The Aging of Bad Outfield Arms

A few weeks back, I took a look at how elite arms age in the outfield.  What was interesting in that little study was that performance peaked in year two and then collapsed.  There was some question as to whether it was merely a regression to the mean sort of phenomenom (although if that was the case the trend should be flat instead of a concentrated and significant increase in year 2).  There was some interest in how poor outfield arms aged and whether it looked like a mirror image of the good arms or if bad arms gain a reputation and will be exploited by runners.

If you did not click on the link above, here is what the elite arm graph looked like.


I used the same methodology of the previous post, but used the nine worst outfield arms as defined by Defensive Runs Saved.

	Runs by Arm per 1400 Innings
	1	2	3	4	5
Andre Either	-9.0	-3.6	-5.1	-6.1	-2.6
Coco Crisp	-3.8	-4.4	-3.1	-1.2	-6.3
Chris Young	-2.2	-6.0	-6.9	3.1	-3.1
Justin Upton	1.6	-1.2	-6.3	-5.1	-3.3
Ryan Braun	1.1	-2.1	-4.2	-6.7	-1.1
Corey Hart	-6.4	-3.1	-3.0	-1.2	-5.2
Jason Bay	-2.9	-9.0	-8.2	-2.3	-4.2
Grady Sizemore	0.0	-3.0	-4.0	-2.1	-8.7
Matt Holiday	-9.2	-4.0	-5.2	-1.0	-2.3

This yields an insignifiant p value (0.39) and the following averages and standard deviations:

	1	2	3	4	5
Average	-3.4	-4.0	-5.1	-2.5	-4.1
StDev	4.1	2.3	1.7	3.1	2.3

What does that all mean? 

I am not sure what this means with respect to this population or the statistically significant differences observed in the elite arms group.  No trends can be measured here or inferred.  Bad arms do not seem to improve as a group and neither are they exploited.  The explanation eludes me.  It may well be that the elite arms group was genuinely a unique occurrence.