Talk:Hit

Condense please
This article is intensive and rather confusing. Basic confusion such as: I don't understand how my chances to miss are greater with dual wield than with a 2 handed weapon. Did bliz really do that?

I get lost in the formulas and I was wondering if someone could condense them so that we can plug in our particular stats and know what will happen. For example, if I trade out an item that increases my agility by X and decreases my hit rating by Y, what will happen?

Also, I have the same question that is addressed below, what happens with a single 1 Handed weapon equipped? (talk) 14:19, March 2, 2010 (UTC)

Moved argument about EJ testing to talk page
I moved the following paragraphs from the main page here, as the arguments are conflicting and probably outdated. Owldark (talk) 00:01, March 1, 2010 (UTC)

(There is no evidence to support a 2-roll attack table save a post from 2006 on the EJ forums. Additionally, the testing itself was neither strict nor rigorous. From a programming perspective, it makes much more sense to use a single calculation rather than having 2 ongoing calculations.)

''Until you do very basic tests and see that there is a very large disparity between special attacks, white attacks and character sheet crit and hit listings indicating there is a special scenario for yellow attacks. From a programming perspective it is no less feasible than checking if each attack is made from behind for consideration of rogues and dodge.''

Changed header level (and also added sub-section on hit rating)
The first section header was at level 2 (==), and the sub-sections incremented from there. That made it a bit hard to tell sub-sections apart as soon as we got to level 4 (====). So I changed all the section levels by 1. The first section header is now at level 1, and the sub-sections increment from there. If I've over-stepped a layout guideline that I'm unaware of, please let me know. -- 22:11, 2 January 2008 (UTC)


 * You have indeed overstepped a layout guideline; there should be no h1s called on a page, except for the title. Bad coding, among other things. See WP:MOS. --Sky (t | c | w ) 22:17, 2 January 2008 (UTC)


 * Sorry about that. -- 04:51, 7 January 2008 (UTC)

Missing when you're over hit cap rating
I'm thinking of reverting the change that mentions that there is still a 0.1% chance of missing, even when you're over the hit cap rating. Unless someone can find a reference for this, or show a WWS log (and if so plz list your hit rating). Here's an example of a WWS log where I had 364 hit for this fight and there were no misses: http://wowwebstats.com/fbijvt41monhi?s=3163-3605&a=6. From reading on www.elitistjerks.com, I've read that they also had 0 misses when they reached the cap. -- 05:06, 7 January 2008 (UTC)


 * Deleted "Note that even if you cap your hit rating, there is always a 0.1% chance to miss a target" as it directly conflicts with "once you reach or surpass the hit cap, you will never miss an attack against a raid boss."--Tiwuno 05:00, 16 January 2008 (UTC)


 * Looks good. Thanks.  --  17:48, 16 January 2008 (UTC)

What about one-handers?
* with two-hander: 5% + (Defense Skill - Weapon Skill)*.1% * with dual-wielding: 24% + (Defense Skill - Weapon Skill)*.1%

What about one-handers? I guess they use the same formula as for two-handers, but it should be clarified/the wording should be changed if it is so. --Deathwinged 15:24, 17 April 2008 (UTC)

Hit rating cap - no Druids?
Hi, it seems druid info is missing in the list. —The preceding unsigned comment was added by.
 * Sign your posts with  See Hit.  21:19, 15 June 2008 (UTC)

Rogue Precision
The talent still exists. I don't know what the hell someone was thinking when they (incorrectly) edited the section. Please take a look at up-to-date talent calculators (or in game talent panel) before making edits on the basis of a talent change.-- Oni Ookami Alfador Talk 00:50, 5 November 2008 (UTC)

One vs Two Roll Tables
I am under the impression that at least physical special attacks all use a single roll system as suggested by the "It is possible and even likely that even yellow attacks use a one roll table" comment. Based on that is there any reason not to remove the two roll table description? As is I think the presentation is confusing and seems to suggest that the two roll system is in place. Zalambar (talk) 06:07, 27 November 2008 (UTC)


 * There is some discussion of the one/two roll system at http://elitistjerks.com/f47/t15596-melee_combat_mechanics/#One_roll_vs_Two_roll. It's not updated for 3.0, so I'm not sure if it's still the case.  As for whether we need it, most of my expertise is on the spell casting side of things, so I can't really say for certain.  -- 10:33, 3 December 2008 (UTC)

