Where is the spell hit cap? - Page 2

epoh · 12/12/08, 11:25 AM

Originally Posted by Pitbuller

You have precision. 9.57% + 3% + 3% + 1% = 16.57%
Only 0.43% chance to miss. I think that mirror images don't get that talent and that's why they missed more or jsut bad luck.
Don't jump to conclusion.

I agree, needs more testing. Plus there is the upcoming change anyway. But at 15% hit I haven't seen a single scorch miss. Haven't thrown out thousands of them or anything though. If you look at that WWS I posted take a look at the other mages and their hit rating. You'll notice Drax, the arcane mage had next to no hit at all (less than 80, iirc) and he missed about 7% of the time. Perhaps that fight is just too short to be much of an indicator.

Asmo · 12/15/08, 11:39 AM

Originally Posted by Malekithe

My experience over the last week indicates that the shadow priest issues with spell hit still exist. I ran 25-man Naxx and Sartharion with 4% hit on my gear and did not encounter a single miss during a boss encounter according to WWS (though I got plenty of unlogged "whiffs"). In light of this, I decided to run a few focused tests against the boss-level target dummy.

First, I stripped off all my hit gear and began chain casting smite. Recount logged the expected miss rate of ~17% over about 300 casts. Then, I began chaincasting Mind Flay with the same 0 hit rating. After 1000 ticks of mindflay, recount indicated a 3.2% miss rate. Note that these are misses of the actual mind flay pulses. I got many "whiffs" over the course of this second test. I didn't keep track, but, as I frequently encountered 2-3 whiffs in a row, the rate was much higher than the actual miss rate of the pulses.

After these two tests I began forming two theories:

1) Shadow Focus is actually granting 10% increased hit chance to shadow spells.
2) The initial application of the Mind Flay debuff does not count as a shadow spell for hit purposes (and, as such, does not benefit from Shadow Focus).

The first theory will need significantly more data in order to prove, but anecdotally, and in the limited amount I've gathered so far, it seems to fit. The second seems like it could be considerably easier to prove/disprove. I stacked as much hit gear as I could and got up over 14%, which, with the 3% from misery, should put me over the cap regardless of the value of Shadow Focus. I then proceeded to chain cast mind flay. After about 100 casts I hadn't encountered a whiff yet. I backed my hit back to 11% and began chain casting again and hit a whiff after about 25 casts. I'll have to do some more testing, but I wanted to get these theories out there so that people could start running their own tests and possibly prove/disprove them.

Heh interestingly before I saw this thread here I made a post on shadowpriest.com that reads almost exactly like the one above. I ran almost exactly the same tests and came to the same conclusions. Not sure how to link to specific posts on that forum so I'm copy pasting it here, hope mods don't mind.

Ok I tested this fairly extensively today which I think will shed some light on what's going on. I tested on the Heroic training dummy which I'm assuming does indeed behave as a raid boss, haven't heard anything to the contrary at least? I had zero hitrating for this to find out what the actual missrate for us vs a boss is. Normal spec otherwise i.e shadow focus and misery. So an expected 11% missrate.

Here is the wws parse. That run was actually 40minutes but forgot to turn on combatlog at first so had even a bit more data in recount:
2453s
MF 38/1197 3,2%
VT 6/165 3,64%
MB 10/299 3,3%
SW:D 1/157 0,6%
Total 55/1818=3,03% miss

Did another a bit shorter run yesterday that I didn't log but noted the recount data:
MF 28/723 3,9%
Vt 4/85 4,7%
MB 4/151 2,6%
Total 36/959 =3,75% miss

Both combined 91/2777 = 3,28% miss

Paladins are experiencing a similar thing which they think is caused by that their hit talent is giving them the old pre 3.0 higher value. These numbers do support the same thing for us, that shadow focus is currently giving us 10% instead of 3% hit. This pretty long test is actually under the expected 4% miss in that case.

The second issue, the failed mind flays with no resist message. I had this happen what felt like a lot on these tests. Many times back to back and one time 3 times in a row. But hard to put a number on with no actual data.

Next to see if these misses or failures were related to hitrating I put on 394 hitrating. With this I ran a 9minute test before I ran oom and got bored, and did not experience this one single time. So I'm pretty confident those failed mind flays with no message are just misses that doesn't show up. But in that case does mind flay now both give these invisible resists on the application in addition to that each tick can individually miss also? That doesn't seem right in that case... But not sure if there is perhaps something more to this also.

