman (lvl57)

TRI Kzarka - Lightning Strike
  Health bar counts: 76
  Health bar average: 91.6%
  Health bar lowest: 85.0%
  Health bar highest: 97.8%
  Complete misses: 0
  Total unique health bars: 10

TRI Kzarka + 10 accuracy - Lightning Strike
  Health bar counts: 78
  Health bar average: 90.7%
  Health bar lowest: 81.5%
  Health bar highest: 97.8%
  Complete misses: 0
  Total unique health bars: 11

TRI Kzarka + 10 AP - Lightning Strike
  Health bar counts: 80
  Health bar average: 91.2%
  Health bar lowest: 84.2%
  Health bar highest: 97.8%
  Complete misses: 0
  Total unique health bars: 11

TRI Liverto + 10 AP - Lightning Strike
  Health bar counts: 82
  Health bar average: 92.1%
  Health bar lowest: 85.0%
  Health bar highest: 97.8%
  Complete misses: 0
  Total unique health bars: 11

Raw data:  raw-data.rar

Not sure how useful this would be for people, but I primarily did this to get a general sense of the practical difference between Liverto and Kzarka when fighting monsters, and also what the general impact accuracy has against lower level enemies, same level enemies, and higher level enemies (Basilisks was a huge annoyance to test with so the sample count for those ended up very small).

Edit: Fixed the character stats for Bashim test (starting accuracy was around 13, not 5).

Edited by FluffyQuack
6 people like this

Share this post


Link to post
Share on other sites

Posted

Thanks for the good work, need a bit more info before I can draw any conclusions though.

What was your source of extra ap and extra accuracy in the bashim/basilisk tests?

 

Share this post


Link to post
Share on other sites

Posted

Thanks for the good work, need a bit more info before I can draw any conclusions though.

What was your source of extra ap and extra accuracy in the bashim/basilisk tests?

 

Bares/Kalis Necklace in Bashim tests. And a whole bunch of Kalis accessories in Basilisk tests.

Share this post


Link to post
Share on other sites

Posted

On second thoughts, let's wait for someone who knows the appropriate statistic test to take a look !

 

I'm assuming you didn't have Bheg's gloves so the +5 (base) came from  precision crystals or something?

Share this post


Link to post
Share on other sites

Posted

First of all,  no matter what your test conclueds, you fail to take in account the synergy between kzarka and full boss armor set.

Also bear in mind that kzarka will recieve a bonus ap increase.  (At least according to the kr rumor).

Also the way to test accuracy is in pvp, not pve (at least on mobs you can kill with a stuck found on the ground) 

All in all, gj for testing neverthelesss.

Share this post


Link to post
Share on other sites

Posted

On second thoughts, let's wait for someone who knows the appropriate statistic test to take a look !

 

I'm assuming you didn't have Bheg's gloves so the +5 (base) came from  precision crystals or something?

Any of the accuracy are the results of this: +5 from guild, various accessories, and whatever amount you get from Kzarka and Liverto.

Share this post


Link to post
Share on other sites

Posted

First of all,  no matter what your test conclueds, you fail to take in account the synergy between kzarka and full boss armor set.

Also bear in mind that kzarka will recieve a bonus ap increase.  (At least according to the kr rumor).

Also the way to test accuracy is in pvp, not pve (at least on mobs you can kill with a stuck found on the ground) 

All in all, gj for testing neverthelesss.

liverto has better synergy with boss set. you only need 2 attack speed gems in gloves to cap attack speed. with kzarka you are left with an awkward 1 attack speed you need. i guess that is trueif you always have an attack speed food running though.

Share this post


Link to post
Share on other sites

Posted

liverto has better synergy with boss set. you only need 2 attack speed gems in gloves to cap attack speed. with kzarka you are left with an awkward 1 attack speed you need. i guess that is trueif you always have an attack speed food running though.

you might as well go zarka+boss+ viper in gloves http://bddatabase.net/us/item/15630/ since some if not most classes dont need the 5crit 

Share this post


Link to post
Share on other sites

Posted

You don't need to cap with gems if you can cap with buff food.

Share this post


Link to post
Share on other sites

Posted (edited)

Where is the Liverto version test for Basilisks ? And why do you spend time to perform an accuracy test on gray mobs ?

Edited by Kargor

Share this post


Link to post
Share on other sites

Posted

Where is the Liverto version test for Basilisks ? And why do you spend time to perform an accuracy test on gray mobs ?

I spend little time on Basilisks because they were an annoyance to test with. I've got the smallest sample size with them, so the numbers I got from those is the least reliable.

And I wanted to see if accuracy has any meaningful impact when fighting enemies of lower level. Although, the results from that test are so weird I'm not sure what to make of it. The most interesting and useful test by far was against the Bashim.

Share this post


Link to post
Share on other sites

Posted

I spend little time on Basilisks because they were an annoyance to test with. I've got the smallest sample size with them, so the numbers I got from those is the least reliable.

And I wanted to see if accuracy has any meaningful impact when fighting enemies of lower level. Although, the results from that test are so weird I'm not sure what to make of it. The most interesting and useful test by far was against the Bashim.

