The Misapplication of BABIP

[rudyjuly2]

I wrote an article for FullCountPitch.com. It's about batting average on balls put in play (BABIP) and why I believe it's crap. Way too many guys are using it and automatically assuming this is a good way to determine the luck of a player. Babip implies that once a ball is hit in fair territory the odds of a player getting a hit is more luck than skill based. Rather than credit a player for a good season I'm reading quite a few comments stating this player was lucky and is due for a dip in performance (Austin Jackson). Or a player's bad year was due to bad luck (Matt Kemp). I believe luck is only a very small part in the equation and that BABIP is very flawed. Read the article for a more comprehensive stat breakdown.

http://fullcountpitch.com/2011/02/08/the-misapplication-of-babip/

[snepp]

Re: The Misapplication of BABIP

I didn't see anything there that shows there's something wrong with BABIP, only it's potential misuse. How is that any different than any other stat that requires appropriate context?

[rudyjuly2]

Re: The Misapplication of BABIP

I didn't see anything there that shows there's something wrong with BABIP, only it's potential misuse. How is that any different than any other stat that requires appropriate context?

I'll disagree in that I think it's a complete waste of time. The entire concept of babip is to quantify luck in baseball. Some people made the assumption that when you put the ball in play the odds of getting a hit are more luck based than skill. For some reason that assumption has stuck without proving anything. But when you look at the long term numbers (which give the same results as the short term numbers) BABIP is highly correlated to avg at 76%. Long term numbers don't lie. If a player's skill didn't matter that much on balls put in play everyone's babip should be close to .305 without a wide spread but we don't see this at all. This indicates luck is not significant which renders babip pointless imo.

Let me ask you this - do you think babip is a significant indicator of luck? When you see a guy with a babip of .330 will you assume he's lucky? If not what information can you determine when looking at it?

[Sportsforever]

Re: The Misapplication of BABIP

I wrote an article for FullCountPitch.com. It's about batting average on balls put in play (BABIP) and why I believe it's crap. Way too many guys are using it and automatically assuming this is a good way to determine the luck of a player. Babip implies that once a ball is hit in fair territory the odds of a player getting a hit is more luck than skill based. Rather than credit a player for a good season I'm reading quite a few comments stating this player was lucky and is due for a dip in performance (Austin Jackson). Or a player's bad year was due to bad luck (Matt Kemp). I believe luck is only a very small part in the equation and that BABIP is very flawed. Read the article for a more comprehensive stat breakdown.

http://fullcountpitch.com/2011/02/08...tion-of-babip/

I read the article. I appreciate you taking the time to look at this, but hoenstly I couldn't disagree more.

First off, most folks acknowledge that BABIP is a little different for hitters than it is for pitchers. For example, if a hitter hits more line drives (a better hitter will), he's going to have a higher BABIP. The problem with Jackson last year was his LINE DRIVE % was insane early in the year when he was hitting out of his mind. Everyone predicted that he would not keep that up and he didn't. When he stopped hitting line drives things started falling back to where they should be.

Secondly, BABIP and luck are more associated with pitchers. This is something that HAS been looked at in depth. You just can't show a correlation between pitching skill and BABIP. Pedro Martinez in 1999 had one of the highest BABIP's in baseball and then the next year had one of the lowest. You see Greg Maddux and Randy Johnson have similar fluctuations.

Finally, I don't like when BABIP is used to discredit a player. For example, in Jackon's case you might hear someone say, "Oh, well his season wasn't very good because he actually got lucky and had a high BABIP." No, he hit whatever he hit and that is what goes on his baseball card. Where I see the value of BABIP is evaluating players going FORWARD. I completely believe that if Jackson does not cut back on strikeouts and increase his walks (basically stop swinging at everything) his batting average will plummet.

Anyway, thanks for writing the article. I always enjoy talking about this stuff.

[rudyjuly2]

Re: The Misapplication of BABIPhttp://www.operationsports.com/forums/newreply.php?do=

I read the article. I appreciate you taking the time to look at this, but hoenstly I couldn't disagree more.

Thanks for your brutal honesty lol. But I actually think we agree on a lot of things.

First off, most folks acknowledge that BABIP is a little different for hitters than it is for pitchers. For example, if a hitter hits more line drives (a better hitter will), he's going to have a higher BABIP. The problem with Jackson last year was his LINE DRIVE % was insane early in the year when he was hitting out of his mind. Everyone predicted that he would not keep that up and he didn't. When he stopped hitting line drives things started falling back to where they should be.

I completely agree with this. Higher line drive hitters will push a BABIP up. But that's not luck imo. That's a reflection of the type of player you are. And if in 2011 Austin Jackson starts popping everything up and his babip goes down I will say it's because he changed his swing and not his luck.

