Understanding crit differences

galahan · August 20, 2008, 06:19 PM

I've been trying to figure out why this other warlock does so much more damage than me on raids. Some reasons are obvious: he has more haste and more damage. But it still didn't explain everything. I found that he was casting a lot more than I was so I focused more on the quartz bar and really boosted my casts. The only difference now I think it due to haste.

But he's still doing a lot more overall damage, which I don't think is fully explainable by gear. One thing I noticed is that he crits a lot more than I do. As an example on one boss fight I cast 24 sbs, he cast 28, but while i only crited for 14% of my bolts, he crited for 40%. (we're both hit-capped btw). I cast 3 immolates because I'm hybrid destro (5 pts in improved immolate rather than cataclysm; the rest is shadow), and crited 33%. He's pure shadow and didn't cast any immolates.

But that's kind of an extreme example. Overall for the for all bosses in the raid he crited on 28% of his sbs, while I did on 20%.

Now the thing is we both have similar tooltip crits. He has 20.05% and I have 20.97%, so I'm actually a little higher. Neither of us were using any crit buffs and we were in the same group within the raid. One thing I found was that he has round 42 more intellect than i do than I do. I found the crit formula of: (Intellect/82) + (Spell Critical Strike Rating/22.08) + Class Specific Constant. But that only represents a 0.5% difference, and I think that might actually be included in the tooltip figure.

So after this long boring dissertation, my question is, why is this person criting significantly more than me in percentage terms. I can understand him doing more damage due to his haste and spell damage advantage, but why the crit difference?

Thanks

Solarus · August 20, 2008, 06:31 PM

2 things in order for us to do the analysis:

An armory link to both of your characters.

A WWS report.

If what you're saying is right and both of you have 20% tooltip crit then that is what you should be getting over the duration of a raid, which i'll assume is 1000 or so shadowbolts (big enough number to even out any differences).

The only way that we can prove otherwise is by statistical evidence.

levinho · August 20, 2008, 07:54 PM

Welll I don't really consider the overall crit for a whole raid to be an interesting mark of performance because you rarely look at an overview to see exactly how someone performed. You should probably look at each fight individually in WWS. Crit can be frustrating because for example I had a crit rate of 20% on Rage and a crit rate of 30% on azgalor. My Crit rate buffed and with devestation should come out to be 31-2% (with AI an MotW etc. . .) Most often on an individual fight that doesn't happen.

The only thing that shouldn't be included in the TT for crit is the 5% for devestation. Other than that your TT does factor in the contribution from your intellect and your crit rating and talents like backlash which do all Trees for spells. Theoretically if you are both around 20% TT then you should both be doing about 25% crits on shadowbolts.

Hmm 24 casts sounds like a wipe to me. It isn't a large enough sample size for your crit rating to approach the TT value. Remember that crit is a function of a Random Number between 1 and 100 that rolls every time you cast. There is never a guarantee that you will get a crit on any given cast. Since it is actually a statistic and not a 1/4 will do more damage thing, you will only find your crit rating when you start to approach a sample size beyond the number of digits which are significant (TT is rounded to hundreths place); therefore, you can only find your crit rating accurately with thousands of casts done. Since in most fights you will be doing only a hundred or so casts you will never see precision and accuracy of crit data for those samples.

galahan · August 20, 2008, 08:11 PM

I can't share WWS reports as per guild policy but if there are any relevant stats I can copy them here.

Levinho, the 24 casts wasn't a wipe. It's just that it was Akil'zon in ZA on a timed-run. We finished the timed part with 13 minutes to spare so the boss fights were pretty quick.

What you say about the lack of statistical significance does make sense though. I guess I need data from more raids to determine whether the rolls just came out that way on this one.

Valimir · August 21, 2008, 03:03 AM

It's just luck, that's all there is to it. This other lock might put out more overall damage if his +spell damage is higher, and he might toss our a few more shadowbolts if his haste is higher but the difference in you crit rate for any boss fight is pure chance.

