THE BOOK--Playing The Percentages In Baseball

Thursday, April 03, 2008

FIP and ERA

Nothing big.  I figured the FIP and ERA of all pitchers since 1994 with at least 60 IP and ran a year-to-year correlation:

That’s 2407 pitchers in all.  Average was 139 IP in each season.

ERA (year 2) = 0.59 * FIP + 1.77
ERA (year 2) = 0.39 * ERA + 2.70

If you use both:
ERA (year 2) = 0.49*FIP + 0.10*ERA + 1.79

As you can see, FIP is far stronger than ERA in year 1, in predicting year 2 ERA.

If I limit it to min 120 IP in each year (average of 190 IP):
ERA (year 2) = 0.54*FIP + 0.05*ERA + 1.84

If I limit it to 40 to 120 IP in each year (average of 70 IP):
ERA (year 2) = 0.37*FIP + 0.11*ERA + 2.18

Basically,you should weight it as 50% FIP, 10% ERA, 40% league average, regardless of number of innings. 

You can bring up selective sampling, but I don’t think that’ll make much difference given these results.