Secondly, BABIP and luck are more associated with pitchers. This is something that HAS been looked at in depth. You just can't show a correlation between pitching skill and BABIP. Pedro Martinez in 1999 had one of the highest BABIP's in baseball and then the next year had one of the lowest. You see Greg Maddux and Randy Johnson have similar fluctuations.

I haven't taken a good look at pitcher's and babip. But I believe there are tangible reasons why a batter's babip goes up and down so I can only assume similar things go on with pitchers. Strike out a ton of guys and your babip goes up. If you give up a lot of singles it goes up. Give up a lot of extra base hits and it goes down. I would assume these things affect pitchers too but I don't think fangraphs lists the complete break down of type extra base hits vs. singles. Not sure if I can test the same things.

Finally, I don't like when BABIP is used to discredit a player. For example, in Jackon's case you might hear someone say, "Oh, well his season wasn't very good because he actually got lucky and had a high BABIP." No, he hit whatever he hit and that is what goes on his baseball card. Where I see the value of BABIP is evaluating players going FORWARD. I completely believe that if Jackson does not cut back on strikeouts and increase his walks (basically stop swinging at everything) his batting average will plummet.

Anyway, thanks for writing the article. I always enjoy talking about this stuff.

This was my biggest pet peeve about babip and why I started looking into this in the first place. I don't normally go into super depth studying baseball stats but I do love stats (I am a statistician). I just got tired of constantly reading excuses about why player X had a bad year. It seems nobody wants to admit players have bad years anymore. One look at a bad player and let's just look at a low babip and blame it on that! I wish everytime I didn't do something well I could just blame it on bad luck.

I also read a couple really questionable things about the analysis of babip on one site where they were misinterpreting regression results. They made it very clear on their website that something was great when in fact it was awful. I just think some people have made assumptions about this without really looking at the results. Thanks for reading and providing feedback.

[dodgerblue]

Re: The Misapplication of BABIP

I enjoyed your article. I can see how BABIP really only diverges from AVE when we are dealing with high strike out players. You mention in your article that singles hitters also see a larger difference. I am interested as to why that is so? Is that just because singles hitters are more "contact" hitters and there is just that many more chances or opportunities for the ball to drop in for a hit?

[rudyjuly2]

Re: The Misapplication of BABIP

I enjoyed your article. I can see how BABIP really only diverges from AVE when we are dealing with high strike out players. You mention in your article that singles hitters also see a larger difference. I am interested as to why that is so? Is that just because singles hitters are more "contact" hitters and there is just that many more chances or opportunities for the ball to drop in for a hit?

Yup. Put the ball in play more and you are more likely to get a hit. Line drive and ground ball hitters will always have higher averages and babips because of it. Fly balls tend to cause outs and have a negative correlation with both avg and babip. It's why the coach in Major League made Wesley Snipes do pushups every time he flied out. When a player hits a lot of singles it strongly indicates they don't hit many HRs either. HRs get removed from the babip formula since that implies no luck/randomness was involved since a defender could never get their hands on the ball. Speed isn't a factor in the formula either and singles hitters tend to be your faster players as well. Speed sometimes gets you infield hits which boost your babip.

Proponents of babip believe that singles are largely luck based. Your talent gets the bat on the ball and then it's just random as to whether you get a hit afterwards. I don't believe that is true. Looking at my career database with guys all over 5000 ABs I separated them into three groups. Singles hitters, normal guys and below avg singles hitters. I only had 59 guys who I could call above average singles hitters and 78% of their hits were singles. They had average of .279 and babips of .298. The normal group of 67% and the low group which had 57%. Again, if getting a hit was just random on singles, these singles hitters should all have similar babips regardless of average. But it's not true.

Ozzie Smith was a singles guy and batted .262 and only had a .275 babip. Tony Gwynn batted .338 and had a babip of .341. Do people really believe Gwynn wasn't skilled at getting hits with singles? You can't be lucky for over 9000 abs. Look at how strongly correlated avg is with babip in this group.



Sometimes I think we just need to use common sense in this area. I do believe we are trying to read too much into this stuff if we are really trying to predict luck. If a batter goes 3-5 with a 1B, 2B and HR along with 2Ks, his babip for the day is 1.000. Do you call that a good day or a lucky day? If a batter goes 0-5 with 3Ks his babip is 0.000. Do you call that a bad day or an unlucky day? And if those same players do something similar over 5 or 10 games are they hot/cold streaks or lucky/unlucky streaks? Most people would say it's ludicrous to call them lucky or unlucky based on a few games. And you can't infer luck for a player over the course of his career because these things even out. So why give so much weight to luck if it happens over 50, 100 or 150 games? In a lot of ways babip is lazy statistic that just creates excuses for things rather than imply anything meaningful.