WAR, what is it good for?

[jstraw]

<!--quoteo(post=81711:date=Mar 5 2010, 12:56 PM:name=leonardsipes)-->QUOTE (leonardsipes @ Mar 5 2010, 12:56 PM) <{POST_SNAPBACK}><!--quotec-->In the data I provided, 2008 stats did not seem much better than blind luck at predicting 2009 performance. We just take it for granted that stats are the best indicator of future performance. But how well do they work? They probably do not work better than human evaluation. Milb stat will be worth even less. 50 players hit at least .300 in high A ball (where Starlin Castro played most of the season) last year. Both Eric Patterson and Andres Blanco had better stats in AAA than Castro did in high A. Stats are not the reason Castro is the hot kid in camp. It is because human evaluators have told us he is.<!--QuoteEnd--><!--QuoteEEnd-->

How about a look at the 2008 numbers for the best or worst players in 2009 from around the league?