Rating Changes

We are taking a look at the ratings decay as you gain levels. Now that the talent trees have lost so much crit, haste and hit, they may be too steep.

That is a quotation from a response to a question about rage generation. The rest of the response is about using rage generating abilities and such and really isn't what I am interested in here. What I am interested in is how the decision about how much rating you need to get 1% crit, hit, or haste is made.

I think we knew this already, but its a little distressing to see it spelled out like this. Apparently they decide things in this order: 1. figure out how big the numbers are going to be on gear; 2. decide how large they want people's crit and haste percentages to be; 3. choose rating conversions that give the desired crit and haste values from the number of rating points people will have.

Absent in that process, is the step where they consider whether crit and haste ratings are good enough to even both putting them on gear. The problem is that if you know how big the numbers on gear will be and you know how much crit you want people to have in the final tier of gear then it is a straight calculation to figure out how many rating points it takes to get 1% crit, there is no room for the value of the rating to come in.

As I've pointed out before, the current rating conversion numbers are going to make it pretty much impossible for crit or haste rating to be useful at all. With an imagined scenario using Arcane Blast I came up with a valuation that put crit at 20% of spell power. But lets look at a more realistic scenario.

From my estimates, a level 80 mage in first tier Cataclysm raid gear should have about 6421 spell power (nearly 1600 less than the Priest because of Inner Fire). A level 85 fireball does 671 to 855 damage, or an average of 763 damage. Now I'm not sure that spell coefficients will actually resemble what they are now in many cases, but fireball isn't a bad bet to still work on the cast time divided by 3.5 model. Since the new fireball is a 2.5 second cast, that would mean that the fireball would do around 5702 damage, about 86.6% of which comes from spell power. It will actually be higher from mastery and talents, but anything that multiplies the entire damage of the spell affects damage from spell power and crit equally, so we can ignore it.

The mage will also have about 5,099 points in ratings. It will take 2611 rating to be hit capped, so presumably the mage will be devoting as much rating to hit as possible. Being a fire mage with big fire mage crits we'll assume that crit is supposed to be an attractive stat. So lets hit cap the mage (a little unrealistic, but that's okay) and split the difference 2:1:2 between crit, haste and mastery. That gives 995 crit rating which is 3.71% crit. Assume a base crit near 5% and a 5% extra spell crit debuff on the enemy. That brings us up to 13.71%. Since the mage crits for 2.1 times normal damage (as far as we know at this point, but this could be as high as 2.8 given current information I have available) the average damage of fireball with crit is around 6561.

So what happens if we gain 100 spell power vs. 100 crit rating. 100 spell power would increase the damage of our fireball by 82.2. 100 crit rating would increase our crit chance by .37% or about 23.4 damage. That puts critical strike rating at about 28.4% of spell power for damage for a fire mage. This is not the vision of bringing stats closer together that they spoke of.

Reevaluating ratings to achieve some goal for critical strike chance might improve this situation a lot. Ratings will increase about 63% over the course of the expansion, but if you stay hit capped then that means a much more than 63% increase to critical strike rating. If we end up with 8311 rating, devote the same 2611 rating to getting hit capped and divide in the same ratio then we end up with 2280 rating, more than double the first tier amount, for a total of 8.51% crit from rating, giving 18.51% crit total. Also spell power increases to 9767.

100 more spell power gives 85.9 more damage with these stats, while 100 crit rating gives 28.9 damage. At top tier crit catch up a bit to be worth 33.6% of a spell power. If they decide that mages critting roughly 30% of the time is a good amount, they could drop the rating required for 1% crit to 114, less than half its current value. Then crit would be worth nearly 79% of a spell power to a fire mage. Not bad, really. If spells in general go to 100% bonus crit damage instead of 50%, crit could actually surpass spell power for fire mages in end of expansion contant.

Which is a pretty nice thing to luck into for the developers, but from the process that used to make the decisions about how much rating it takes to get 1% crit, haste or hit, we know that it is just luck if the numbers end up being close together. Rating values should be set to make stats balanced against one another, because stats that are close in value are the right thing for the game. If doing this would end up meaning that players will end up with 150% crit, and it might, depending on how you set the values, then there are other knobs you can turn to fix that problem, such as increasing the base damage of spells, giving more innate attack power based on level, or just having players' base stats account for more of their damage than they do. In classic your base stats from your race, class and level might have accounted for a third of your total stat value. In wrath they aren't even a tenth.

The desired relative value of stats should determine the rating conversions. Something so arbitrary as rating conversions should not determine the relative value of stats. It looks like, for at least some classes, we might end up with reasonably balanced stats after all. Ultimately, though, I wouldn't expect non-red gems to auction for much over their vendor value.