HIT

It's becoming pretty widely accepted that the missrate for a level 60 hunter against a boss (considered to be level 63) is 8.6%.
Last I checked, it's also known that 1 Defense gives you, amongst other things, a 0.04% higher chance to be missed in melee.
In terms of hits and misses, 1 Skill negates 1 Defense.
For melee, chance to miss against a boss would thus be 5.6% (5 Defense gained per level).
This would mean that for ranged, one of the following is true:
* Ranged has a miss penalty equal to 1% per level difference.
* 1 Defense gives more than 0.04% higher chance to be missed (possibly 0.24% - 1.2% per level, 3.6% against boss)

Note: Any way to add bulletpoints on this forum?

If the latter is true, one would also expect +skill to give more +hit than 0.04%.

Trying to test this, I ran around shooting bosses in MC, BWL and AQ, recording the stats I found.
I did not record anything from bosses with adds that had a level lower than 63 if I needed to shoot them (example would be C'thun).
If there was any chance of Multi-Shot hitting something other than the boss (for example Mind Control, critters roaming around, lower level adds), I did not use Multi-Shot at all.

Setup A: +6% hit
Total of 1727 shots fired.
Total of 44 misses.
Total Missrate: 2.55%

This seemed to fit pretty accurately into the known formulae.

Setup B: +6% hit, +4 skill
Total of 1313 shots fired.
Total of 17 misses.
Total Missrate: 1.29%

Using Auto Shot only, my missrate was 2.64% and 1.32% respectively.

While setup B is a fairly small sample, it's being added to whenever possible. I wouldn't reccommend using the missrate here to determine the effect of +skill for ranged, however, it's more than accurate enough to tell that +skill adds more than 0.04% chance to hit.
For reference, the exact same mobs were hit, and my gear was always consistant. If someone wants, I can note down the exact gear I used for both setups.
As a little note, after about 1.1k hits in the first test (+6% hit), the missrate never changed more than 0.2%. The values in the second test (+6% hit, +4 skill) change a bit more (due to 1 miss being a higher percentage), but I'm pretty sure results after ~2k shots fired with this setup will be within 0.5% of this even is a worst-case scenario.

It seems Defense (and thus skill) is more important for pure hit purposes for ranged than melee. If my above guess with 0.24% per level is correct (currently tested difference per level = 0.32% per level), my missrate should hover around 1.64% with setup B.


CRIT

With +skill obviously working differently in ranged than in melee (most likely due to no glancing blows, parrys, etc), how is ranged crit affected by it? The same as melee (0.04%), or the same as ranged hit (possibly 0.24%)?
Due to not having an easy way of checking chance to crit for ranged (melee can simply hover the Attack icon in spellbook), I decided to find something, and shoot it for a while...

Target: Servant of Grol, Badlands, level 53 humanoid.
No buffs or debuffs
Auto Shot only (Mend Pet when needed, Rapid Fire when available)
Again, exact gear for both setups can be posted by request.

Setup 1:
19.96% melee crit (24.96% ranged, due to talents)
Total shots: 5939
Total crits: 1584
Total critrate: 26.67%

Setup 2:
19.90% melee crit (24.90% ranged, due to talents)
+4 skill
Total shots: 2959
Total crits: 770
Total critrate: 26.02%

Both critrates are higher than the spellbook would indicate. This is due to the level difference. Since both tests were made on the same target, the increase in crit should be constant.
If crit was affected by +skill in the same way as hit, I would expect to see a fairly large difference in critrate.
Critrate in setup 1 was ticking down (critrate after 2k shots was about 28%), while critrate in setup 2 was ticking upwards (although it seemed much more stable than the first test). I am planning to do another 3k shots test with the second setup, possibly later tonight, to get more accurate results.

Since I find it highly unlikely that +skill decreases your crit chance, let's look at the problem from a different angle.
Using data from setup 1, due to bigger sample.

My ranged crit chance should be 24.96%. However, this is against a level 60 mob. However, the mob in question was level 53. Under the assumption that skill and defense counteract eachother, we get the following:
Level difference of 7. This leads to my weapon skill being 35 higher than the mob's defense skill.
This would imply the following:
CritModifier = (WeaponSkill - DefenseSkill)*x
RealCrit = Crit + CritModifier = 24.96% + 35*x

If the same applies for melee as for ranged:
RealCrit = 24.96% + 35*0.04% = 26.36%
Compare to what I got in setup 1, 26.67%. Difference = 0.31

Theoretical crit rate for setup 2:
RealCrit = 24.90% + 39*0.04% = 26.46%
Compare to what I got in setup 2, 26.02%. Difference = 0.44

Although my results are a bit off (too small samples to tell accurately, it seems), it would seem that, for crit purposes, +skill affects melee and ranged the same.

Educated guess as a conclusion:
+1skill = ~0.24% ranged hit and 0.04% crit.

[EDIT:] I did actually go back and test more, and the results I got for both hit and crit approached my above guess more and more as total shots increased.

As a gun using dwarf, I have noticed that I have less misses then the others, compared to +hit, or, so it seems (how things seem however, are a very risky thing to go by since the mind likes playing tricks on you). As for the crit, while there is no reason for them to make +skill do less crit for ranged, there might very well be bugs present that lead to this (while the mind does indeed play tricks on you, my dmeter comparisons in bwl tends to have the other hunters consistently critting more then their spellbooks should indicate, while I consistently crit lower. But unfortunately 9 out of 10 times this is for single bossfights, so the data is insufficient to draw any conclusions).

