|
The usefulness of haste is a contentious issue. While it increases DPS via more casts, some argue that you only see the gain from this if you get an opportunity to make that extra cast (ie: you would need to make 100/haste% casts to see extra cast). [...]
|
The above describes the well known fallacy that haste is only useful for longer fights and/or if you have a lot of haste. The problem with that logic is that it considers fight length to be a fixed value when instead it is a random variable that follows a certain distribution. Because fight length is not static, haste always increases the probability that you get another spell off, even if you only have a single point of haste and even if the fight is very short.
Consider the following example:
Shaman A with no haste and 2.0s lightning bolts and Shaman B with a little haste and 1.9s lightning bolts are nuking the same trash mob. If the trash mob dies after 10.0s, both shamans will successfully cast 5 lightning bolts (for simplicity, we are ignoring the traveling time of the lightning bolt). If the mob however already dies after 9.8s, Shaman A will only cast 4 lightning bolts successfully whereas Shaman B hits with 5 bolts.
The key really to understand the usefulness of haste is to realize that fight length is a random variable and that the effect of haste is therefore probabilistic in nature. Therefore, even a single point of haste is useful, and even when dealing with very short fights.
This fallacy used to pop up once in a while in the mage threads and has gotten numerously debunked since.