Accuracy of ratings per 1%
The rating required for 1% hit at level 80 may be 32.79, not 32.77 as stated - see Talk:Combat rating system. I'm adding an accuracy tag until I or someone else has time to verify the numbers. -- 10:20, 3 December 2008 (UTC)


 * I looked at my hit rating along with some of those in my guild and with friends I play with and the ratio looks to be 32.77. Paly 1 (talk) 22:45, 3 December 2008 (UTC)


 * Did any of you use API_GetCombatRating and API_GetCombatRatingBonus to calculate the number with high precision? There's not enough accuracy in the information provided by the default UI to reliably achieve 4 significant figures.  Having just done the precise calc through the API, 159 hit rating (4.85% melee hit on char sheet) is actually 4.8490409851074% chance to hit, and the rating for 1% is 32.789988884055, i.e. 32.79 to 4 significant figures.  The magic incantation to get the client to reveal high precision numbers via the API for melee hit is:
 * I will update this article soon, unless someone gets there before me. -- 01:36, 4 December 2008 (UTC)
 * I will update this article soon, unless someone gets there before me. -- 01:36, 4 December 2008 (UTC)

If you do the math for 9% hit rating you have 32.77*9 = 294.93 and 32.79*9 = 295.11 you are only talking about a .18 difference. So for 9% 295 is a good number. No one is going to try to hit cap themselves to 296 to make sure they get that last .11 points past 295. The miss is so miniscule for .11 points of hit rating. At 25% it is 32.77*25 = 819.25 and 32.79*25 = 819.75 and that is a .5 difference. For 25% hit rating a 820 is fine. I really do not see your point since our hit rating on our gear, gems, and enchants are in whole integers and not fractions. I do not see any +12.3912 gems etc... out there to make us perfectly hit capped. Paly 1 (talk) 04:22, 4 December 2008 (UTC)


 * It's not really a question of how big a difference it makes (I agree it's a small difference), it's a question of accuracy. The evidence direct from the game API suggests that 32.77 is not an accurate representation of the conversion from combat rating to effect, and that 32.79 is the accurate number.  When it comes to theorycraft, precision and accuracy are very important, and it looks bad when WoWWiki has incorrect information.  More importantly, it causes confusion when the numbers here do not agree with other highly respected sources of theorycraft information (e.g. Elitist Jerks).  With accurate and precise information, people can make their own choices about the importance of being hit capped or being very nearly hit capped.  At the high end of the game, some people spend a huge amount of effort to squeeze every last drop of potential from their gear, and inaccurate information just causes confusion, or even needless expenditure (getting those last few rating points can be very expensive when you are looking at the likes of epic gems).  -- 06:08, 4 December 2008 (UTC)

I added a hit rating chart that goes out 2 decimal places using the 32.79 number. People can get a better idea of what the actual hit rating vs. the rounded number. I hope that helps. Paly 1 (talk) 23:36, 4 December 2008 (UTC)


 * Yeah, that seems like a reasonable idea, thanks. I have expanded it to include numbers for level 70.  I've also gone through the article and changed all instances of 32.77 to 32.79.  I have not checked or changed the various individual cap numbers - I'm assuming that they are ok.  accuracy tag removed, as I think my original concern is suitably addressed.  -- 06:52, 5 December 2008 (UTC)

Class specific hit requirements
The class specific entries were starting to take up a significant portion of the page, contained inconsistent data, and in some cases contained outdated information (Shaman reaching the hit cap through talents alone). I have therefore decided to be bold and consolidate most of that information into a single table to display the hit rating required to cap special attacks against a raid boss. This leaves out some data regarding the hit cap of dual wielding auto attacks modified by talents but I do not believe this is a significant loss as the unmodified auto-attack hit cap is still explained on the page and it is generally not desirable for characters to attempt to reach the hit cap with auto attacks. If anyone would like this data back I suggest adding a second similar table for auto-attacks. --Zalambar (talk) 10:47, 5 December 2008 (UTC)

Thanks for expanding on my idea of the hit rating chart. It looks a lot cleaner with your consilidation of the class information into a single chart. Paly 1 (talk) 11:59, 5 December 2008 (UTC)

Yes, I like that change, and it's actually something I'd been contemplating myself. To be honest, the only reason I hadn't done it so far was that I was still trying to figure out if there was a benefit to the table per class layout. -- 03:40, 6 December 2008 (UTC)

Updated with new information
There has been a lot of debate recently about a seeming change to the hit cap for melee classes, and possibly spell casters as well. In The Burning Crusade expansion it was shown to be at 9% for melee. However there is very compelling evidence to show that some change has been made. I am not really a theorycrafter myself however I do value the accuracy of WoWWiki. There is still some conflicting evidence coming in for some classes such as Retribution Palladins, however Druid and Hunter theorycrafters have already redone their models based on a 8% hit cap against level 83 raid bosses.

