Anybody got some info on streakiness?

[M Skjellyfetti]

Just curious if anyone has any links to an article or two comparing the actual effect of streakiness. For instance, a comparison of a guy who hits with an OPS of 1.000 every other month and .600 every other month compared to a guy who hovers within .050 points of .800 every month. Just looking for some legitimate evidence to prove to others that streakiness doesn't really affect much of anything.

[1992casey]

I'd also like to see information which can tell us whether or not a player actually is streaky.

[rluzinski]

As for measuring streakiness, the book "Curve Ball" formulated a way to quantify it using running averages (Bill James did something similar as well).

 

Here's a graph of Jenkin's 10 game running batting average last year:

 

http://photos1.blogger.com/blogger/7391/569/400/jenkins.0.jpg

 

Here's a graph of a randomly generated running average or a player who has a 29.2% chance of getting a hit average AB (Jenkins batted .292 last year):

 

http://photos1.blogger.com/blogger/7391/569/400/robojenkins.jpg

 

ABB: Jenkins and Consistency

 

I explain some of the methodolgy in the comments section but haven't finished the study yet. Some day soon I hope. Jenkins certainly more inconsistent than expected last year but not by as much as some think.

 

As for the benefits of a hitter being "consistent", I'm not convinced it was much value anyway. 9 typically inconsistent players will average out to a relatively consistent offense anyway. I know of no studies looking at the topic but it could easily be done with a simulator.