The rating formulas are designed to give winning expectancies and don't specfically address draw chances.
Rating vs. draw chances
Sort:
Then should the fact of a draw affect the rating of players?
Mar 24, 2016
0
#4
I think high rated players should be expected to draw more often because most end games in relatively blunder free games start with equal material and reasonable positions, and require one player to make a mistake for the other player to win.
Mar 24, 2016
0
#5
All games require a mistake for the other player to win. 
In many games there are multiple mistakes.
As to the question of whether or not draws should induce a rating change, yes, most of the time. Players very close in rating would likely have a low magnitude change and those farther apart would have a larger one. That particular discussion has been on the forums before, at least once.
What is the frequency of draws at various Elo levels, like 1500 vs 1500, 2500 vs. 2500, etc.? Does draw chance vary in any systematic manner?