Wednesday, February 2, 2011

kwERA and bbFIP: ERA Estimators for the Oriole Staff

Lucas Apostoleris of Beyond the Boxscore has a post up about league leaders and trailers in a couple of lesser known ERA estimators, kwERA (strikeout-walk ERA) and bbFIP Batted Ball Fielding Independent Pitching), both created by Tom Tango. Both work like FIP but with different data. kwERA is ridiculously simple, using only strikeouts and walks as criteria and bbFIP is a bit more complex but used various batted ball data to come to its conclusion.

I'll let you read more about the calculations for these stats over at BtB or on Tom Tango's site but Lucas posted a spreadsheet with 2010 kwERA and bbFIP numbers for every pitcher in the league and since I am mildly obsessed with ERA estimators, I wanted to take a look at how a few Oriole pitchers fared.

              ERA     kwERA    bbFIP
Millwood     5.10      4.32     4.71


Kevin Millwood pitched a bit better for Baltimore than he gets credit for. His actual ERA was worse than both of these predictors as well as FIP and xFIP. The defense let Millwood down a bit and he was slightly unlucky perhaps. Again, I'm not saying he was good but he was a sliver better than we give him credit for.

              ERA     kwERA   bbFIP
Matusz       4.30      4.07    3.86


See a pattern here? Defense matters and the Orioles were not great in that facet of the game in 2010. kwERA came pretty close to estimating his ERA but the batted ball data suggests Matusz deserved a lot better from his fielders and Lady Luck.

One would have to think that the infield of Lee-(healthy)Roberts-Hardy-Reynolds would have to be better than the Wigginton-Lugo-Izturis-Tejada configuration we saw for most of last year. And the pitching numbers should be better for it.

             ERA   kwERA   bbFIP
Guthrie     3.83    4.23    4.44


Jeremy Guthrie doesn't give a damn about your ERA estimators.

Like all the others, Guthrie outperforms these estimators too. kwERA comes the closest and it's still .40 off. Kind of interesting that the estimator that uses solely strikeouts and walks is the best at quantifying what he does. I might have to apply it to previous seasons as well to see if there's a correlation.

            ERA   kwERA   bbFIP
Hernandez  4.31    4.11    4.38


He is what we thought he was. These numbers are pretty right on and even FIP does a good job. (Only xFIP was totally out of whack.) So he's an average-ish reliever with some potential upside as a late inning strikeout guy. I'm still OK with trading him for a couple years of Mark Reynolds.

           ERA   kwERA  bbFIP
Tillman   5.87    5.22   5.60
Arrieta   4.66    5.14   5.23


These two are linked due to their upcoming spring training duel for the 5th spot in the Oriole rotation.

As you can see, the two aren't as far apart in performance as their ERAs would suggest. There is a bit of sample bias since Arrieta faced nearly twice the major league batters that Tillman did in 2010. I still like Tillman's progress a little better at this point but...flip a coin with these two. We're likely to get the same 2011 performance from both of them.

           ERA   kwERA   bbFIP
Berken    3.03    3.84    3.52



Jason Berken wasn't quite as stellar as his sparkling 3.03 ERA but he was legitimately very good. Hopefully, he fully recovers from his injury and pitches at that level again.

           ERA   kwERA   bbFIP
Gonzalez  4.01    2.70    2.91



The contract was still not good...but Mike Gonzalez pitched pretty well over the second half of last season. It gives you hope that he can actually get results like he did in Atlanta. These numbers are giving me hope for the bullpen. Speaking of which...

           ERA   kwERA   bbFIP
Uehara    2.86    1.85    2.45



As spectacular as Uehara was when healthy last season, he may have been even better than we thought. All these numbers are really making me question our defensive team in 2010. Was it really that bad? Maybe.

(By the way, Dave Mc of Weaver's Tantrum has a nice breakdown of what defensive improvement can do for a team. Maybe the O's are setting themselves up for a big move for the better on D?)

           ERA   kwERA   bbFIP
Bergesen  4.98    4.79    4.68


That's about right for Brad. But hopefully the Bergy of the second half is what we get in 2011. It's no surprise that Brad fares slightly better with the metric that used batted ball data.

Three things to think about from this data:

1) For the first time, I'm beginning to think we might mave a pretty good bullpen.
2) It appears that our defense in 2010 may have been really, really bad. And that our defense in 2011 looks like it will be a sizable improvement. And the pitching numbers could improve with it.
3) I still can't pin down the intangible quality that makes Jeremy Guthrie so good. (I guess that's why it's intangible...)

No comments: