I read on Aerys Sports today a post in a series of articles explaining the advanced baseball metrics. One of the comments asked why Carlos Pena was offered and accepted a one-year $10 million deal from the Cubs last offseason. I took it upon myself to write up a response.
cubneil writes: “I’d like someone to figure out what part of the .190 Carlos Pena hit in 2010 qualified him for $10m in 2011, when he ballooned to .225.”
Yes, there was many a baseball folk that cried foul when the Cubs signed 1B Carlos Pena to a one-year, $10 million contract before the 2011 season. After all, when should a batting average under the Mendoza line be worth an eight-figure salary? As the title of Megan’s post goes, the devil is in the details. Here I plan to debunk the fact that the Cubs were blowing money away money when they signed Pena through two routes: (a) the Cubs weren’t paying for Pena’s meager batting average, and (b) 2010 was a very unlucky year for him.
First, if you recall an earlier post, we concluded that batting average was an outdated and inaccurate statistic for determining a player’s offensive production. Today I’ll tackle Pena’s other metrics to justify his contract.
If you include his entire 2010 slash line of BA/OBP/SLG, Pena hit .192/.325/.407, and we’ll look at the latter two first. Obviously the .192 BA wasn’t a good indicator of Pena’s run producing ability. The big difference between batting average and on-base percentage immediately tells me that Pena might be drawing a lot of walks. A quick check to my favorite baseball stats site (www.fangraphs.com, on Twitter at @Fangraphs) shows that Pena walked 87 times in 582 PA, for a BB% of 14.9%. In 2010, this would fall above the 90th percentile in all of baseball. When you’re better than more than 90% of your contemporaries, it’s not easily replaceable. Further, Pena’s OBP was above the 2010 AL average OBP (.322), giving us an OBP+ of 101 (this of course is a primitive measure as it’s not adjusted due to strength of schedule and park factors).
Next, we see Carlos Pena slugged at a .407 clip in 2010, implying an OPS of .732. The .407 was exactly in line with the league average for the AL in 2010, and his OPS was 2 points lower than league average (.734). However, Baseball Reference is kind enough to make OPS+ freely available, as well as the park factors that go into the calculation. Tropicana Field has historically been a pitcher’s park and 2010 especially so, with a Batting Park Factor of 92, meaning based on environment alone, the Trop produced about 92% of the offensive production compared to the MLB average. By comparison, historically hitter friendly Coors Field had a Batting Park Factor of 118. Getting back to Pena, after adjusting to park factors, Pena’s OPS+ for 2010 was 103. So by this metric, he had a slightly above average season.
Now we get to the real meat of the issue. Pena’s BABIP in 2010 was a paltry .222, a career low. Keeping in mind that he was entering 2011 with a below average career BABIP (.279 in 4295 PAs from 2001-2010), the .222 indicates an extreme bit of bad luck. As a very crude form of regression, we will say that Pena’s BABIP is 75% explained by the league average (using Megan’s .300 from above) and 25% “BABIP skill.” That is, Pena has 25% percent control over this BABIP.
Using those averages, Pena’s 2010 mean BABIP should be .295; so if we were to simulate Pena’s 2010 season hundreds or thousands of time, under these assumptions we would expect him to average a BABIP of .295. The .222 figure that was actually recorded (.222) was just 75% of what we expected. Now I don’t have the time to calculate BABIP variance, but will say that Pena’s .222 was WAY less than expected, and we should expect a statistically significant change.
Lo and behold, Pena saw a bump in his 2011 BABIP (.267), and a subsequent bump in his slash line (.225/.357/.462), and a WAR, depending on where you go for it, of 2.6 (FanGraphs) or 2.2 (Baseball Reference), 3rd on the team behind Ramirez and Castro. So on a $10M salary, the Cubs were paying between $3.84M and $4.55M for each additional win that Pena brought to the team. Among the blogosphere the general rate is $4M to $4.5M per win above replacement, so you could argue that Pena was properly or even underpaid for his services in 2011.