/cdn.vox-cdn.com/uploads/chorus_image/image/54873795/usa_today_10011687.0.jpg)
If you rub elbows with the more analytically-minded fans — those who frequently visit this site definitely do — you have heard a lot about how a certain player may be “lucky” or “unlucky.” One of the primary objectives of advanced metrics is to strip away all of the outside variables that influence our statistics in ordder to isolate the things that a player has direct control over. In doing this, we can more accurately measure his skill.
With the advent of these statistics, we are much better equipped to quantify the luck that is inherent in baseball. It isn’t a new concept — players have been rifling line drives directly to a defender since the game was invented, and they always knew it was just bad luck. But now we have methods to measure that luck, and to estimate the effects of luck over time.
Recently, the most common way to do this was with batting average on balls in play (BABIP), which simply measures the percentage of batted balls that fall in for a hit. Certainly a line drive is more likely to become a hit than a fly ball, but the idea is that over a large enough sample, the line drives and fly balls even out. The league-average BABIP becomes a good measure of what should happen when a ball is put in play and, combined with line drive rates, can paint a decent picture of a player’s luck.
And that system is fine. It is far from perfect, as most stats are, but it does give us a general sense of whether a player has been lucky, unlucky, or somewhere in between. However, now that fans have access to Statcast data, there are new and more accurate ways to quantify the effects of luck.
Among other things, Statcast measures the speed of a batted ball (exit velocity) and the trajectory (launch angle) as it leaves the bat. With this data, each batted ball is assigned a percentage based on how often comparable balls have become hits in the past. If a player rifles a line drive directly to a defender, we can now definitively say that similarly struck balls are a hit X% of the time, and the player was indeed unlucky. Or if a soft ground ball finds a hole, we know how lucky the hitter was.
Using the same methodology, each batted ball can be assigned an expected Run Value. Over time, those run values add up to give us a clearer picture of a player’s luck. Expected wOBA (xwOBA) does exactly that, and it tells us what a player’s weighted on-base average (wOBA) should be based on the batted ball data combined with real-world walks, strikeouts, etc. It’s the same concept as using BABIP in that we are removing defense from the equation, but now we base it on empirical data on the quality of contact, instead of the actual outcomes.
xwOBA is as prone to inaccuracy in small samples as any other stat, but with most teams having reached the quarter-season mark, we are starting to see some reliable trends. If you find the difference between each player’s xwOBA and their real-world wOBA, you can create a leaderboard of baseball’s most unlucky hitters.
xwOBA-wOBA
Player | PA | wOBA | xwOBA | Diff. |
---|---|---|---|---|
Player | PA | wOBA | xwOBA | Diff. |
Nick Castellanos | 178 | 0.307 | 0.393 | 0.086 |
Miguel Cabrera | 120 | 0.331 | 0.416 | 0.085 |
Alex Avila | 78 | 0.465 | 0.550 | 0.085 |
Todd Frazier | 132 | 0.281 | 0.353 | 0.072 |
Khris Davis | 169 | 0.340 | 0.408 | 0.068 |
Kendrys Morales | 164 | 0.322 | 0.386 | 0.064 |
James McCann | 112 | 0.303 | 0.367 | 0.064 |
Matt Joyce | 135 | 0.284 | 0.346 | 0.062 |
Stephen Piscotty | 98 | 0.343 | 0.404 | 0.061 |
Kyle Schwarber | 174 | 0.296 | 0.357 | 0.061 |
The top of the leaderboard is packed with Detroit Tigers. Nick Castellanos, Miguel Cabrera, and Alex Avila lead the league in xwOBA-wOBA, each by at least 13 points over fourth-place. Castellanos has been particularly unlucky, as his fourth-ranked 50.9 percent hard-hit rate has resulted in only a .695 OPS. According to xwOBA, Castellanos has been hitting the ball well enough to be 40 percent more productive than an average hitter, yet his results have been 10 percent below average.
Cabrera’s 85-point differential ranks second in the league, coincidentally the same position he found himself at the end of the 2016 season. Last year, Cabrera posted a 152 wRC+ and received MVP votes for the 14th consecutive, but he could have been even better. According to xwOBA he hit the ball better than anyone in baseball (again), including the American League MVP, Mike Trout.
Third-place in bad luck, and leading the league in overall xwOBA, is Avila. He currently boasts the fifth-highest wRC+ in the league, and it should be even higher. A lot higher. Avila leads the league in hard hit rate at 57.8 percent, and ranks second in line drive rate. When he isn’t crushing the ball, he has been drawing a walk in 40 percent of his remaining plate appearances, too. It’s hard to imagine a way for Avila to have done better with his limited playing time.
And of course, ranking seventh on our leaderboard is the reason why Avila’s playing time is limited. As the Tigers’ primary catcher, James McCann is hitting only .196, providing a prime example for why batting average is severely limited in measuring offensive value. McCann has only managed 19 hits in his 112 trips to the plate, but seven of them were home runs. Ranking 20th in line drive rate and 21st in hard-hit rate, he isn’t making quite the same ridiculous contact as Avila or Castellanos, but he has been hitting the ball well enough to be a top-five hitter among catchers.
Flipping the leaderboard upside-down shows that the Tigers’ lack any hitters on the “lucky” side of the ledger. Andrew Romine ranks as the Tigers’ luckiest hitter with an xwOBA only four points below his actual figure. Without running the numbers, it’s safe to say that the Tigers have been the unluckiest offense in the league thus far, yet they still rank third in the AL in runs per game. If more of their batted balls start falling for hits the way they should — not to mention the terrifying return of J.D. Martinez — opposing pitchers better look out.