The "major" DPS boost you get using Kzarka is because of the accuracy. As the Liverto data is missing it is hard to make a comparison.

Share this post


Link to post
Share on other sites

Posted (edited)

The "major" DPS boost you get using Kzarka is because of the accuracy. As the Liverto data is missing it is hard to make a comparison.

I think this is true. If you cast lightning strike, it has 3 hits, but did you account for all 3 hits? Missing 1 might look like all 3 landed, but you do obviously will do less damage. If you look at your single hit %s (magic arrow/staff attack), they are the same. When you move to multi hit, the difference is because you miss some of the 3 hits on the liverto but not on the kzarka. The difference is small because you have a high hit rate (0 complete misses), but if it was on a harder mob, the difference would be greater.

Edited by Azveki

Share this post


Link to post
Share on other sites

Posted

 

Also bear in mind that kzarka will recieve a bonus ap increase.  (At least according to the kr rumor).

 

That's a fake rumor there's no info in korea about that , that won't be happening 100%

Share this post


Link to post
Share on other sites

Posted

The difference in damage is around 1%, and its not hidden damage its from the accuracy.  Which you can offset with other gear anyways.

You should do this on two classes that have bad accuracy.  Witch has some of the best passive skill accuracy in the game.

If you really want to see Kzarka in action use it on a warrior or valk

Share this post


Link to post
Share on other sites

Posted (edited)

  I'd like to take this opportunity to express my distaste for hidden stats. Not being able to make an informed choice in items and gems in an ARPG is beyond asinine. It'll never happen but I'll hope anyway that it will change. >:(

Edited by Kung__Lao
because

Share this post


Link to post
Share on other sites

Posted

Nice work again, Fluffy. I'll crunch the data tomorrow while at work to look for significant findings, and attempt to estimate the damage bonus of the accuracy increase.

2 people like this

Share this post


Link to post
Share on other sites

Posted (edited)

I spend little time on Basilisks because they were an annoyance to test with. I've got the smallest sample size with them, so the numbers I got from those is the least reliable.

And I wanted to see if accuracy has any meaningful impact when fighting enemies of lower level. Although, the results from that test are so weird I'm not sure what to make of it. The most interesting and useful test by far was against the Bashim.

Initial results: testing the effects of accuracy on grey enemies. The null hypothesis is that accuracy has no effect on damage.

 t-tests used for comparing 0 accuracy with +10 accuracy, and 0 accuracy with +31 accuracy. No ANOVA performed as the samples did not have normal variance.

0 vs +10 accuracy: t-statistic: .320; p value: .751. p> 0.05, thus null hypothesis is accepted: +10 accuracy has no effect on damage against grey enemies.

0 vs +31 accuracy: t-statistic: -.470; p-value: .641. p. 0.05, thus null hypothesis also accepted: +31 accuracy has no effect on damage against grey enemies.

 

Rather unfortunately, the basilisk test also showed no significance, either as a t-test or a non-parametric test - likely due to the sample size (or at least, that's my hypothesis: these tests require large variation when the sample size is small, and I can well see how doing these tests on Basilisks would be a pita).

[edit]: To get reliable data on the Basilisks, given the small differences between the +0 accuracy and the +18 accuracy, we'd need a minimum of 60 tests of each accuracy level. That's a real pita....

I'll get on to the Bashims a bit later.

 

 

hypo.png

Basims results:

testing effects of extra accuracy: t-test on Kzarka vs Kzarka +10 accuracy:

t-statistic: -2.776. p-value: 0.00301. the result is significant at p<0.05, which it is, so the +10 accuracy noticably increases average damage. the average damage increase of the +10 accuracy against mobs of the same level was 12.6%.

 

testing effects of extra ap: t-test on Kzarka vs Kzarka +10 AP:

t-statistic: -1.142. p-value: 0.1276. the result is NOT significant, as p>0.05.

Thus: for PVE on similar level mobs, +10 accuracy does more damage than +10 AP.

 

testing effect of Liverto vs Kzarka:

Kz 0 vs Liv +10 AP: t-statistic: -0.879. p-value: 0.19034. the result is NOT significant as p>0.05. the average damage of the Liverto is lower than the average dam of the Kzarka, but not to any significant value.

Kz +10acc vs Liv +10AP: t-statistic: -3.358. p-value: 0.0005. the result is significant as p<0.05. This result is expected given the result of the Kzarka +0 acc vs Kzarka +10 acc test. the average damage increase of the +10 accuracy Kzarka over the Liverto +10 AP was 18% (17.97%).

Kz +10AP vs Liv +10AP: t-statistic: -1.878. p-value: 0.03108. the result is significant as p<0.05. the average damage increase of the +10 accuracy Kzarka over the +10 accuracy Liverto was 10.7%, which similar to the earlier test that gave a TRI Kzarka a 11.2% damage increase over a TRI Liverto when using lightning attack.

Edited by seanas
8 people like this

Share this post


Link to post
Share on other sites

Posted

liverto has better synergy with boss set. you only need 2 attack speed gems in gloves to cap attack speed. with kzarka you are left with an awkward 1 attack speed you need. i guess that is trueif you always have an attack speed food running though.

The same cannot be said for casting classes, though. 

Share this post


Link to post
Share on other sites

Posted

The same cannot be said for casting classes, though. 

Honestly, both have a place, and which one is most efficient depends on how much you value other stats like evasion, or critical rate. 

When it comes to optimizing, your gems for either look something like

Assuming you value critical rate:
Kzarka + Full Boss
Cast speed: 

Gloves: 2 BMC Valor
Weapon: 1 BMC Addis, 1 RBF Power/BMC Precision
5CS, 5CR, 2AP + (2ACC+10%CC, or 5AP+2HD) with room for 10 evasion in helmet

Attack Speed: 
Gloves: 2 BMC Valor
Weapon: 1 BMC Carmae, 1 RBF Power/ BMC Precision
5AS+5CR+2AP+(2ACC+10%CC or 5AP+2HD) with room for 10 evasion in helmet

Liverto + Full Boss
Cast speed:
Helmet: 2 BMC Memory
Gloves: 1 BMC Valor, 1 BMC Precision
Weapon: 2 RBF Power/ BMC Precision
5CS+5CR+2ACC+(4ACC+20%CC or 10AP+4HD) with NO room for 10 evasion

If you don't value critical rate (Warrior/Valkyrie):
Kzarka + Full Boss
Attack/Cast Speed:
Gloves: 1 BMC Viper, 1 BMC Precision
Weapon: 2 RBF Power/BMC Precision
5CS+4ACC+(4ACC+20%CC, or 10AP+4HD) with room for evasion

Share this post


Link to post
Share on other sites

Posted

@seanas very nice test man! I'm interested in knowing whether or not there's diminishing returns on stacking accuracy, was that in some way implied in your test? If not, what's the best way to test it?

Share this post


Link to post
Share on other sites

Posted

@seanas very nice test man! I'm interested in knowing whether or not there's diminishing returns on stacking accuracy, was that in some way implied in your test? If not, what's the best way to test it?

There's a test on reddit here: https://www.reddit.com/r/blackdesertonline/comments/5oxtk2/accuracy_kzarka_vs_liverto_kutum_tests/ that shows the effects of stacking accuracy in PVP: +1 accuracy = +1% hit rate. The test Fluffyquack did above doesn't test (and can't test) hit rate, but it can imply increase in hit rate by increase in damage.

So for PVE, what you could do, using the method above, is test a setup with baseline gear; +10 accuracy to that gear, and +20 accuracy to that gear - you'd get a trend line from that; if you kept adding accuracy and testing, you'd find a point where the increased damage from the increase accuracy started to fall below the trend line: at that point, you'd be close to/ at the point of max hit rate (at which point, more accuracy is not going to increase damage). The Bashim Kzarka vs Kzarka +10 accuracy test showed a 12.6% damage increase for the +10 accuracy: so you'd expect around +25% damage increase with +20 accuracy *if* you hadn't reached the max hit rate. A damage increase significantly less than that would imply you've reached the max hit rate.

However, that point of diminishing returns is going to be different for each target you hit, depending on their DP, how much of their DP is Evasion (assuming that's a variable), and how much AP you give up to get the extra accuracy (remilafo made a similar point a few days ago in a different thread).

1 person likes this

Share this post


Link to post
Share on other sites

Posted

The Bashim Kzarka vs Kzarka +10 accuracy test showed a 12.6% damage increase for the +10 accuracy: so you'd expect around +25% damage increase with +20 accuracy *if* you hadn't reached the max hit rate. A damage increase significantly less than that would imply you've reached the max hit rate.

Actually you would expect a bit less than a +25% damage increase. Linear accuracy bonuses imply decreasing damage bonuses. Imagine that your base accuracy is 10%, adding another 10% would double your damage. Adding a further 10% would only increase your damage by a third, and so on. +10 accuracy increasing damage by 12.5% suggests that base accuracy was around 80% (assuming +1 acc = +1% chance to hit, of course).

Share this post


Link to post
Share on other sites

Posted

Actually you would expect a bit less than a +25% damage increase. Linear accuracy bonuses imply decreasing damage bonuses. Imagine that your base accuracy is 10%, adding another 10% would double your damage. Adding a further 10% would only increase your damage by a third, and so on. +10 accuracy increasing damage by 12.5% suggests that base accuracy was around 80% (assuming +1 acc = +1% chance to hit, of course).

that's... a really good point, actually. That would imply that +20 accuracy from a TRI Kzarka was functionally 100% hit rate (or asymptotically approaching it) for fighing against on-level mobs at lvl 57 (and probably a close enough approximation for most PVE purposes).

So that would be: TRI Kzarka + Bhegs (10 acc) +10 more accuracy for effective max hit rate; or TRI Liverto + Bhegs +15 more accuracy if using Liverto.

I wonder if that's a good enough rule of thumb for all mobs? or all lvl 56+ mobs?

Share this post


Link to post
Share on other sites
Sign in to follow this  
Followers 0