THE BOOK--Playing The Percentages In Baseball

Is this the first thread on the B James good plays vs misplays system? It’s a long time coming, since he published it a few years ago....

In the short run, any kind of a visual assessment is going to destroy any of the play-by-play fielding metrics.  Many times while watching a game, I’ll see a fielder make a very good or very bad fielding play, and I will wonder to myself how UZR would rate that one particular play.

For example, an outfielder will take a horrible route to a fly ball, and it will fall in for a double (for some reason, official scorers will rarely give an error on an easy fly ball in which the outfielder takes a horrible route as long as the OF’er doesn’t touch the ball).  I will mentally say to myself that the average fielder would catch that ball 95 times out of a hundred.  My guess is that UZR would have the catch rate at maybe 60% or 70% at the most because the hang time for that particular fly ball might be 1.5 seconds but the UZR bucket includes balls with hang times of 1 second and 1.2 seconds, etc.

In the long run, I don’t know how the “objective” metrics would compare to the “visual” ones, and I certainly don’t know what the “long run” even means.

Ideally, without “field f/x” data (position of fielders and hang time), visual and objective evaluations would be best, I think, although I am not sure how one should combine them.

From there you need to have a baseline.  Are those all good fielding 3B or just the ones who play the most?  If good fielding plays are more common than misplays, then it’s mostly a counting stat.  For a guy leading like Longoria or right behind him like Wright, I’m sure it means he’s having a good defensive year.  But I don’t know if Mark Reynolds is actually above average or just durable.

yes, of course you need a baseline to make much sense of these numbers.  Similar to his “scoops.” He never presented a baseline as far as I know, which rendered those numbers almost meaningless (in that case, we didn’t even know how many “scoop opportunities” each 1B got)…

On that note, I LOVE that Fangraphs carries “scoops,” but am disappointed that they don’t carry the opportunities as well.

Anyways, I hope we see more articles using GP/DFM in the future.  This stuff is pretty cool.