[Rogue/Warrior] Weapon Skill Adjustment Discussion - Page 19

Olgas · 08/29/07, 5:58 PM

Originally Posted by sp00n

The 1 miss in 20500 attacks is under a 0.4% theory.
Under this, with 350 hit rating I should see 1 miss in exactly 20500 attacks, with 351 hit rating I should see none.

Oh, OK, I get it now.

Originally Posted by Karmon

I will try this friday on our Karazhan run with 343 and 344 hitrating.

I was just thinking of one other test you might try, Karmon....

At 351 hit rating and 353 weapon skill, you saw zero misses (2542 attacks).
At 350 hit rating and 353 weapon skill, you saw zero misses (3989 attacks).

So, what if you were to try 349 hit rating or 348 hit rating??

Under the .26667% theory, you should see a miss with 349 hit rating after about 1464 attacks. Or, with 348 hit rating, you should see a miss after about 759 attacks.

I guess I'm just wondering if it might make sense to fully disprove the .26667% theory before jumping to testing the .4% theory?

In any event, I know all this testing is a pain, and all data points are valuable. So, I'll of course be happy with whatever test you decide to do. Just making a suggestion.

Morlark · 08/29/07, 9:48 PM

Originally Posted by Olgas

But I got lucky -- one of those shots was a MISS ! And it only takes 1 miss to prove the point.

I've seen this argument used a few times recently (and I'm not trying to argue with its validity here), but pre-TBC I was always told that even while hit-capped there was always a very small (but not precisely determined) chance to miss. And I have certainly heard many (admittedly anecdotal) stories of people missing while hit-capped, pre-TBC. Given this, are we entirely certain that the hit-cap has changed at all? From what little I remember of reading the theorycrafting pre-TBC, a lot of the hit-cap calculations were based on assumptions extrapolated from what we knew of how defense works. Was the hit-cap absolutely definitely 8.6% pre-TBC?

Edit: Crap, didn't notice I was several pages behind. That's what I get for posting while tired.

Olgas · 08/30/07, 12:56 AM

Originally Posted by Morlark

I've seen this argument used a few times recently (and I'm not trying to argue with its validity here), but pre-TBC I was always told that even while hit-capped there was always a very small (but not precisely determined) chance to miss. And I have certainly heard many (admittedly anecdotal) stories of people missing while hit-capped, pre-TBC. Given this, are we entirely certain that the hit-cap has changed at all? From what little I remember of reading the theorycrafting pre-TBC, a lot of the hit-cap calculations were based on assumptions extrapolated from what we knew of how defense works. Was the hit-cap absolutely definitely 8.6% pre-TBC?

Yes, you can eliminate ALL physical damage misses with sufficient hit rating. This has been well established by tons of evidence and testing. It was the case before TBC, and it is still the case now. I have certainly heard the anecodotal "there is always some small chance to miss" argument before, too. It is simply not true.

However, there are curses and debuffs that can affect your chance to hit. So, even if you are "hit-capped," you will still miss when you are affected by these debuffs.

(As far as I can tell, the misinformation that "there is always some small chance to miss" always seems to arise either from hit-capped people who were not paying enough attention to their debuffs at the time they experienced a miss, or from people who were confusing the physical damage system with the magic damage system, where there is always a 1% chance to miss with spells.)

As for your second question, was 8.6% really the hit cap pre-TBC? The truth is, I've been wondering about that myself a lot lately. Pre-TBC, you could only add +Hit in even percent integers. That is, you could equip +8% Hit or +9% Hit, but nothing in between. You couldn't equip exactly 8.6% Hit, because the Hit Rating system didn't exist yet. The common wisdom at the time was that 8.6% was the hit cap, but it couldn't really be tested exactly. Folks tested at +9% Hit and found that they would never miss with 9% equipped and this served as validation of the 8.6% theory. But I suppose the hit cap may actually have been 9% all along and we just never knew it!