Here is a link to elitist jerks with some pretty compelling numbers: http://elitistjerks.com/f31/t37856-melee_hit_cap_really_9_a/

Wasn't sure how exactly to go about pointing this out so if I messed up I apologize in advance. Furlion (talk) 18:10, 16 December 2008 (UTC)


 * It would be nice if a Blue would post to confirm the change if it did do so. I think unless a Blue posts it will safely remain at 9%. Paly 1 (talk) 23:23, 16 December 2008 (UTC)


 * I wouldn't hold your breath on a Blue post on something like this. While they have posted some details of combat mechanics in the past, it's quite rare that they do so.  I've not read the whole EJ thread referenced above, but it's probably the best place to find out, once they have finished analysing and debating the finer points of it.  It could be an undocumented change, or it could be a bug / mistake like the old "ghost hit" with frostbolts.  -- 00:22, 17 December 2008 (UTC)


 * You don't need to wait for the blue post, which is possible that there won't be any about this issue. Test it yourself. My hunter is 8.08% hit atm. No miss so far in my last few Naxx runs. Nevertheless I won't conclude it would be the case for every class, so test it yourself with 8% hit, and if you see any miss, just go back to 9%.
 * I think adding a small reminder may help atm. After all the evidence is conclusive, we can rewrite all the related value. --WakemanCK (talk) 05:16, 4 January 2009 (UTC)

Now that Blizzard finally fixed the "ghost hit" today the guys on EJ can finally once and for all figure out what our hit rating is really at. Paly 1 (talk) 03:41, 9 January 2009 (UTC)


 * More of the EJ guys have been testing it and pouring over WWS reports in Feb/March and it seems 8% is the number. I used the wordage "near certainty" since there's been a couple unexplained misses pop up after a few thousand swings/shots in Naxx.  Scifischizo (talk) 10:18, 15 April 2009 (UTC)

Difference between soft cap and hard cap
I don't actually understand why this page makes a difference between the soft cap and the hard cap. As far as I understand this they both are at 263 hit rating. Or am I mistaken here? If I'm not I'd suggest using another wording for these caps as "soft" suggests that the cap can be "bend"/extended whereas the "hard" cap cannot. Manski (talk) 08:15, 19 June 2009 (UTC)


 * Both the soft and hard cap for some classes such as hunter and retadin are 263. For dual-wielding classes (e.g. rogues, fury warriors) and those with lot of spell damage (e.g. unholy DK), soft and hard cap are different. WakemanCK (talk) 13:59, 21 June 2009 (UTC)

Example of hit-reducing debuff
changed the example of hit-reducing debuff to a more modern one (Archavon the Stone Watcher's gas cloud)

Discussion of hit vs. other stats too simplistic
The discussion of "why one wants hit" simplifies way too much. One does not always just want hit to maximize your DPS - there are some abilities that need hit (like taunts, or hunter stings needed for boss-specific fights) that simply must always land, or the raid will be wiped. This is a topic that I've had to explain to too many new people joining one of my raids. Raiding DPS and tanks must hit cap before anything else, period, imo. Babasyzygy (talk) 04:27, June 11, 2010 (UTC)

Outdated?
The chart says my hunter needs 8% / 246 hit to never miss. She's at 9.82% / 304, but the character sheet shows a miss chance of 17.18% with normal attacks vs a level 83. Special attacks are all at 0% miss, though. What gives? --Azaram (talk) 11:05, 24 November 2010 (UTC)
 * If you're dual-wielding, your miss chance skyrockets. -- Dark T Zeratul (talk) 11:07, 24 November 2010 (UTC)
 * Mm, that may be it. It was under the 'ranged' section, though. --Azaram (talk) 07:12, 1 December 2010 (UTC)

Calculating Miss Rate
I am not sure this is needed anymore. The formulas will not be consistent with MoP when the cap will be 7.5% and not 8%. I think it can be cleaned up or removed. Neo (talk) 14:53, 30 August 2012 (UTC)