If the options are being 2 hit below hit cap and gaining 9 spell power vs. using a veiled and being capped but wasting 8 hit, under most gear options the former is the better choice. Hit cap is not something special.

Imagine a fight that lasts for 5 minutes, that's 300 seconds, or 100 fireballs (for the point of the thought experiment we'll ignore everything else). With a 1% miss rate you'd expect to lose one spell during that fight, for a 1% DPS loss. Sometimes you'd miss none, sometimes two, etc, but it would average out to 1% lost.


Now imagine a fight that lasts for 3 minutes, that's 180 seconds, or 60 fireballs. Consider the DPS loss of 1 spell missing in that situation, 1/60 > 1/100, each spell is more important to your average as the fight grows shorter. Extended over many fights you'd average out to the same 1% loss that you observed in a 5 minute fight, but the impact on a 3 minute fight that gets unlucky could mean you underproduce.

Now consider instead of just fireballs, you're also casting HS pyros and LBs, which if they miss are a bigger hit to your DPS than just a fireball. So if you cast 100 spells in the fight, 10 of them being LBs, if you miss one of those LBs it will be a greater than 1% loss for that particular fight.

That's what the concept of hit capping is all about, removing the potential to under perform. The same as increasing your crit is increasing your potential to over perform.

All that said, 1000 spell power vs 1% hit, take the damn spell power. But when it comes down to gemming the last few slots up, hit capping is never a bad idea. There are certain qualitative considerations that are difficult to manage in the quantitative world of theorycrafting if all you live by is the average DPS number, this is one of them.

The question isn't if hit capping is good. Obviously it is. But it is not always optimal. If you are hit capping just for the sake of hit capping, sometimes you will be doing so at the cost of some personal dps. Missing occasionally is not the end of the world, when it is made up elsewhere. That is all we are saying. You shouldn't be "laughed out of a raid" if math and computers are on your side.

Hit is not a magical Jesus stat.

If you are more concerned about keeping up a scorch stack for the benefit of raid dps, hit capping when it isn't optimal would just be another way of sacrificing your already compromised personal dps. I believe RAWR is only concerned about your personal dps, scorch duty or not.

Korey, there is a difference between "Jesus Stat" and "Qualitative difference not handled by average DPS calculations".

Imagine you are in an encounter that requires at least X health, your existing raid gear gets you to within two SP/Stam gems of that limit. So, you swap your SP/Spi gems out to meet the requirement, even though it is a DPS loss to do so. Why? to use an old Molten Core adage, dead mages do no damage. Or, you can focus on the average DPS number, and hope that you get the low end of the damage randomizer when the effect requiring X health happens.

Imagine now you are in an encounter where mana regeneration is key, you find yourself consistently just short of being able to maintain DPS through the fight. You have a blue slot in your equipment that you've been gemming straight SP into, but that has a socket bonus of 5 spirit. You gem a SP/Spi gem into it and that pushes you just over the cusp of being able to maintain your mana for the fight. Or, you stick with the straight SP gem because your average DPS number is higher with it.

Somewhere between these two ridiculous examples of qualitative vs quantitative is the situation at hand, hit.

This argument between individuals focusing on the average DPS number and individuals seeking to not miss is silly. If you aren't bothered by a miss every so often, follow the hardline average DPS guidance of the TC tools. If you are bothered by it, then maintain hit cap.

Neither is the absolute right answer since it's entirely based on your personal preference. Your super averaged out average DPS number is no more the "Jesus Value" than hit is the "Jesus Stat".

How accurate is rawr's simulation models for the Arcane changes coming with the next patch? The reason I ask is because last I checked, from what I understood fire and arcane were going to be on par with single target dps come next patch, but when I put rawr into 3.2.2 mode and switched from 19/52/0 spec to 57/3/11, my dps went from 8700->9300.

Is this correct for next patch? Arcane will be that much better of dps?

Kavan's on vacation for a week, so has not gotten to update it with the fact that Arcane Power is now stacking additively with Arcane Blast, rather than multiplicatively.

Regarding Combustion value I'd trust simcraft at the moment, just because it's using simulation instead of a modeling approach and in this case the model is relatively hard to make accurate. Any insight into what might be causing the difference is welcome.

This is currently not supported. The absorbs in survivability tab are assumed to be external. Rawr currently supports fire/frost ward only in terms of regen. This means that if the regen itself is worth the gcd then it will be used, but it does not consider effect of incanter's absorption on ward.

What is the actual dps when you change number of points in IA? Are you sure you don't have any other differences between the 2 talent specs you're comparing in 'talent spec' comparison tab? IA should not be highlighted in talents chart until you get to 3/3.

Living bomb is only modeled as a single target spell. It's not modeled as an aoe spell. Pretty much there was no work on the aoe modeling for more than a year. If I would start any new work on this I would probably rework the whole thing. The main thing that concerns me is that in different fights the value of aoe on off-targets varies greatly so if I redesign things I would definitely want to take that into account.

Yes the talents chart shows the value of next point. When I had the talent chart as a custom mage only chart I used to show both the value of last point (colored as equipped) and value of next point for partial talents, but that part wasn't replicated when the charts were made general for all models. If you would like that functionality I'd suggest making a suggestion in the general Rawr discussion. If others are not opposed to it I can add that back in.

Version 2.2.19 seems to have a problem with the "arcane cycles" window. Both I and another user from the arcane thread experience crashing/constantly heavy CPU load upon trying to open that window. Is this a known bug or what is it related to?

Arcane cycles is not really something you need to click on yourself, since the stats tab summarises the information for you. It's mainly for Kavan's own use in calculating which spell cycles he should include in the solver.

The reason it may be heavy on CPU and/or crash Rawr is due to the complexity of these calculations, and especially the increased number of states due to AB stacking now to 4 instead of 3. If you are really curious what happens when you press this button, Kavan posted his latest results from it in the arcane thread The Arcane thread

Things changed in 3.2.2. Check the filters as well. If you still have an issue, we have to see your character file.