Next test would be to put on close to 105 hitrating(4%) and see if we indeed get zero misses then, either normal or these invisible mf resists.

Did part 2 of this testing. Wearing 107hit (4.08%). WWS.

Recount breakdown
Test3 947s 107hit 4,08%
MF 756
VT 495/5=33
MB 165
SW:D 91
DP 1
0/1046 = 0% miss

So while I guess 1046 events aren't enough to say 100% sure it's good enough for me at least to assume a hitcap of 107 or less at the moment.

The invisible mind flay resists did however still happen, fairly often. Tested again with 293 hit, so just over the non bugged 11% hit cap and it still happened although what felt like considerably less. And like I wrote in the previous post with 394 hitrating I did not see it once in 9minutes, before I stopped.

So while I can't be certain my theory would be that mind flay has double resist checks one at first on the application which you don't see in the combatlog that appears to be working of some strange hit value. Possibly not taking shadow focus in account at all, so maybe 14% missrate there then. Then every other spell including the individual mind flay ticks are using a bugged hit value which appears to be giving 4% missrate or less.

If you assume this to be true and want to optimize gear for while these bugs last I guess we need to figure out how much hit is worth if it is only affecting the MF application and no other spells (assuming you're not managing to get hit as low as under 105).

Strange that we both ended up at almost the same 3,2% missrate and not 4%. Any statistics math guy can figure out what the likeliness of getting a 3,2% instead of 4% value over roughly 4000 events are?

Additional information since. I tested with a level 70 on ptr (EU character copy broken) and mind flay is giving a miss event on the application there. Attacking the +10 lvl 80 dummy you clearly saw almost all mind flays missing, and when one landed the individual ticks would also miss. So that spell definitely has 2 resist checks, just the one for the whole spell is currently invisible on live and appears to be not benefiting from talented +hit which explains why people are seeing them miss while over the expected hitcap. I'm betting this same thing is happening with other spells also like I see warlocks are reporting this with immolate and paladins with judgments I believe?

Physicist · 12/16/08, 9:49 AM

Originally Posted by Asmo

Strange that we both ended up at almost the same 3,2% missrate and not 4%. Any statistics math guy can figure out what the likeliness of getting a 3,2% instead of 4% value over roughly 4000 events are?

As a rule of thumb, error goes as the square-root of the sample size. So square-root of 4000 gives an error of +/- 63, or about 1.6%. So 0.8% discrepency is not distinguishable from random noise.

deadjon · 12/16/08, 10:44 AM

Originally Posted by mek

The Mind Flay "whiff" bug is a separate issue, as it has been observed to occur regardless of level of spell hit.

As you can see from my armory I'm well over hit cap currently (like 309, misses some raids, and some meh drops..) however I still see this "whiff" issue several times per raid. Do we know what is causing this? What can I do to help figure out this issue? My WWS show 0 misses on Flay, but I'm unsure if that's WWS issue or combat log issue. I cast flay, GCD is triggered but nothing comes out.

Asmo · 12/16/08, 12:30 PM

Originally Posted by deadjon

As you can see from my armory I'm well over hit cap currently (like 309, misses some raids, and some meh drops..) however I still see this "whiff" issue several times per raid. Do we know what is causing this? What can I do to help figure out this issue? My WWS show 0 misses on Flay, but I'm unsure if that's WWS issue or combat log issue. I cast flay, GCD is triggered but nothing comes out.

Read mine and Malekithe's posts above. There's some fairly good evidence for that the mind flay whiffs aren't getting any benefit from shadow focus and has a 14% missrate with misery.

Physicist: Ok thanks, in other words nothing to suggest it being even lower than a 4% missrate then.

Physicist · 12/16/08, 12:35 PM

That's correct. 3.2% +/- 1.6% is consistent with a 4% miss-rate (and a 2% miss-rate for that matter). You need more data to measure these things more precisely.

SentinelBorg · 12/17/08, 8:11 AM

Originally Posted by Rhaegal

Unfortunately it's not as easy to "autocast" against a target dummy and go afk for a few hours as a caster like physical DPS can do, but I've asked all of my caster friends who are raiding already to start regularly recording their combat logs.

