Glancing Blow Formula - New Findings

[VeryRare]

Some research indicating that the current Glancing Blow formula is incorrect has been found.

First source: https://web.archive.org/web/20070105001816/http://evilempireguild.org:80/guides/glancing.php

2nd source which confirms the same table: https://web.archive.org/web/20060516141644/http://forums.worldofwarcraft.com/thread.aspx?FN=wow-warrior&T=942763&P=1

lvl 63 Mobs have -30% damage against level 60 players and glance 40% of the time. +0 skill = -30% damage +1 skill = -27% damage +2 skill = -24% damage +3 skill = -21% damage +4 skill = -18% damage +5 skill = -15% damage +6 skill = -12% damage +7 skill = -9% damage +8 skill = -6% damage +9 skill = -3% damage +10 skill = -0% damage

3rd source: http://www.lurkerlounge.com/forums/printthread.php?tid=4500

Note that with +0 skill, glancing blows were doing about 2/3 normal damage to Onyxia, whereas with +10, there was no damage reduction at all.

+0 wep skill:

+5 wep skill

+10 wep skill

4th source: http://leftoversraiding.org/index.php/topic/839/1/glancing-blows-and-you.html

If you are fighting a level 63 boss, once your Weapon skill reaches 310, your glancing blows now do 100% damage (although they still cannot crit). This appears to be a linear, non-tiered scaling, so each point of Weapon skill raises your glancing blow reduction by 3% (a human with a sword has +5 Swords, making his glancing blows do 85% damage; if he equips Maladath with +4 Swords, he now has +9 Swords and his glancing blows do 97% damage).

This means the current cap of +15 wep skill and the non-linear formula are not blizzlike. They are using the TBC mangos formulas.

[GuybrushGit]

Really weird data, his glance chance is 70% when the 40% cap is pretty well documented, maybe he's not counting misses / crits / etc. Plus his sample size is tiny, 150-170 hits.
The formula used on this core is not from TBC, its based on Athene's research in July 2006 with a 3k - 9k sample size: http://wowwiki.wikia.com/wiki/Talk:Weapon_skill

Really hard to tell if the diferences are due to being done in a different patch, one of them using wrong methodologies or just due to the small sample size.

[Scyllaqt]

@GuybrushGit there's definitely a small sample size in the screenshots. If you read the Lurker Lounge thread he says each screenshot is from one Onyxia kill with three different members of the raid wearing varying amounts of wep skill. That being said, there's a strange amount of support for the linear formula from the sources above as well what was on wowwiki at the time. If you go to the bottom of the wowwiki revision there are a few sources listed, all which seem to be inaccessible via archive.org unfortunately. I wouldn't be surprised if the way glancing blows and wep skill worked was changed towards the very end of vanilla. All of the other sources do say there's a 40% chance to glance though, unsure why the one test had 70.

[VeryRare]

I don't see any actual data for testing weapon skill glancing blow damage reduction from the source you linked.

Athan's 3k-9k swing tests are all using 300 weapon skill.

It's also littered with conflicting statements by Athan, for example:

My data suggest (NPCLvl - YourLvl) * 10 as the %age reduction, approx., but that's really not a very good fit.

Would be 30% reduced damage for glancing blows vs a 63 mob.

Another user has a chart with +wep skil and it's affect on glancing blows but I don't see any actual data for it. Within the chart itself is conflicting formulas.

Additionally if you check the actual article from Nov 2006, you will see that it's using 310 wep skill as a cap: http://wowwiki.wikia.com/Weapon_skill?oldid=326000

I found even more sources with +10 wep skill/100% glancing blow: 5th source: http://leftoversraiding.org/index.php/topic/839/1/glancing-blows-and-you.html 6th source: http://wow.allakhazam.com/forum.html?forum=243&mid=1154446920253254922 7th source: http://web.archive.org/web/20061115223930/http://forums.wow-europe.com/thread.html?topicId=14381707&sid=1

For those that are interested, yes I have personally tested most of this. No, not as extensively as I would like. The kudos for most of the work behind this goes to others, mainly in the US who have more thoroughly tested this than I and published their results. They have my thanks, as I all I have done is compile and render unto formula. Part of that is that it's a work in prgoress. I am 100% certain of the mecahnics versus equal or lower level mobs. The glancing mechanic I am certain produces 40% glancing blows with 30% mitigation for 0 +WS against a level 63 mob and that +!0 Ws reduces that to 0% mitigation

Classic WoW chose to use the 310 weapon skill cap, you can see them discussing it here (in German): https://community.mmobase.de/forum/thread/6947-waffenwertung-skaliert-zu-hoch/?pageNo=1 https://github.com/ClassicWoW/Nefarian_1.12.1_Bugtracker/issues/17

After some bug reports they chose to stick with the 310 wep skill cap, as it has the most evidence.

[VeryRare]

Okay I went over the wowwiki and realized that in a way he did test 300, 305 and 310 weapon skill against level 63 targets.

300 wep skill vs a 61 elite should be the same results as 310 wep skill vs a 63 elite. 300 wep skill vs a 62 elite should be the same results as 305 wep skill vs a 63 elite. 300 wep skill vs 63 elite

So if we take the results from the first test which is the same as 310 wep skill vs a 63 elite: Average Glancing: 148 Average Hit: 153

So glancing blows are 96.9% of normal hits or 94.4% with crits added as well.

According to Athan's formula it should be 95% but according to the numerous sources I've linked it should be 100%.

IMO, 96.7% of averaged hits is within the margin error of both 95% and 100%.

I still see no tests of any kind past 310 wep skill on this source.

Where as I've linked 7 different sources all with the same conclusions, some using swing tests, etc.

[Scyllaqt]

If anyone finds an archived version of this rogue thread that would be great. Apparently the author compiled a ton of data to support the linear formula and it's a thread that many of the sources linked above reference.

[GuybrushGit]

Some more data:

+4 levels (attacking from behind, no block / parry): 
111 miss / dodge, 387 glancing, 239 hits, 0 crits. 
total hits = 737. The range of values of Glanging Rate for which this would not be statisticly unlikely (less than 5% chance to happen) is 49.5% - 55.4%. The average damage from hits was 34.6, and from glanging blows was 15.8, which is 45.8%. The total loss, neglecting crits being overwritten, is approximately equivalent to +26% miss. 

+2 levels (attacking from behind, no block / parry): 
254 attacks, 71 glancing. The range of values for which this is a feasible result is 23.7% - 32.7%; likely values are 25% or 30%. The average damage of the glancing blows was 85% of normal hits. Estimated loss is about +4% miss.

http://web.archive.org/web/20060516141657/http://forums.worldofwarcraft.com:80/thread.aspx?FN=wow-warrior&T=942763&P=2