Of course, it's impossible to go back and test now, but it's funny to wonder about that...

Darkrenown · 08/30/07, 7:09 AM

This may be a stupid question, but do your MH and OH share the same hit chance? What happens if, say, an Orc has a axe MH and a sword OH?

Giske · 08/30/07, 7:42 AM

Dual wielding gives you a 19% miss chance on both weapons so its 24% total with the 5% base, regardless of weapon type. However, chance to hit with weapons could be different if you have effects that only apply to a certain weapon type. Like Orcs with their Axe racial.

rooppa · 08/30/07, 8:03 AM

Something has been bothering be about the amount of +hit the 1st 5 points in weapons skill awards (going from 350 to 355).

If we assume that you have 20% (or slightly more) hit through hit rating, which should be in the region of 316 hit rating, (using the rough guide of 20*15.77= 315.4, and then rounding up the 0.6 to ensure the 20%+ hit and not 19.9746% with 315 hit rating) then you take into consideration the 5% hit which is generated through the rogue talent of precision. Going off the extremely nice work of others in this thread, there should still be a 3% base miss rate for the player attacking a level 73 mob.

Now if a human rogue was to attack this raid boss, without any additional points in weapons skill other than there base +5 racial (so obviously they would have to use dual swords/maces)

They "should" not encounter a miss if indeed the 1st 5 points in weapons skill is (meaning going from 350 to 355) worth 3% hit.

This would therefore also mean that a human rogue with 315 hit rating and precision does not need to spend 2 talent points in Weapons expertise. (In relation to the hit cap that is, obviously weapons skill does more than just add +hit)

Not exactly game altering, but 2 talent points are 2 talent points.

But then something else occurs to myself.

The "hit cap" mantra that I have always been told is that a rogue requires, as in "308 hit rating, 5 points in precision and 2 points in WE in order not to miss".

But if increasing you’re hit rating by a measly 8 rating (1 gem) means that you can save a talent point in weapons expertise.

Therefore in regards to human rogues, is it really worth them going for weapons expertise at all? Or in the case of normal rogues, spending a 2nd talent point in weapons expertise? I am under the impression that the extra difference from going higher than 355 weapons skill to say 360 or 365 is not really worth the DPS difference for a rogue compared to other (although very limited) talents that rogues would be able to spec into with the spare talent points. This, of course, is simple speculation until the complete understanding of what + weapon skill actually awards in other areas such as crit/anti-dodge. But with people testing as regular as has been seen, am sure someone would be able to come up with a comprisable DPS figure for NOT taking the 2nd (or any at all in a human case) point/s in weapons expertise talent and placing them into another talent.

sp00n · 08/30/07, 10:08 AM

Originally Posted by rooppa

...

Correct, for a human with swords 1 point of weapon skill is basically worth only 0.1% to hit + the still to determine part of dodge/parry/block.
So their racial is worth 3% hit (didn't Blizzard want to decrease the value of racial skills?).

With 2 points in WEx they reduce their chance to miss by further 1%, so they'd only require 300 hit rating to match the now 24% (19%) miss rate.

The real question is, what other talents are available to increase DPS besides WEx. With a standard sword combat build we only have one of the two poison increasing talents available, improved eviscerate, or other talents that do not increase DPS, but rather concentrate on survivability (parry, dodge, endurance). Or talents more suited for farming/pvp like remorless attacks, expose armor and stealth.

One could even take improved backstab to have some variation in his attacks, but without dagger specialization and without a dagger racial and the lacking skill for daggers, it would be only for 'fun' reasons, as it gives no real advantage.

Punscho · 08/30/07, 1:09 PM

There aren't that many places to put those 2 talent points in raid DPS wise. For PvP there are imp sprint, imp kick, nerves of steel and such, but from a pure raid DPS standpoint I'd say you're better off taking the 2 points of WEx and using those 2 gem slots for AP or something.

