THE BOOK--Playing The Percentages In Baseball

Thursday, October 28, 2010

True Talent v Observations

Here is what JC wrote, in his new book, about Replacement Players:

This is wrong.  For those of you who think he’s right, I’ll explain to you why he’s wrong.  And, it has to do with the difference between true talent and observations. 

Note: I’m going to use “win%” not in its literal W/L assignment rules sense, but as a proxy of his peripheral stats.  That is, convert his OBP and SLG allowed into runs allowed, and convert that into an effective win%.

If you have a bunch of players with a true talent level of a .400 win%, you are going to OBSERVE some of them PERFORM at a below .300 win% level, and you will find a similar number performing at above a .500 win% level.  But, their actual talent level is .400.

What JC should be stating here is that 32% of players were OBSERVED to PERFORM at below the replacement level.  These players are NOT necessarily inferior to the large population of players available for the league minimum.  But, JC neglects this critical distinction.