Monday, December 19, 2011

Looking back at the 2011 ZIPS projections

So earlier today, Dan Szymborski posted the 2012 Astros ZIPS projections which, as clack says at TCB:

ZIPS is one of the more respected systems for projecting player performance. I won't say it's the best system, because that is hard to determine. But the godawful Astros' projections of the last couple of years have identified some problem areas for the roster that would materialize during the year.

Let's see what ZIPS projected for 2011 vs. what really happened (selected players):














NameSetBA/OBP/SLGOPS+K:BB
PenceZIPS.277/.326/.463111113:44
Pence2011.314/.370/.502138124:56
LeeZIPS.275/.319/.45910850:35
Lee2011.275/.342/.44611760:59
WallaceZIPS.261/.324/.41498129:34
Wallace2011.259/.334/.3699691:36
BarmesZIPS.245/.299/.3858487:28
Barmes2011.244/.312/.3869388:38
MartinezZIPS.272/.323/.39593115:37
Martinez2011.274/.319/.42310448:13
SanchezZIPS.258/.306/.3317280:31
Sanchez2011.240/.305/.2856544:27


I'm not committed enough to look at the pitchers, but we can break this down: ZIPS-projected ERA+ and actual ERA+:

Wandy: (ZIPS) 109, (Actual) 109
Myers: 101, 85
Happ: 94, 71
Norris: 91, 100
Lyles: 84, 71

So take from this what you will. No projection system is perfect, as it can't figure in playing time. But ZIPS is historically pretty good.