For warriors it's a whole different thing since the fury/arms build don't have any extra weapon skill so humans and orcs should really focus on getting the right weapons since it yields them 3% hit. And of course orc shamans aswell.

For the 2h warrior builds (arms/fury) it's easier to cap out those 9% because of the weapon mastery talent that yields +4 skill points and the total lower amount of hit required.

Kandiru · 08/30/07, 2:23 PM

Paladin tanks have a choice of getting 3% to hit from precision talent or gaining 6 weapon skill from 3/5 in weapon expertise using the "standard" cookie-cutter build .

From the current results it seems that the 6 points in weapon skill will be superior on bosses from the point of view of hit, but this is from a rear attack perspective, has there been any results for how much it will lower /block/dodge/parries from the boss?

From the results it seems for level 70-72 mobs the 3% to hit is better, but against raid bosses the weapon skill will be best?

GSH · 08/31/07, 6:37 PM

Apologies if this has already been covered. As I understand it, the current theory is that dual-wield miss rate is:

70 mob: 24%
71 mob: 25%
72 mob: 26%
73 mob: 28%

For rogues against a 73 mob:
/w 1/2 WE: 26%
/w 2/2 WE: 25%
/w 2/2 WE and Precision: 20%

Now, a 20% miss rating implies that you need 316 hit rating to never miss. Yet the accepted rogue hit cap is 308. Is 308 correct?

I'm asking because I maintain a list of hit caps at Blessing of Kings: +Hit Caps for Bosses, and I would like to make sure it is accurate. (Though I have no idea how I'll cleanly represent the new weapon skill info.)

sp00n · 08/31/07, 7:14 PM

Originally Posted by GSH

Apologies if this has already been covered. As I understand it, the current theory is that dual-wield miss rate is:

70 mob: 24%
71 mob: 25%
72 mob: 26%
73 mob: 28%

For rogues against a 73 mob:
/w 1/2 WE: 26%
/w 2/2 WE: 25%
/w 2/2 WE and Precision: 20%

Now, a 20% miss rating implies that you need 316 hit rating to never miss. Yet the accepted rogue hit cap is 308. Is 308 correct?

It's:

70 mob: 24%
71 mob: 24.5%
72 mob: 25%
73 mob: 28%

For rogues against a 73 mob:
/w 1/2 WE: 25%
/w 2/2 WE: 24.5%
/w 2/2 WE and Precision: 19.5%

19.5*(10*82/52)= 307.5

Olgas · 08/31/07, 7:23 PM

Originally Posted by GSH

Apologies if this has already been covered. As I understand it, the current theory is that dual-wield miss rate is:

70 mob: 24%
71 mob: 25%
72 mob: 26%
73 mob: 28%

For rogues against a 73 mob:
/w 1/2 WE: 26%
/w 2/2 WE: 25%
/w 2/2 WE and Precision: 20%

Now, a 20% miss rating implies that you need 316 hit rating to never miss. Yet the accepted rogue hit cap is 308. Is 308 correct?

I'm asking because I maintain a list of hit caps at Blessing of Kings: +Hit Caps for Bosses, and I would like to make sure it is accurate. (Though I have no idea how I'll cleanly represent the new weapon skill info.)

You've got the numbers a little wrong. First, just to make sure there is no confusion, mob level actually has nothing to do with miss rate. The only determining factor is Weapon Skill v. Defense Skill. However, if we assume a standard Weapon Skill for a Level 70 character (i.e., 350 Weapon Skill), then, against mobs of the following levels the base miss rates for that Level 70 character would be as follows:

v. Level 70 mob: 24.0%
v. Level 71 mob: 24.5%
v. Level 72 mob: 25.0%
v. Level 73 mob: 28.0% (this applies to Level 73 mobs and skull bosses)

If a Rogue has 2/2 in WE, then his Weapon Skill is 360 instead of 350. At 360 Weapon Skill, his base miss rate against a boss will be 24.5%. If he also specs into Precision, this will reduce his miss rate to 19.5%. So, if this Rogue then equips 308 Hit Rating (+19.53% hit), he will be able to eliminate all misses.

