THE BOOK--Playing The Percentages In Baseball

Interesting, but subject to the following questions, as I don’t think the data directly support the conclusion:

1) relief specialization started in the early 90s/late 80s as far as I am aware, yet the difference didn’t improve till late-90s-2000. Why should that be?

2) with so few data points, how much does the one 2011-so-far (is that weighted appropriately) account for the recent inflection point?

So:

What’s the r value on that chart, and sans 2011?

Could the following explain the findings:
Reliever mania kicked in leading, not to improved efficiency via specialization, but just a shifting of better pitchers to the relief role.
Now (as in just beginning to take hold but not yet), as the trend of having your best pitchers pitchers the most innings and not focus so much on saves, we might except the difference to actually shift back and we won’t have the best pitchers wasted on 60 innings a year anymore.

As if by magic, this just showed up:
http://www.fangraphs.com/blogs/index.php/teams-continue-to-overpay-relievers-why/

So, as long as relievers are overpaid, you’ll get what you find in the above chart—so, an effect due to specialization is unnecessary.