Actually there is. Enhancement shamans have their Flametongue weapon enchant, which uses spell mechanics including spell hit. You just need two fast weapons, both enchanted with that buff and you can autohit the dummy all night long.

Handyhoof · 12/17/08, 8:28 AM

Flametongue would fall off in 30 minutes.

deadjon · 12/17/08, 1:27 PM

Originally Posted by Physicist

That's correct. 3.2% +/- 1.6% is consistent with a 4% miss-rate (and a 2% miss-rate for that matter). You need more data to measure these things more precisely.

Naxx25 full clear. Flay had a .6% miss rate.

Wow Web Stats

However this does not explain the "whiff" mentioned above when you cast Mind Flay, no "miss" shows up in combat log (or scrolling text) and uses up a GCD.

Physicist · 12/17/08, 8:53 PM

There were 918 hits. That gives you a measurement of 0.6% miss +/- 3.3%. Maybe you're just outside 4% miss-rate, but it's sure not conclusive. I wouldn't call this a statistically significant miss-rate. You just got lucky.

SentinelBorg · 12/18/08, 3:38 AM

I started a test with two 1.3 speed daggers, dual Flametongue and 6.98% spellhit at the heroic dummy yesterday.

When I finished it, I had over 7000 Flametongue Procs and a miss rate of 10.3%. I will resume this test run tonight and then post the WWS here.

Karoshi · 12/22/08, 6:40 AM

Here's a log over 6 hours, 24,9 million damage with 150-160 spellhit, 0,3% miss on MF and 1,2% on Mind Sear (both on trash as far as i can tell), 0% miss on anything else. http://wowwebstats.com/o5a5rbhi5fv4q?a=x1680acd
There was a bunch of MF whiffs, I'd say around 1-2%, but that's just a guess.

This log from the week before was recorded with 298 spellhit Wow Web Stats
MB 1,5% miss, MF&MS each 0,7%

Something's very wrong with shadowpriests...

Starfox · 12/22/08, 6:54 PM

Originally Posted by Karoshi

Here's a log over 6 hours, 24,9 million damage with 150-160 spellhit, 0,3% miss on MF and 1,2% on Mind Sear (both on trash as far as i can tell), 0% miss on anything else. http://wowwebstats.com/o5a5rbhi5fv4q?a=x1680acd
There was a bunch of MF whiffs, I'd say around 1-2%, but that's just a guess.

This log from the week before was recorded with 298 spellhit Wow Web Stats
MB 1,5% miss, MF&MS each 0,7%

Something's very wrong with shadowpriests...

Be warned, the 'Missed%' column in WWS is not the same as the (NOT shown by default) 'Miss%' column
Missed%: miss + evaded + immune (+dodge+parry)
Miss%: miss
If you activate the 'Other Miss' column you see that your Missed% is from evade/immune kind of events, not actual misses. So in fact you did not get a single miss in BOTH wws reports, weird.

Karoshi · 12/22/08, 8:47 PM

Just to add another log (though I doubt there needs to be any more evidence for weird spellhit things going on with sPriests) Wow Web Stats

21,5 million damage with 170 spellhit, 0% misses. I counted 9 MF whiffs over the 52 pulls. 6 of them on Shardron, 3 of them in a row.

What seems pretty interesting to me is that Shadow Torment missed with 17,5% (yes, I should be more awake and not try to kill myself... I know). I guess it just totally ignores any player's stats like spellhit (and class), but that would again hint to unchanged miss-values.

dukes · 01/08/09, 8:16 AM

For those that haven't seen it:

During the transition from patch 2.4.3 to 3.0.3 the following talents were not unlearned properly by some players, resulting in those players still gaining the benefits despite the talents no longer existing in the talent trees. As a result, we've pushed a hotfix which has corrected this matter and affected players will longer receive credit for having these talents.

Warlock: Improved Lash of Pain
Warlock: Fel Stamina
Shaman: Totemic Mastery
Shaman: Improved Lightning Shield (Rank 3)
Paladin: Precision
Rogue: Fist Weapon Specialization
Priest: Shadow Focus (Ranks 4 and 5)

World of Warcraft - English (NA) Forums -> Hotfix: Talent Issue

This pretty much confirms the issues in this thread with spell hit, although whether Mindflay still shows strange behaviour is another matter.