GSH · 08/31/07, 7:51 PM

Thank you for the corrections.

Crimsonstorm · 08/31/07, 8:00 PM

So are the following Hypotheses true:

1) Miss % is based on the difference betwee nweapon skill adn the mob's defense skill, not specifically on level difference.

2) Miss % with 350 weapon skill is:

vs level 70 mob: 5%
vs level 71 mob: 5.5%
vs level 72 mob: 6%
vs level 73 mob: 9%

3) Each point of weapon skill appears to reduce the mob's chance to dodge by about .2%? Is this correct. Also, it seems to have no effect on parries, and seems to increase crit chance by .04%.

4) Each point of weapon skill from 351-354 appears to reduce miss chance by somewhere around .4%? The point gained from going from weapon skill 354 to 355 appears to have an exceptionally large reduction is miss rate?

Based on all that, I would hypothesize that weapon skill has the following effect vs a Level 73 Mob:

Weapon skill points 351-354: .4% miss reduction(!), .2% dodge reduction, +.04% crit
Weapon skill point 355: 1.4% miss reduction(!!), .2% dodge reduction, +.04% crit
Weapon skill points 356-365: .1% miss reduction, .2% dodge reduction, +.04% crit
Weapon skill points 366+: .04% miss reduction. .2% dodge reduction??, +.04% crit

Vs a level 72 mob, the first 10 points of weapon skill give .1% to hit, and after that, .04%
Vs a level 71 mob, the first 5 points of weapon skill give .1% to hit, and after that, .04%

Do we have any data which contradicts my hypothesis?

sp00n · 08/31/07, 8:12 PM

I have now gathered some more data with 351 hit rating (eagerly awaiting Karmon's similar testings) and am now at 2276 attacks without a single miss.
This pretty much rules out a 0.267% theory, because as said above, I would have expected a miss after around 480 attacks. Of course there is a chance that some of these attacks were made while improved faerie fire was active (+3% hit), but I only ever saw once on my debuff filter list an active faerie fire which was not noted as (Feral), so the impact shouldn't be too severe.

Also as mentioned above, this leaves us with the following situation:
2 points of weapon skill are worth more than 0.7415% to hit and less than 0.8049% to hit.
Resulting in between 0.3707% and 0.4024% per point.

Postulating a 0.4% theory would look like this (for a rogue, dual wielding, with precision and against a level 73 mob, as always):

Skill    Hit   Miss    Total    Decrease
+ 0   362.69   28.0%   + 4.0%   0.0%  [most likely confirmed]
+ 1   356.39   27.6%   + 3.6%   0.4%
+ 2   350.08   27.2%   + 3.2%   0.4%  [most likely confirmed]
+ 3   343.77   26.8%   + 2.8%   0.4%
+ 4   337.46   26.4%   + 2.4%   0.4%
+ 5   315.38   25.0%   + 1.0%   1.4%  --> 1% jump [hit rating confirmed]
+ 6   313.81   24.9%   + 0.9%   0.1%
+ 7   312.23   24.8%   + 0.8%   0.1%
+ 8   310.65   24.7%   + 0.7%   0.1%
+ 9   309.08   24.6%   + 0.6%   0.1%
+10   307.5    24.5%   + 0.5%   0.1%  [confirmed]
+11   305.92   24.4%   + 0.4%   0.1%
+12   304.35   24.3%   + 0.3%   0.1%
+13   302.77   24.2%   + 0.2%   0.1%
+14   301.19   24.1%   + 0.1%   0.1%
+15   299.61   24.0%   + 0.0%   0.1%

Now, concerning dodge rating.
I have added up all my tests so far with 352 weapon skill, and since hit rating doesn't affect dodge at all, it is unimportant how much hit rating I actually had in these tests.