I'm planning doubling the sample sizes there, getting a total of 9k shots for each (crit) setup =)
Thank [preferred deity] for being able to watch Farscape in windowed mode while shooting ;)

One a minor note, if you're using pet, and treating pet data as your own, your crit rate will probably be much lower than your spellbook. I seem to remember my pet having less than 10% chance to crit during this testing (although I admit, this is of the top of my head, and might be horridly wrong).

On the racial note, if I'm right with the 0.24% per skill, it would also explain why trolls / dwarves don't need as much +hit gear as others, while still using the standard skill vs defense system.

Something interesting to note is that not all mobs have level * 5 defense. I'm assuming this because you can get glancing blows even against some level 60 and 61 mobs, and the amount of glancing blows you get on level 63 mobs can vary from one to another.

Do you have the humanoid slaying talent? That would affect your Blasted Lands testing.

Interesting that +skill helps more than I thought it did.

So by your estimate:

Barb of the Sand Reaver = 82 AP, 0.83% crit
Eye of Nerub = 90 AP, 1.07% crit, 0.96% hit

Interesting.

Wait, dumb question: What's the AGI/crit equivalency for hunters again?

60, no?

52.88 at level 60.

52? 53? Something like that.

EDIT: What the guy above me said.

Ah, I'd always heard 50, which I guess was just an approximation -- ok, that explains the difference in our numbers.

It's still inferior to two crit/AP 1h weapons with 15 agi each, but the main thing I was trying to get at was the direct comparison to Barb, since some hunters in EJ have questioned the extent to which Barb-->Eye is really an upgrade. I think that 0.25% crit and (effectively) 1% hit on top of 8 AP for a MM hunter, plus higher raptor strike damage for gimmick meleeing, is a reasonable upgrade for that slot.

I've always looked at the polearm as a more PvP weapon, but the skill / hit relation makes it more interesting than I previously thought, moreso for encounters heavy on the enviromental damage, due to the stamina I think it has :P
I'm not taking it over any LR hunter, and I'll probably not take it at all, due to my DKP saving schemes.
Still, very nice weapon.

Well - barb never, ever impressed me. It is just not good enough. It's only an odd 24ap more than FotF + Silithid Claw, and you could even switch FotF for Hatched of Sundered Bone making the difference in ap a mere 16, with 1.17% extra crit.

Eye of Nerub sure looks interesting, but with tier3 loot you already have near max +hit, and those stats are just wasted on nerub then. I don't like either of those.

And just now I noticed this post. Ignore my former post.

the problem is that you lose a lot of stamina going from barb to eye, and stamina is the real reason (at least PVE-wise) that you'd give up dual wielding for the superior dps stat, and PVP-wise stamina is the main reason i wear barb too. maybe there's fights in which you have to melee or something :-P

For encounters heavy on the enviromental damage you often have to wear some resist gear and lose some %hit and sta. So i think this weapon may be quite usefull in some situations.

Back to topic, being a Troll hunter, i cant remeber missing a tranq shot with +7%hit while using a bow. I dont have any record, and may have been very lucky, but until now +skill has often be underrated for ranged weapons.
Before that topic i was just thinking that someone with +5bow was a kind of lv 61 player for miss and crit rate. So i'm very interested in all the record that can be collected (i'm no more playing my hunt atm ).

Btw, thx a lot for all the work youre doing Lactose.

But if +5 skill makes you a level 61 player for miss rate, then your expected results would be a 7.4% miss rate vs. level 63s (down from 8.6%, or 5*0.24, which is consistent with the numbers in the OP). But then you'd still expect to miss with just +7%, even if only very rarely....

Surely figuring out the value of +hit required for a troll with a bow to never miss a lv63 mob can't be too hard....

i dunno - i've made this point before, but there are definitely fights in which i value a marginal increase in HP over a marginal increase in damage. c'thun is one of those, because of the high amount of environmental damage. ouro is another for the same reason. sometimes i get sand blasted/hit by scarabs/burrowed to really low health, and if i had less stam on my gear i'd just be dead.

my point, though, is that the reasons i use barb in PVE are all survivability based, not big-damage-two-hander based, so losing 9 stam for 4 agi and some weapon skill/topend is a weird way to 'upgrade' from barb.

Yeah, with the right equipment, we're generally swimming in +hit-- so the +skill from Eye is only really a +crit % bonus or an excuse to wear a piece of gear with slightly upgraded stats but no +hit (Band of Reanimation, etc.).

2 1H's of high calibur (Silithid Claw, Hatchet of Sundered Bone...) are still better for PvE unless you've got a real itch for Stamina. In PvP, there's no question either Barb or Eye are superior for melee purposes and for the extra HP.

That is the main reason I don't see a point in "upgrading" from Barb to the Eye of Nerub. The primary reason I equip a Barb is for Stamina, so why would I want an Eye of Nerub when it's a downgrade in Stamina? If I want damage, I have my DW combo.

I whole-heartedly agree. Eye of Nerub was some kind of ill-conceived hybrid-- and there's no use for a hybrid when you have the options available (either Barb or DW).

It's fine for people who don't have Barb... but picking up the Eye when you already have the Barb doesn't really get you anything.