Why does Combustion have a 3 min cooldown?

There are two slightly entwined problems here. One is Combustion just not keeping up with other mage talents. That's a problem we want to fix. While there are plenty of examples of talents that aren't as popular as they should be, we consider Combustion to be a pretty core talent to the Fire mage experience, and thus we consider it a bigger problem. (And no, we aren't ready to announce a change yet.)

The other is that character stats are much higher than we originally predicted. This is less of a big deal for things like Str and Sta that can just keep growing, but it gets messy with the ratings like crit and avoidance. While more crit is generally always better, you do run into weird scenarios when players can realistically think about 100% crit chances (if not now, perhaps in Icecrown).

The reason the stats got so high is that the Ulduar model of gear (where hard modes offer higher ilevel) was something we added after ship and therefore after the core combat mechanics of Lich King were already in place. We designed the system with lower numbers from gear than we are ultimately going to end up offering. You might be able to guess at the highest ilevels in Icecrown and those are higher than we originally planned on them being. The alternative was that hard modes didn't offer better loot. :(

Anyway, we can tweak the combat formulae to some extent, but short of "Icecrown Radiance" that drops your crit by 10%, some of these issues aren't going to go away until the level cap increases and we can adjust all of the combat ratings again.

Q u o t e:
This was a fantastic post. Why can't other issues gets post of this magnitude letting us know what the devs are thinking.

Scourge Strike comes to mind. Atleast you are letting mages know that Combustion while probably not in line for a fix in WOTLK is on the radar to be fixed in Cata. Why can't something along these lines be told to DKs in terms of diseases cleansing and Scourge Strike?

Good posts are in the eye of the beholder. Read the responses above yours. Several of them conclude that we don't know how to make gear or we hate mages. :)

[Not tracked]

Q u o t e:
Inflation is a nasty sounding word, and your argument predicts a dps cataclysm, but where exactly is gameplay and balance suffering atm? - besides Combustion and a few other 'outliers' (to use GC's favorite word).

Some stats can grow forever and yet the abilities they affect generally behave the same way. Things like crit, haste, armor pen and avoidance (generally the numbers that convert into a percentage) start to get weird at very high numbers. We design using the range of those values that make sense. To use an analogy, if your thermometer only goes up to 105 F, that's probably fine for Michigan but lacking in Arizona. The range on the instrument was designed for one range of numbers and might not work for another.

The problem in LK was we ended up growing the gear by more than we expected. Had we known that in hindsight, the rating conversions would probably have been less favorable to begin with so that we'd be talking about mages with 50 to 60% crit rates now instead of 70 to 80% crit rates.

The avoidance rates on tanks are probably more problematic than the crit rates on dps specs -- high avoidance led us to Sunwell Radiance in the past. The alternative is that bosses have to hit so hard that if you ever fail to avoid twice in a row, you wipe (or the bosses are just very easy). Tanks should be avoiding at say 30% not 50%.

Just remember all this if the ratings are even steeper in Cataclysm and you see your crit rate drop with each new character level. The alternative is that you get to 100% crit too quickly.

Addressing another poster, it was never our intention that getting very high crit rates is a way to opt out of the RNG of Hot Streak and other mage mechanics.

Q u o t e:
This was the assumption I was under and stated it earlier. When the 2pc and 4pc was designed, if you were aware of this "crit soak" issue then why make the 4 pcs add in more rng and stack more crit?

A major goal of 3.2 was to implement a fairly simple raiding tier (when compared to 3.1 and 3.3) so that we could focus on development of the new BG. The relatively simplistic set bonuses reflect that general philosophy. Set bonuses are among the most challenging parts of itemization: they are technically hard to implement, it's tough to come up with good ones that you haven't seen a hundred times, we need something that can benefit a variety of specs, and they have to actually be desirable. Perhaps in retrospect, more crit was not the best thing we could have put on mage gear, but at the same time, we need a system where we don't have to avoid giving a class a stat it actually wants. :)

P.S. You shouldn't read too much into my comments above about how various content patches will be developed in the future. We tried an extreme raid with Ulduar and Icecrown will also be a blow-out in that sense. That doesn't mean the pattern will always be large / small / large or anything. We like to experiment with different things to see what works and also so we don't fall into too much of a predictable pattern.

Q u o t e:
It's not that you don't know how to make gear, its that you are offering higher item level gear through T9 content when in fact it was completely unnecessary, and then blaming gear inflation for scaling problems.

It was necessary in our opinion. Just because you would have handled things in a different way doesn't mean it was a misstep on our part.

We have, and have had, a target for the final level 80 mage, and we're going to overshoot that target. Adding additional tiers of gear is a factor, but so are other things such as the big jump that was Naxx gear over blue heroics, more generous talents, the inclusion of glyphs, and consolidated raid buffs. Perhaps in 4.0 we'll include more of a buffer.

In any case, all the blame still falls on our shoulders. I was just attempting to explain why we're in this situation where mages are looking at potential 100% crit rates and then wondering why we'd allow that if it wasn't intentional.

Q u o t e:
For avoidance ratings it looks like pure DR might be the only option, leading to a gearing strategy of "reach the avoidance soft cap / sweet spot, then focus on effective health."

Well, technically that's not the only way to do it. Consider that more advanced tiers of mobs hit harder and have more health but otherwise don't generally improve. Mobs don't scale with gear. :) An alternative design might have the Icecrown mobs with more armor, more expertise, etc. I'm not suggesting we're going to do that, but it would mean that your experience would be a little more linear instead of critting / avoiding the easier mob 50% of the time and the harder mob 70% of the time.

Q u o t e:
That's why I suggest bosses that are even higher levels than the ones we have now, to be the essential check against players' scaling.

What Muphrid says.

Q u o t e:
Is the mage class balanced against having "average" luck or having "very good luck?"

You're balanced the same way all other classes are. We don't take the extreme cases and assume that represents the target number. It's the same reason we don't balance just around best in slot gear -- by the time you have that gear, the encounters are probably a whole lot easier. It's when you don't have the gear that balance matters a lot more.

Q u o t e:
From the development side, knowing the absolute maximum amount of critical strike that can be attained for any class at any time (and for what length of time) is so simple it's tough to fathom how it could get out of hand to begin with.

Just because the ilvl of the gear has gone up doesn't mean the cumulative amount of crit has to go up also. Just look at the T7.5 to T8 crit transition for balance druids.

Something like 'Icecrown Radiance' would be a ham-handed solution culled from the same negligence that created the problem.

I don’t know that Icecrown Radiance is anything we’d actually do. More likely we’ll just live with the current situation. But the fact remains that adding more crit rating as long as mobs have the same chance to be crit will just inflate your chance to crit. Otherwise you’re not really upgrading your gear.

You can call it “negligence” if that makes you feel better, but we understand exactly how we got here and knew every step of the way what was going to happen. If the final raid-buffed mage crit target was 50%, then it would be really lame to be a fresh level 80 mage in Halls of Stone with a crit rate of 15% or whatever. It’s not a case of us just not setting the low end low enough. It’s not a trivial problem to solve, which is why we were content to just go with what we had.

Q u o t e:
id also love to see where gc is pulling these 80% crit rates

Those are a bit of arm-waving numbers, but mage crit rates are pretty high considering hard modes for 3.2 just unlocked and we have a whole tier to go.

Q u o t e:
I have lost count on how many times GC has responded to this issue with "...nevertheless, the difference between a skilled and a bad mage will always show on DPS".

Yup. If players are really beating Yogg in blue gear, it speaks volumes to the degree to which skill and coordination can trump gear (or bad luck).