I have 5578 white attacks and 1524 yellow attacks (Eviscerate, Sinister Strike and Hemorrhage).
For white attacks, I had 248 dodges, meaning a dodge rate of 4.446%.
For yellow attacks, I had 88 dodges, meaning a dodge rate of 5.774%.
Combining it, I am looking at 336 doges in 7102 attacks, and a dodge rate of 4.731%.

The difference between white and yellow attacks is rather large (1.3%), but this could be due to the much smaller sample size of the yellow attacks (sample size for white attacks is over 3.5 times as large as for yellow attacks), but it is rather obvious, so there might be a different mechanic at work, but this subject to futher testing.

I'm beginning to wonder if the (maximum) dodge rate for white attacks is fixed to 5%. At least I'm quite positive that there is no such dramatic increase like for the miss rate (for yellow attacks, see above). If it is set to 5%, 1 point of weapon skill is worth 0.277% 'anti-dodge' in this example. If it was set to 5.6% (as believed pre TBC), the value of 1 point would be 0.577% 'anti-dodge'.

However, the results I have do not match the summary made here, which is based on combat logs of other people in this thread.

Right now I assume that Captain Galvangar and Drek'Thar are no good targets for measuring dodge rates. Galvangar can be stunned, which eliminates dodge all together, and there is a chance that while Drek'Thar is performing Whirlwind, he may also be unable to dodge (which I don't know exactly).
A quick look at my combat log approves this. On Galvangar I have only 3.64% dodges, and on Drek'Thar 5.05%.

Taking Galvanger out of the equation, I still have 3191 white hits however, and the dodge rate between white and yellow attacks still differ by over 1% (1.38%).

At the moment I'm not sure how to integrate this data set.

@Crimsonstorm
We have no idea how weapon skill affects crit up to now.

Vs a level 72 mob, the first 10 points of weapon skill give .1% to hit, and after that, .04%
Vs a level 71 mob, the first 5 points of weapon skill give .1% to hit, and after that, .04%

1 point of weapon skill always gives at least 0.1% to hit.

Karmon · 08/31/07, 8:30 PM

Well, I decided to run Karazhan with 348 hitrating, 353 skill, 5% precision, dualwielding daggers

Results (excluding Attumen and Netherspite) are
714 crits, 1299 hits, 0 blocks, 713 glances, 0 misses, 137 dodges, 24 parries
2887 total attacks
853208 damage, 295.534465 damage per swing
342154 hit damage, 263.397998 hit damage per swing
138175 glance damage, 193.793829 glance damage per swing
24.732% crits
44.995% hits
0.000% blocks
24.697% glancings
0.000% misses
4.745% dodges
0.831% parries

So the 0.26666 multiplikator seems to be disproved.
I should have seen about 4 misses with that samplesize if the 0.26666 multiplicator is valid.

I might get a shot on Netherspite tomorrow. Then I will try 343 hitrating.

Darkrenown · 08/31/07, 9:35 PM

Originally Posted by Giske

Dual wielding gives you a 19% miss chance on both weapons so its 24% total with the 5% base, regardless of weapon type. However, chance to hit with weapons could be different if you have effects that only apply to a certain weapon type. Like Orcs with their Axe racial.

Thanks, but I'm aware of all this

All I was looking for was a confirmation of if hit chance is the same for MH and OH or not. I would expect the bonus from weapon skill would only affect the relevent weapon, but untill recently there's been nothing that would affect the hit chance of one and not the other, so I'm used to thinking of it as just one value and I'm not sure if the game tracks them seperetly either.

sp00n · 08/31/07, 9:59 PM

Originally Posted by Karmon

I might get a shot on Netherspite tomorrow. Then I will try 343 hitrating.

With 343 hit rating you should see a miss after 2050 attacks, so in only one fight you might not encounter it.
Also it was mentioned in this thread that Netherspite has some weird mechanics, possibly meaning that he has a lower defense value than standard bosses.

