Last week, I put up a poll on Twitter asking for people to vote on what stat they wanted me to cover next, and BAbip, or Batting Average on balls in play, won handedly, so that is the topic today/this week.
(Ignore the 12 votes thing. This never happens I swear.)
There was a dissenter however: my nemesis – and coworker at both The Fordham Ram and WFUV – Pat Costello.
Obviously, I don’t think it’s dumb. But, there is a significant subset of people (including Pat), who think all of the advanced statistics “invading” baseball are dumb. They argue that FIP is unnecessary because ERA has worked for so long, that BAbip is just overcomplicating batting average, and that WAR is trying to abstractly quantify something that doesn’t require quantifying. Basically, it all circles back to the idea that pervades – and honestly, harms – baseball: that things should remain the way they are because that’s how they’ve always been and baseball has lasted this far without advanced statistics.
That’s basically the point of this post, the FIP post, and any other impending stat explanation posts (and there will be more): to try and convince you all that these advanced statistics are grounded in things that happen on the field, and provide us with a better feel for the game. And that if I can (sort of) handle the math involved, then there is at least a solid backing to them, despite how complex they may seem.
Anyway, onto an actual explanation of BAbip. As mentioned above, it is batting average purely on balls in play. This means it eliminates strikeouts and walks, but does include sacrifice flies, along with the other usual hits. The formula is pretty straight forward:
(Hits – Home Runs)/(At-bats – Strikeouts – Home Runs + Sac Flies)
The main thing that sticks out from the formula is that home runs are not included as a “ball in play”. This is because a homer is obviously a hit, regardless of where beyond the fence it lands, making it not truly “in play”. Nothing about the defense can keep a home run from being a home run. And be quiet smart ass who’s saying “But they can rob it!” Then it’s not a home run, is it?
BAbip, despite being simple, can show a lot of different things. The main use is to figure out how “lucky” or “unlucky” a player is being at the plate, or a pitcher with the hits he is giving up. The league average BAbip sits just around .300 every year. So by just glancing at a player’s BAbip and seeing if he is well above or below .300, you can expect him to either over the course of the rest of the season or the next season to regress more towards the mean.
Take, for example, Nick Markakis of the Braves. Let’s compare his 2015 season BABip and batting average to his numbers so far in 2016.
His BAbip in 2015 was well above .300, resulting in a .296 batting average that was his highest since 2011. So far in 2016, his BAbip has regressed to right around the league average, explaining in part his plummeting average (though that is also due to the highest strikeout percentage of his career so far in 2016). In other words, a drop in the average should have been expected for Markakis.
However, there is an important caveat for BAbip: a single season should also be compared to a player’s career BAbip. This is generally a better tool for determining potential regression or not. You most often see an inflated career BAbip on speedy players. Lorenzo Cain of the Kansas City Royals is a great example of this.
LoCain is fourth in the majors in BAbip over the last three seasons – really his time as a star player for the Royals – with a .360. His .380 in 2014 was very high, and the odds for regression figured to be very high, as he generally hovered around league average as well. However, as the Royals improved at small ball so did he. In other words, his batting average rings true despite his high BAbip. He’s sustained a high average on balls in play over the last three years thanks to plus speed, plus an ability to spray the ball over the ballpark. Cain is interesting in the sense that if it weren’t for his high strikeout totals (108 in 2014, 98 in 2015, 48 so far this year) his batting average would more accurately reflect his BAbip, but by being higher instead of lower.
Comparing single-season BAbip to career BAbip is the most effective use. It makes it possible to weed out whether a player is getting “lucky” with his hit location, or if he really is fast enough/good enough at hitting the gaps for power.
In the same way we use BAbip to quantify the luck of batters, it can be used for pitchers as well. A good example of this is San Diego Padres pitcher Drew Pomeranz, who is currently having the best season of his career.
As with batters, the league average BAbip generally sits right around .300, so Pomeranz has for the most part come in under that mark. However, this .223 BAbip is second in the majors right now behind only Marco Estrada, who has a FIP of 3.82 and should be feeling some regression soon (though this is his second straight year with a stellar BAbip; he had the 7th best since integration in 2015 with a .217).
What is interesting with Pomeranz is that while that low figure is definitely unsustainable, he’s having the best season of his career by far. His strikeouts are way up (though he already has as many walks this season as he did in all of 2015), and has a sub-3.00 FIP currently. However, this is also just his second year of full-time play in the bigs, and he’s doing it in Petco Park, a notoriously pitcher-friendly park.
The Pomeranz example illustrates an important facet of not just BAbip, but baseball statistics in general: everything must be taken in context. (This also applies to such scenarios like “Is every Trump/Hillary/Bernie/blow-up-the-government supporter the same as this one example I saw?” and “Is a hot dog a sandwich?”)
Unless Pomeranz is suddenly the next Clayton Kershaw, his BAbip is most likely going to falter from his incredible .223, which would be one of the best of the post-integration era. That being said, his numbers across the board have improved in his first season in San Diego, and second season as a full time starter. His strikeout percentage is up to almost 30%, he has that 2.99 FIP I mentioned before, and his stuff also passes the still-important eye test.
Basically, he’s getting lucky, but even when his luck evens out, it looks like he should still shape up to be at least an above-average starting pitcher, if not better.
Pat, any more doubts?
@jfmclooney not useful at all. Who cares if they’re getting lucky. That’s a big part of the game.
— Pat Costello (@PatCostello20) June 2, 2016
Okay, good point. Luck is a major part of the game. But why does that mean we can’t quantify it? Just like FIP, BAbip helps illustrate whether or not a player’s performance is due to his own skill instead of luck. Over time, luck evens out, while skill is a relative constant. Players improve (or falter), which is why contextual cues for the numbers we have are important.
The common refrain is that “stats are killing the game”, but that’s just the dying cry of the ever-shrinking anti-change faction. But that just makes no sense. Stats make us smarter viewers of the game, and also makes GMs and managers better at valuing the players in their lineups and rotations. I would argue that only improves the game. Suck it Pat.
Any comments or questions about this post, or anything else, can be most quickly answered/rebutted through my Twitter account, which is littered all over this page. If you have any other ideas for topics for me to cover, whether they be stats explanations or something else, let me know, because I am a man of the people. Unless your ideas suck. Keep those away.