Toying with BABIP

People always cite BABIP as a means of easily telling how lucky or unlucky a hitter has been. .300 is generally accepted as the average to which most (if not all) hitters regress with their BABIP. If you're over .300, you're lucky. If you're under .300 you're unlucky. Clear as day.

Quite frankly, it's an easy way to sound smart when talking about baseball. You're acknowledging that luck can play a large part in skewing a player's stats one way or another.

But not all BABIPs are created equal. For example, this season, Dan Uggla has a BABIP of .302. Pretty average. But he's only hitting line drives at rate of 10.1%, well below the league average of around 18-19%. So his average BABIP is in fact, pretty lucky. Contrast that with a guy like Derrek Lee, who has a below average BABIP of .275 but is hitting line drives at 23%.

I've always said you can't look at BABIP and draw accurate conclusions unless you couple it with line drive percentages. The same goes for judging the luck of pitchers.

There was discussion on the 126th best Cubs blog today about what Aramis Ramirez's woeful batting lines would look like if you regressed his BABIP and batted ball percentages to the average. Harry Pavlidis, Lord of the Internets, was quick to point out that regressing hitter's power numbers to the mean may not be the best idea because unlike pitchers, that's an individual skill.

I would tend to agree. But there's no reason to think that something as simple as batting average couldn't be normalized by taking a deeper look at BABIP and LD%.

The standard I've always heard to calculate expected BABIP (xBABIP) is LD% + .120. I've never completely comprehended the thinking behind that, but I'm sure someone much smarter than me came up with it. So we'll use that.

I also came across what I think is a better system on the Twins' blog Twinkie Town this afternoon. Essentially, you multiply the batted ball type (line drive, ground ball, flyout, pop fly) by the percentage that those batted balls typically fall for hits. The calculation for this expected BABIP, which we'll call eBABIP to avoid confusion is then as follows: (LD% * 0.720 + GB% * 0.231 + FB% * .171 + PF% * 0.019)

Using both of these methods, we'll take the average and find a new BABIP (nBABIP) that might be more representative because they're based on what types of batted balls the players are hitting.

With a new BABIP, we can calculate the new number of hits that player should have, based on the number of balls they put in play (not counting strikeouts, walks, or home runs, but adding in sacrifice flys). By adding home runs and strikeouts back into the equation, we can get a new batting average (nAVG), a number that we can use to easily compare how lucky or unlucky a batter has been in terms of not getting out when they hit the ball.

I've also included dBABIP and dAVG to represent the difference between the current numbers and my adjusted numbers.



Using these numbers, you can see that Xavier Nady has been the unluckiest hitter on the team. He's batting .222 on the season but in a normalized situation he should be batting .313, 91 points higher.

It should be noted that Nady only has 72 at bats on the season, so these numbers aren't as reliable as other regulars on the team. The more at bats a player gets, the more representative his batted ball percentages will be, and the more accurate these adjusted numbers become. Take note of the aforementioned Derrek Lee, who by these calculations should be batting near .280.

The luckiest hitter so far for the Cubs has been Marlon Byrd, who's .299 average drops to .268 when we normalize his batted balls. Kosuke and Starlin Castro appear to have been lucky in the early going this season as well.

I'm not going to bother trying to normalize these player's power numbers, or try to adjust using UZR or park or anything like that. This exercise is purely looking at batting average, admittedly a flawed stat. This is something I'm going to try and update throughout the season. We'll see how these numbers change as the season goes on.