Olgas · 08/31/07, 11:24 PM

Originally Posted by Karmon

So the 0.26666 multiplikator seems to be disproved.
I should have seen about 4 misses with that samplesize if the 0.26666 multiplicator is valid.

Sp00n, Karmon:

Awesome tests, guys. I definitely agree that the .26667 theory has been disproven now. And I also think we definitely know that the per Level value can't be greater than .4024. I think Sp00n's miss at 350 proves that.

But, the one leap I don't think is fully warranted yet is the conclusion that: "1 point of weapon skill is worth more than 0.3707%" (not an exact quote, paraphrasing Sp00n's conclusion). Looking at these recent tests, a per level value of .35% (ugly as it is) could explain Karmon's run for sure, and most likely Sp00n's run, too. Now, .35% is certainly an ugly number -- almost as bad as .26667. But I just want us to stay pretty scientific about this before we rush to the .4% conclusion.

However, IF the correct number is .4%, then I guess that would leave us with the following formulas for calculating miss rates:

When Def Skill - Weapon Skill <= 10:
5% + (Def Skill - Weapon Skill)*.1%

When Def Skill - Weapon Skill > 10:
7% + (Def Skill - Weapon Skill - 10)*.4%

This is exciting guys, we've almost solved the whole puzzle!

Dontmindme · 09/01/07, 12:25 AM

Originally Posted by Crimsonstorm

So are the following Hypotheses true:

1) Miss % is based on the difference betwee nweapon skill adn the mob's defense skill, not specifically on level difference.

2) Miss % with 350 weapon skill is:

vs level 70 mob: 5%
vs level 71 mob: 5.5%
vs level 72 mob: 6%
vs level 73 mob: 9%

3) Each point of weapon skill appears to reduce the mob's chance to dodge by about .2%? Is this correct. Also, it seems to have no effect on parries, and seems to increase crit chance by .04%.

4) Each point of weapon skill from 351-354 appears to reduce miss chance by somewhere around .4%? The point gained from going from weapon skill 354 to 355 appears to have an exceptionally large reduction is miss rate?

Based on all that, I would hypothesize that weapon skill has the following effect vs a Level 73 Mob:

Weapon skill points 351-354: .4% miss reduction(!), .2% dodge reduction, +.04% crit
Weapon skill point 355: 1.4% miss reduction(!!), .2% dodge reduction, +.04% crit
Weapon skill points 356-365: .1% miss reduction, .2% dodge reduction, +.04% crit
Weapon skill points 366+: .04% miss reduction. .2% dodge reduction??, +.04% crit

Vs a level 72 mob, the first 10 points of weapon skill give .1% to hit, and after that, .04%
Vs a level 71 mob, the first 5 points of weapon skill give .1% to hit, and after that, .04%

Do we have any data which contradicts my hypothesis?

1) Yes, that appears to be the case.
2) Yes.
3) I agree with crit, not so sure on the dodge though. I'm personally leaning to 0.1% based on attacks against level 62 and level 63 mobs with 360 weapon skill. I didn't see a miss against 62's, but did against 63's. Assuming 5% base dodge rate, at 0.1% per weapon skill explains this.
4) No, it's a flat 0.1% per weapon skill all the way to 376, with the odd exception at over 10 weapon skill difference. This has been tested.

Dontmindme · 09/01/07, 2:18 AM

While we are on the topic of interesting numbers, assuming weapon skills are linear from 351 to 354, the number 0.375 seems interesting as well (can be expressed as 15/40). Given that 0.400 and 0.375 seem to almost bracket the low and high ends of our current theory, I decided to chart out what effect these 2 possibilities would have on hit rating:

0.400 factor, 351 weapon skill = 357 hit rating not to miss
0.375 factor, 351 weapon skill = 357 hit rating not to miss

0.400 factor, 352 weapon skill = 351 hit rating not to miss
0.375 factor, 352 weapon skill = 351 hit rating not to miss

0.400 factor, 353 weapon skill = 344 hit rating not to miss
0.375 factor, 353 weapon skill = 345 hit rating not to miss

0.400 factor, 354 weapon skill = 338 hit rating not to miss
0.375 factor, 354 weapon skill = 340 hit rating not to miss

So, it appears that at least 353 weapon skill will be needed to demonstrate a difference with 354 being even better (although more difficult to piece together in the way of gear).

