| Author |
Topic:
Weaponskill comparison test
|
GotharNW
collection | wishlist
Title: Nonchalant pragmatist
Posts:
25,800
Registered:
Feb '02
|
Date Posted:
2/4/06 6:17pm
Subject:
Weaponskill comparison test
|
I've heard a lot of people claim there was some hidden damage progression between classes outside of weaponskill. I've always assumed weaponskill was the only determinate, but there have always been dissenters. I'm not sure if this consensus is a majority or a minority, but I did a brief test that shows a comparison of weaponskill and base damage. I've only done a little over 1,000 samples and there appears to be still some discrepancies.
The comparison was done comparing a Blademaster and a Warden. Two classes near their respective ends of the weaponskill progression table.
The Data
Blademaster:
139 str
765 weaponskill
1650 damage
88 quickness
15+6 weapon
16.5 4.1spd 100 quality weapon
Warden:
130 str
764 weaponskill
1650 damage
88 quickness
34+13 weapon
16.5 4.1spd 100 quality weapon
Blademaster text file
Warden text file
Using Neill's log parser, I reached the following results:
Blademaster:
1076 attacks
55.4 avg. base dmg
Warden:
1082 attacks
56.7 avg. base dmg
Conclusion
As can be seen the weaponskill of both classes are not the same. The Blademaster's weaponskill is one point higher. I tried to balance them out exactly, but it's a little hard without the ability to create items like TLs can. What this test does prove is that there is no hidden damage outside of weaponskill. However, there is an obvious error in that the Blademaster had one more higher weaponskill, but failed to have equal or more base damage. The only conclusion I can make is a further test is needed of 5,000 or more attacks. Looking at the specs and data there is no other conclusion I can derive.
If we assume people are correct, and there is some hidden damage progression outside weaponskill, then the Blademaster should be ahead of the Warden despite the equal weaponskill. Unless, there is some secret damage progression outside of weaponskill and Wardens are at a higher progression.
Obviously, I dismiss the latter notion and assume the former in that the sample size of ~1,000 attacks was not sufficient data to formulate an exact result. Also, the difference may be assumed based on the different weapon specs, but one handed weapons do not derive a spec bonus like two handed weapons do.
Any further discussion or questions on the results (preferably on an explanation as to why they may be different) can hopefully be focused in this thread.
-----signature-----
"It is only when the people become ignorant
and corrupt, when they degenerate into a
populace, that they are incapable of
exercising their sovereignty."
--James Monroe
|
Locked Topic |
Active Topic Notification |
Private Message |
Post History
|
omSilver
collection | wishlist
Posts:
4,651
Registered:
Feb '04
|
Date Posted:
2/4/06 8:01pm
Subject:
RE: Weaponskill comparison test
|
|
although they have the same weaponskill, the blademaster's cap should be less than the warden's due to weapon spec, that may be why theres some varience, but it should show alot more than 2%.
-----signature-----
|
Locked Topic |
Active Topic Notification |
Private Message |
Post History
|
GotharNW
collection | wishlist
Title: Nonchalant pragmatist
Posts:
25,800
Registered:
Feb '02
|
Date Posted:
2/4/06 8:44pm
Subject:
RE: Weaponskill comparison test
|
omSilver posted:
although they have the same weaponskill, the blademaster's cap should be less than the warden's due to weapon spec, that may be why theres some varience, but it should show alot more than 2%.
I don't see how weapon spec would have any influence on the outcome. Weapon spec results in weaponskill which should result in overall damage. As far as I'm aware there is no benefit with one handed weapons and weapon spec.
-----signature-----
"It is only when the people become ignorant
and corrupt, when they degenerate into a
populace, that they are incapable of
exercising their sovereignty."
--James Monroe
|
Locked Topic |
Active Topic Notification |
Private Message |
Post History
|
Aeiryn
collection | wishlist
Posts:
3,527
Registered:
May '04
|
Date Posted:
2/5/06 12:02am
Subject:
RE: Weaponskill comparison test
|
... Except that below 51 (or 50 for below RR5) composite weaponspec the variance range would exist in a lower damage area.
Lets say the unstyled damage cap is 200 for that 4.1 weapon.
Let us also imagine a test target where someone with 51/51 weapon spec will JUST cap full damage on the #1 best roll of variance, so their range of hits would be from 140 to 200 (melee variance range is about 30% IIRC). Call him Bob 
15+6 = 21, which is 21/51 or about 20% of capped variance, so that blademaster is probably hitting at the 20% to 50% section of possible damage, or something like that. E.g. He'd be hitting magical test target Bob for hits between 40-100 damage.
The warden doesn't cap his variance either, getting 47/51 but it comes pretty close. So in their example they'd be at like 68% to 98% of that variance line, on magic Bob, a couple points shy of 200 on their best swings but they'd get there with a few more RR.
I'm not explaining this well, but do you see what I mean? Basically the test you did is flawed because the BM has lower average melee DPS induced because of lack of spec points, and that's only if the tests were unstyled swings. Obviously the style damage would be nerfed for the BM too having about 1/3 of the weaponspec (aka style bonus modifier) of the warden.