The way crit works (feel free to correct me if I'm wrong anyone) is that each shadowbolt you cast has (in your case) a 25% chance to crit and the crit chance of your next cast has nothing to do with the previous one. So if your last SB crit then the next one is not mroe or less likely to crit again, the chance remains 25%. In theory this should even out over the course of a fight so you see a crit rate that is in line with your character's crit, but you are not necessarily going to see that. It's like tossing a coin. Each toss has a 50/50 chance of being heads or tails but just because you have got heads 10 times in a row does not make the next toss any more or less likely to be tails, it's still 50/50.

I've seen some very odd examples of this. My worst was a Prince fight where I had 9% crit but I've also had over 40% on others.

Just do the things you are doing. Try and keep up with the other lock in terms of the number of SB's cast and that's all you can do other than improve your gear. I'd consider dropping Immolate from your rotation though.

galahan · August 21, 2008, 09:47 AM

Thanks. Yeah the probability aspect all makes sense. I was just wondering if there was another aspect to it that I wasn't understanding.

Concerning immolate, I think you're right. I modelled my talents after the "hybrid" destro spec in Leulier's (sp?) spreadsheet and it came out as having higher damage than pure shadow but it doesn't seem to be working out. I also wanted more variety in my spell rotation than CoR (other lock is CotE) and sb. Maybe I should go affliction for a little variety at some point.

qwwerty · August 21, 2008, 10:57 AM

yeah its random. last night we were badge run in kara and I am at 25% with devastate and i rocked a 38% crit on curator. Was funny we usually down him at end of 1st evocation he was down with about 25% left. I didn't even realize it. till my GM sent me a tell. needless to say next fight I was at 20% crit.

August 20, 2008, 06:19 PM	#1 (permalink)
galahan Lurker	Understanding crit differences I've been trying to figure out why this other warlock does so much more damage than me on raids. Some reasons are obvious: he has more haste and more damage. But it still didn't explain everything. I found that he was casting a lot more than I was so I focused more on the quartz bar and really boosted my casts. The only difference now I think it due to haste. But he's still doing a lot more overall damage, which I don't think is fully explainable by gear. One thing I noticed is that he crits a lot more than I do. As an example on one boss fight I cast 24 sbs, he cast 28, but while i only crited for 14% of my bolts, he crited for 40%. (we're both hit-capped btw). I cast 3 immolates because I'm hybrid destro (5 pts in improved immolate rather than cataclysm; the rest is shadow), and crited 33%. He's pure shadow and didn't cast any immolates. But that's kind of an extreme example. Overall for the for all bosses in the raid he crited on 28% of his sbs, while I did on 20%. Now the thing is we both have similar tooltip crits. He has 20.05% and I have 20.97%, so I'm actually a little higher. Neither of us were using any crit buffs and we were in the same group within the raid. One thing I found was that he has round 42 more intellect than i do than I do. I found the crit formula of: (Intellect/82) + (Spell Critical Strike Rating/22.08) + Class Specific Constant. But that only represents a 0.5% difference, and I think that might actually be included in the tooltip figure. So after this long boring dissertation, my question is, why is this person criting significantly more than me in percentage terms. I can understand him doing more damage due to his haste and spell damage advantage, but why the crit difference? Thanks

August 20, 2008, 06:31 PM	#2 (permalink)
Solarus Member Character Info Solarus 70 Human Warlock Zenedar Euro PvP Talent Spec: 6/44/11	Re: Understanding crit differences 2 things in order for us to do the analysis: An armory link to both of your characters. A WWS report. If what you're saying is right and both of you have 20% tooltip crit then that is what you should be getting over the duration of a raid, which i'll assume is 1000 or so shadowbolts (big enough number to even out any differences). The only way that we can prove otherwise is by statistical evidence.