GSH · 09/01/07, 3:03 AM

Originally Posted by Dontmindme

So, it appears that at least 353 weapon skill will be needed to demonstrate a difference with 354 being even better (although more difficult to piece together in the way of gear).

You could test 354 with a warrior fairly easily. It would be an odd spec, but their Weapon Mastery adds 4 weapon skill. Rogue with 2/5 Mace spec would also work.

sp00n · 09/01/07, 8:20 AM

Ok, I took the 2 free talents points from not having WEx and put them into mace spec.
Now I'm going to run around in AV with stick grenades from WW2 ([Cudgel]).

If I don't encounter a miss (which could take some time in AV...) we've narrowed it down to 0.3915% - 0.4024%.

Oh well, there goes another 50 gold for respeccing afterwards.

Leguaris · 09/01/07, 12:48 PM

Since I don't have time to run tests, I thought I might contribute in another way: digging though WWS for data. The following is melee data for main tanks versus various bosses. Since I lack a way to reverse engineer a combat log .txt file or a way to see only a particular attack by specific people against specific targets I had to limit the fights to those I could be certain the MTs where only attacking the boss.

Human data:
Normal + Crit: 1725 (49.870%)
Glancing: 866 (25.036%)
Miss: 170 (4.915%)
Parry: 491 (14.195%)
Dodge: 207 (5.984%)
Total: 3459

Nonhuman data:
Normal + Crit: 2907 (49.022%)
Glancing: 1530 (25.801%)
Miss: 485 (8.179%)
Parry: 641 (10.809%)
Dodge: 367 (6.189%)
Total: 5930

-All attacks were made facing a boss target.
-Hit % data is useless, as each attacker has various (and possibly) changing levels of hit rating. Rather, this data would probably be more useful when considering dodge and parry.
-The human sword/fist racial is the only weapon skill involved in the test.
-All human attacks were made with [King's Defender].

First question: what is going on with those parry numbers?

Also,

Weapon skill points 366+: .04% miss reduction. .2% dodge reduction??<italics mine>, +.04% crit

I'd imagine that weapon skill would instead .04% dodge reduction (as well as .04% parry reduction), which would make it the exact opposite of defense.

sp00n · 09/01/07, 1:28 PM

Hell I don't know. From tests before we're quite certain that dodge is being reduced by about 0.1% if you're at the same weapon skill value as defense or higher (read: you're attacking a mob at the same or lower level). As for higher levels, this data seems to indicate a 6% dodge rate against bosses. I mean, 5930 and 3459 attacks are quite a few, although we e.g. see a variance of 0.8% for glancing blows.
The problem is finding a formula that gives about 1% for 15 points of defense. 0.06%? 0.067%? 0.07%?
Are we even sure about the 5% standard dodge chance of mobs with the same level?

And parry... I'm completely lost here. The nonhuman data shows less parries than the human data with 5 additional skill.

The original data from our two crazy paladins show a 11.5% and 11.7% parry rate, with 15 weapon skill providing a 0.2% difference.

But I'm still not sure if they changed the combat mechanics since then. They both have 1% less glancing blows and their dodge rate seems to be rather low as well.

I. Just. Don't. Know.

And now for something completely different, with 600 swings on 338 hit rating with 354 skill I haven't encountered a miss so far. The 0.35% theory would require about 603 swings for a miss to happen, and 0.375% about 1519 swings.
Going to continue clubbing some poor Orcs.