Haste makes... headaches?

Haste, how I love thee, let me count the ways. Fast.

How much haste is good for a Boomkin? Is spell power better? How about crit? Or Int? Well, many people are fond of plugging in their stats at MaxDPS.com to get an idea of what is going to give them the most bang for the item budget buck. Its a pretty cool website, I use it often.

However, don't let it run your life. Because it says that one piece of gear is going to give you a 1.55233332555 DPS increase over another doesn't mean it is so. If one piece of gear shows as having a 45 DPS increase over another, yeah, its probably better. But always check the stats, and figure it out yourself. God gave you a brain, use it!

The most interesting part of MaxDPS.com, at least to me, is the little thing on the right that shows you how much DPS 10 points of each stat will give you. At least in theory. Its pretty cool to give you a general idea, but again, use your brain.

So, on to the whole haste thing. If someone can clear this up a bit with some solid math or extensive testing, please do! No, posting that "I red 1 tiem on teh WoW forums that hastes r rediculously gud" doesnt count.

A fellow Druid and I have been hashing out how to maximize our laser turkeys a lot lately, and the results have been excellent. Both of our DPS has gone up a big chunk quickly. We got talking about haste, and the more I got in to, the stranger it got. Here is a PM I sent:

Haste: (WowWiki says that 32.79 haste rating = 1% at level 80.)

Lets consider a spell with a 3 second cast. If I get enough haste to cut that down 10%, I am shaving off .3 seconds for a 2.7 second cast time. 10% more haste got me exactly 10% faster than what I had.
Now lets say I already have a massive chunk of haste, say enough to make that same spell a 1 second cast. I add an additional 10% here and I shave off another .3 seconds for a .7 second cast time. 10% more haste got me 30% faster casts than I just had!
Looking at it this way, haste would be more valuable the more of it you already have, at least until you reach the GCD.
This would be a logarithmically scaling valuation for stacking haste.

Looking at it from another perspective, mana issues aside, casting 10% faster *should* net exactly a 10% damage increase. This would be a linear valuation for stacking haste.

I was going to write about other stats, then compare to haste, to arrive at some relative value for each at various levels of spell power, crit, etc. However, the napkin math I just laid out about haste seems enough of a brain twister for now!

I checked the EJ forums briefly and didn't find anything about this. It would almost seem that haste would have to have built in diminishing returns to *not* be logarithmic in it's scaling. IE, if you have 0 haste, 32.79 haste gets you 1%, but if you have 500 haste, maybe 60 haste is needed to add another 1%. However, I have not seen anything written by anyone that indicates this is the case. This would probably be easy enough to check though. Take off haste gear, see what the Stafire cast time is in the tooltip. Add a little haste, check it again. Add a ton of haste, check it again. Then compare the results. Maybe I will try this later!

My brain is going to explode.


Anonymous said...

Not quite right because the haste formula for spells goes as follows :
cast_time = base_time / (1 + haste%)

From 0 haste to 10% you are indeed casting 10% faster. From 10% to 20% however you get :
1.2/1.1 = ~9% faster
and it decreases the further you get.


Anonymous said...

Woops, i dont think i was comparing the right things.
Still, if you compare 1/1.1 and 1/1.2 (cast time with 10% and 20% comapred to original cast time), you'll see that you aren't gaining the same amount of time by adding 10% to the first 10%.

Anonymous said...

Who knows where to download XRumer 5.0 Palladium?
Help, please. All recommend this program to effectively advertise on the Internet, this is the best program!