August 20, 2008, 07:54 PM	#3 (permalink)
levinho Member Character Info Ceausescu 80 Blood Elf Warlock mug'thol US PvP Guild: Defiant Profile: Blizzard Armory Talent Spec: 56/0/15	Re: Understanding crit differences Welll I don't really consider the overall crit for a whole raid to be an interesting mark of performance because you rarely look at an overview to see exactly how someone performed. You should probably look at each fight individually in WWS. Crit can be frustrating because for example I had a crit rate of 20% on Rage and a crit rate of 30% on azgalor. My Crit rate buffed and with devestation should come out to be 31-2% (with AI an MotW etc. . .) Most often on an individual fight that doesn't happen. The only thing that shouldn't be included in the TT for crit is the 5% for devestation. Other than that your TT does factor in the contribution from your intellect and your crit rating and talents like backlash which do all Trees for spells. Theoretically if you are both around 20% TT then you should both be doing about 25% crits on shadowbolts. Hmm 24 casts sounds like a wipe to me. It isn't a large enough sample size for your crit rating to approach the TT value. Remember that crit is a function of a Random Number between 1 and 100 that rolls every time you cast. There is never a guarantee that you will get a crit on any given cast. Since it is actually a statistic and not a 1/4 will do more damage thing, you will only find your crit rating when you start to approach a sample size beyond the number of digits which are significant (TT is rounded to hundreths place); therefore, you can only find your crit rating accurately with thousands of casts done. Since in most fights you will be doing only a hundred or so casts you will never see precision and accuracy of crit data for those samples.

August 20, 2008, 08:11 PM	#4 (permalink)
galahan Lurker	Re: Understanding crit differences I can't share WWS reports as per guild policy but if there are any relevant stats I can copy them here. Levinho, the 24 casts wasn't a wipe. It's just that it was Akil'zon in ZA on a timed-run. We finished the timed part with 13 minutes to spare so the boss fights were pretty quick. What you say about the lack of statistical significance does make sense though. I guess I need data from more raids to determine whether the rolls just came out that way on this one.

August 21, 2008, 03:03 AM	#5 (permalink)
Valimir Member	Re: Understanding crit differences It's just luck, that's all there is to it. This other lock might put out more overall damage if his +spell damage is higher, and he might toss our a few more shadowbolts if his haste is higher but the difference in you crit rate for any boss fight is pure chance. The way crit works (feel free to correct me if I'm wrong anyone) is that each shadowbolt you cast has (in your case) a 25% chance to crit and the crit chance of your next cast has nothing to do with the previous one. So if your last SB crit then the next one is not mroe or less likely to crit again, the chance remains 25%. In theory this should even out over the course of a fight so you see a crit rate that is in line with your character's crit, but you are not necessarily going to see that. It's like tossing a coin. Each toss has a 50/50 chance of being heads or tails but just because you have got heads 10 times in a row does not make the next toss any more or less likely to be tails, it's still 50/50. I've seen some very odd examples of this. My worst was a Prince fight where I had 9% crit but I've also had over 40% on others. Just do the things you are doing. Try and keep up with the other lock in terms of the number of SB's cast and that's all you can do other than improve your gear. I'd consider dropping Immolate from your rotation though.

August 21, 2008, 09:47 AM	#6 (permalink)
galahan Lurker	Re: Understanding crit differences Thanks. Yeah the probability aspect all makes sense. I was just wondering if there was another aspect to it that I wasn't understanding. Concerning immolate, I think you're right. I modelled my talents after the "hybrid" destro spec in Leulier's (sp?) spreadsheet and it came out as having higher damage than pure shadow but it doesn't seem to be working out. I also wanted more variety in my spell rotation than CoR (other lock is CotE) and sb. Maybe I should go affliction for a little variety at some point.

August 21, 2008, 10:57 AM	#7 (permalink)
qwwerty Member Character Info qwwerty 70 Undead Warlock gorefiend US PvP Guild: Vex Profile: Blizzard Armory Talent Spec: 0/21/40	Re: Understanding crit differences yeah its random. last night we were badge run in kara and I am at 25% with devastate and i rocked a 38% crit on curator. Was funny we usually down him at end of 1st evocation he was down with about 25% left. I didn't even realize it. till my GM sent me a tell. needless to say next fight I was at 20% crit.