What do we expect from Astros' batters, according to BABIP?

It's the time of year to sum up the players' seasons.  And inquiring minds want to know which hitters are likely to improve on this year's performance, and which are likely to suffer regression.  To suggest some answers, I will analyze the Batting Average on Balls in Play (BABIP) and batting average on line drives. 

A similar review after last season revealed  Hunter Pence, Michael Bourn, and Carlos Lee as likely to have a rebound in 2011 over their 2010 batting stats, while Chris Johnson appeared to be a candidate for significant downward regression.  And it turns out that Pence, Lee and Bourn each enjoyed 30 - 32 point improvements in their batting average. Chris Johnson's 2011 batting average declined 22 points, compared to 2010.  For this article, I will omit players, like Bourn and Pence, whom were traded away, since they will not be part of the 2012 Astros' team.

BABIP

BABIP tells us about the apparent "luck" experienced by batters and hitters on batted balls.  The luck can take the form of fortunate bounces of the ball, good or bad defense, etc.  For pitchers, BABIP is expected to be distributed in a narrow range around league average BABIP.  This is due to the limited extent that pitchers can control whether batted balls are caught or not.  Batters have more control over how hard the ball is hit, and the trick is figuring out the normal or "expected" BABIP for a particular hitter.  Hitters with BABIP far above or below the expected level in 2011 are  likely to see a reversion to their normal level in 2012.

For this article, I have estimated the expected BABIP (x-BABIP) in 2011 for each Astros' batter and compared that to the hitter's actual BABIP in 2011 to determine whether the BABIP can be sustained in the future.  I used a BABIP calculator to estimate the expected BABIP for 2011.  The BABIP calculator is a simplified formula based on statistical regression analyses published in the Hardball Times. The formula, which is park adjusted, is a strong predictor of future performance.  The inputs to the formula are batted ball data and other peripheral stats for the hitter in 2011.

In the summary below, I consider a difference of less than 10 points between x-BABIP and BABIP as neutral.  Higher differences are indicative of a potential correction in the direction of x-BABIP in the future.

BABIP Higher Than Expected

Bogusevic

Downs

Paredes

BABIP Lower Than Expected

Lee

Barmes

Sanchez

Altuve

Schafer

Towles

Corporan

Johnson

Quintero

Neutral BABIP

Martinez

Wallace

Bourgeois

Despite an overall team BABIP higher than average, the BABIP calculator indicates that a substantial number of the returning players have an expected BABIP higher than what they produced in 2011.  The most significant candidate for downward regression is Jimmy Paredes; while he may be capable of sustaining a relatively high BABIP, his .383 BABIP sticks out as unsustainable, with or without the assistance of the BABIP calculator.  Matt Downs was one of the success stories of 2011, but he may have been hitting over his head somewhat.  Bogusevic's .355 BABIP looks unsustainable, but Bogusevic's x-BABIP is high enough that the regression may not be as severe as it seems.

J.R. Towles appears to be the unluckiest of the hitters who fell short of their x-BABIP, with a deficit over 100 points.  I'm not sure what it is about Towles' batting luck, but every year it seems like he has more than his share of "at 'em" balls.  Perhaps Towles gets frustrated too quickly when he has bad luck, and his ML seasons take a downward spiral.  I have no idea if Towles will be with the Astros next spring; but I think there is still a chance that he will have a nice rebound season at the plate.  (Behind the plate.....that may be a different story.)    Angel Sanchez, Carlos Corporan, and Humberto Quintero also appear to have suffered a lot of bad luck at the plate.  If Carlos Lee and Chris Johnson improve on their 2011 numbers, as suggested by their x-BABIP, that could help bolster the middle of the lineup.

Batting Average on Line Drives

Another approach is to disaggregate batting averages on batted balls.  Batting averages on line drives are high--an average of.722 in 2011--because the balls are hit hard.  The percentage of hits on line drives consistently falls within the .720 - .730 range each year, and the distribution of batting average above or below that level appears to be more or less randomly distributed.  (However, an all fields line drive hitter may be able to sustain a higher batting average than a pull hitter.)  Hitters whose line drives are caught at a much higher or lower rate than the MLB average probably saw more than their share of good or bad luck. 

Because so many of the Astros' hitters had small-ish sample sizes this season, I don't put a lot of weight on this statistic right now.  Line drives are a narrow slice of all batted balls, and if the number of at bats is small, the impact of a high or low batting average on line drives can be misleading.  Many of the  hitters remaining after the trades, like the minor league call ups, played for a partial season.  Carlos Lee and Angel Sanchez, who were with the pre-trade Astros team, were unlucky on hits off their line drives, while Clint Barmes was somewhat lucky.

Despite the limitations for using this statistic, we can use it to confirm the x-BABIP conclusions, when the results point in the same direction.    The line drive batting average supports the "lucky" BABIP for Jimmy Paredes and Matt Downs, and confirms the "unlucky" BABIP for the following: Lee, Bourgeois, Sanchez, Schafer, Towles, and Johnson.

The data is shown below.