clock menu more-arrow no yes

Filed under:

A wOBAr Update

New, 26 comments

Before the update, a refresher on what exactly wOBAr is. What I attempt to do with wOBAr is take a hitter's batted ball types and use his own past success rates on each type to generate an expected wOBA line. It judges the player against himself when possible and substitutes in league average rates to fill in the gaps, sort of like the frog DNA in Jurassic Park. If you want a more detailed process, continue through the jump.

Last year's Mariners ended up with baseball's biggest difference between their park-adjusted wOBA and their wOBAr. The team fell about 86 runs short of where their batted ball tendencies would have led you to believe. That shouldn't be that surprising. That was not an offense assembled that was going to true talent its way to just 513 runs scored.

Casey Kotchman was the poster child for last year's team. Despite a career normal line drive rate, a slightly above career average ground ball rate and a vastly lower infield fly rate, Kotchman's BABIP was just .229 despite a career mark in the upper .270s. Kotchman wasn't a good hitter last season, but he was significantly robbed of a number of line drives. The American League had a .716 BABIP on line drives in the 2010 while Casey was at an amazingly low .507. He lost 37 points of batting average that way alone. Dummy shouldn't have picked 13 to wear as a uniform number.

This year's team is off to a rousing start at following their past selves' footsteps. The Mariners right now are -18.5 runs below league average according to park-adjusted wOBA, but only 5.2 runs below by wOBAr. Their 13-run spread is third in baseball behind the Braves and Nationals. The "luckiest" team so far has been the Reds and then the Astros which is hilarious because the Astros are bad enough as it is.

Unsurprisingly, topping the list of unlucky Mariners so far in 2011 is Chone Figgins. Ryan Langerhans, Michael Saunders, Miguel Olivo and Brendan Ryan are all down so far as well. Justin Smoak and Milton Bradley are the only two Mariner hitters going the other way, but neither by a large amount. The takeaway point is the 5.2 figure above. The offense has been lackluster at scoring runs so far. They're not going to transform into a potent offense, but ending the year 50 runs below league average is a lot more tolerable than 187 would be. 

wOBAr Process in a nutshell:

The first step is to generate league average batted balls rates using data covering the 2007-present time frame. This will be the base line used to regress individual players back to in the event that do not have a large enough sample. For example about 32% of all batted balls are classified by MLBAM as a ground ball. Roughly 22% of those have gone for a single, 2% for a double and 3% for an error. 

Repeat the above step but for the player in question. If the player falls below the sample size that I set, his rates get mixed (regressed) with the league rates until the sample size is sufficient. The rates are then park-adjusted and then run through the wOBA calculator again to form wOBAr. The only inputs affected are the players single, double, triple and reached on error rates. His walks, hit by pitch and home runs are left as they are.