You'd have to match the exact specs of the test subjects and alter stats to balance WS for a fair investigation. Try buffing the warden and aug str but not on the BM, should get close that way. Use this calc to help estimate http://haldar.newmail.ru/daoc/ws
-----signature-----
Aeiryn-Guinevere, Acu-Gareth, Breirise-Percival, Garten-Ector http://gimpchimp.etilader.com/display.php?user=Brei
|
Locked Topic |
Active Topic Notification |
Private Message |
Post History
|
asilithiel
collection | wishlist
Posts:
1,997
Registered:
Jun '03
|
Date Posted:
2/5/06 9:17am
Subject:
RE: Weaponskill comparison test
- Date Edited:
2/5/06 5:01pm (1 edits total)
Edited By:
asilithiel
|
The current theory of DAOC melee damage that I operate from is this (assuming a pure-STR weapon):
1) Strength score scales the average damage I do.
2) Weapon spec improves my damage variance, and thus average damage over time.
3) "Weaponskill" from the character window is a composite of these two independent numbers, estimated by Mythic.
My theory for different base class types (tank vs. rogue) is that there's a simple linear scaling factor to the str-based damage result. Tanks have 1.x times the damage of rogues, warriors have 1.05 times the damage of other tanks, etc.
I think the test you're doing is interesting -- trying to get an equal UI weaponskill number on classes whose weaponskill component due to strength is different. I think your initial results suggest that average damage is not necessarily predictable based on that UI number.
The test I would be interested in seeing is constant str and weapon spec on both classes, at several levels of weapon spec. You can record the UI weaponskill, but I think it would be informative to see if they keep the same ratio of damage independent of what str, spec level you measure at. If they keep the same ratio at various spec levels, I think that supports the theory that tanks simply do 1.x times more damage than rogues, etc.
The other test I would be interested in is measuring different levels of weapon spec, with constant str, on both classes. If the damage goes up by the same percentage for the same increases in spec on both classes, that supports the idea that spec & strength are independent factors in damage calculation. It would help show that class type (warrior/tank/rogue) is one scaling factor, strength is another, and weapon spec is a third scaling factor.
Edit: Looking back, I realize I described the same test in two different ways. It would require keeping strength constant for both characters, and picking a few different spec levels at which to test each of them. You could get two sets of data from this -- one set of percentages for how much a particular spec level increases damage, and another set indicating what percent more damage the tank does than the rogue/hybrid at each level. My theory is that the percent gains from spec would be the same for both toons, and that the percent difference from tank to hybrid would be the same at all specs.
-----signature-----
Alb Tristan <Wraiths of Albion>
Asilithiel, Speakssoftly, Kegnor, Kesteral, Badgers, Amarynthia
Hib Galahad <Juggernaut>
Hulainn, Hansil, Cythrael
"Friars are in great need of a donkey. We need something to carry our ale."
|
Locked Topic |
Active Topic Notification |
Private Message |
Post History
|
Kelevandross
collection | wishlist
Posts:
1,990
Registered:
Aug '03
|
Date Posted:
2/5/06 4:31pm
Subject:
RE: Weaponskill comparison test
|
i've seen the BM hit for a range of 43-63 while the warden for 50-63.
from this i can only deduct the warden is a better meleer than the BM but since it is the opposite i can only say that wyrd spec theory about variance is true.(read above poster)
-----signature-----
Semel in anno licere insavire
|
Locked Topic |
Active Topic Notification |
Private Message |
Post History
|
Aeiryn
collection | wishlist
Posts:
3,527
Registered:
May '04
|
Date Posted:
2/5/06 4:55pm
Subject:
RE: Weaponskill comparison test
|
Asilithiel speaks my thoughts as well (and with a far more articulate tongue) 
-----signature-----
Aeiryn-Guinevere, Acu-Gareth, Breirise-Percival, Garten-Ector http://gimpchimp.etilader.com/display.php?user=Brei
|
Locked Topic |
Active Topic Notification |
Private Message |
Post History
|
silkensmooth
collection | wishlist
Posts:
1,166
Registered:
Aug '05
|
Date Posted:
2/5/06 11:12pm
Subject:
RE: Weaponskill comparison test
|
The reason the bm is doing less damage is because variance is based on spec. If you are specced less than 50% you will do 50 to 100% damage, if you are specced to 75% then your variance is 75% to 125% and if you have 50 modified then your variance is 100% to 150% damage.
The warden in this example has 47 modified weapon spec, and so he is specced at or above 75% and so his variance will be from 75% to 125%.
The blademaster has only 21 modified weapon spec,21 being less than 50% and so his variance will be from 50 to 100%.
50% to 100% < 75%to 125%
So for example if cap damage per swing was 100, the warden with 47 modified weapon spec would be able to hit for 75 to 125 points of damage, whereas the blademaster would only be able to hit for 50 to 100 points of damage with his 21 modified weapon spec.
-----signature-----
Deluvian rr9l3 Hunter Gareth
Leyeah 50 Enchanter Gareth/cancelled
Juliana 50 Mentalist Gareth/canceled
Torl rr7L1 Spearo Gareth// Icuras rr5L6 bard Gareth
Silken rr9L3 ranger Gareth
|
Locked Topic |
Active Topic Notification |
Private Message |
Post History
|
RSS Feeds