clock menu more-arrow no yes

Filed under:

Exploring advanced stats - xwOBA

Colorado Rockies news and links for Monday, February 7, 2022

Baseball is a fickle game. Inches, and even fractions of inches matter a lot. An inch is the difference between a ball or a strike. A fraction of an inch is the difference between a line drive and a pop up.

To twist things even more, that line drive and pop up can go completely different places from there. Liner into the shift? Easy out. Pop up into no-man’s land in shallow left? Bloop single. Being off by a fraction of an inch can easily result in a better outcome, at times. An 0-4 game at the plate can consist of four hard hit balls to unlucky places, while a 4-4 game can be filled with terrible contact. But what matters at the end of the day? The hits - creating actual offense, not just loud outs.

Fortunately, baseball being a game of large sample sizes, these things tend to even themselves out relatively quickly. Baseball is also a game of data, and finding hidden value in players is a trademark of successful teams. So predicting what contact should and shouldn’t be hits, compared to what actually becomes a hit, is a valuable and useful science. Enter xwOBA, Expected Weighted On Base Average.

Weighted On Base Average (wOBA) is a measure of a hitter’s offensive contributions by weighting at bat outcomes (singles, doubles, homers, walks, etc) by their run values. xWOBA, then, is a Statcast based stat that ignores the actual outcome of the at bat, but rather predicts outcomes using the Statcast measured launch angle and exit velocity, plus the batters sprint speed. So, when a line drive down the third base line is saved by a web gem from the third baseman, wOBA counts it as an out, while xwOBA gives it an expected outcome based on similarly hit balls.

There’s an argument here for defensive positioning being based on player tendencies, but one of the goals of expected hitting stats, and analytics in general, is to decouple a hitters performance from the defense or pitcher they are facing. It’s not the batter’s fault that the third baseman who caught the line drive is an all-world talent, so the hitter should not be punished. On the flip side, the hitter shouldn’t be rewarded for dropping a ball in front of an outfielder who is pulled back to cover an expansive field.

Over the course of the season, wOBA and xwOBA converge, with slight discrepancies here and there. The Rockies mostly outperformed their xwOBA last year.

Rockies 2021 xwOBA

Player wOBA xwOBA Diff
Player wOBA xwOBA Diff
Cron, C.J. 0.383 0.362 0.021
Blackmon, Charlie 0.333 0.357 -0.024
Story, Trevor 0.341 0.333 0.008
McMahon, Ryan 0.334 0.324 0.01
Díaz, Elias 0.33 0.317 0.013
Rodgers, Brendan 0.341 0.309 0.032
Daza, Yonathan 0.304 0.292 0.012
Hampson, Garrett 0.288 0.29 -0.002
Tapia, Raimel 0.305 0.279 0.026
Nuñez, Dom 0.294 0.262 0.032
Fuentes, Joshua 0.261 0.23 0.031

Interestingly, the Rockies outperformed most of their expected stats (xWOBA, xSLG, xBA) in 2020. While that doesn’t inspire much hope for the team’s run totals going forward, it could very likely be due to one of the many intricacies of Coors Field. The league leaders in xwOBA are the usual suspects - Juan Soto and Bryce Harper top the list, but strangely just four of the top 25 were underperforming by wOBA, the rest had higher expected than actual rates.

What should you take from this? Whatever you want. Expected statistics are meant to illuminate and illustrate, but definitely not be the be all and end all. Again, baseball is a fickle game. If you’re tied in an NL Wild Card game and your typically light hitting catcher hits a grounder up the middle, you expect an out, and xwOBA would likely agree with you. But, as we all know, anything can happen.

★ ★ ★

Rockies revamp front office with eye toward analytics | The Denver Post ($)

It’s been no secret that the Rockies analytics department has seen some ups and downs (well, mostly downs) over the past few years. Fortunately, it looks like the new administration has been able to fill some of the numerous holes in the department in an attempt to get everyone moving in the same direction. The Post has a good recap of the Rockies recent front office moves, detailing each hire and promotion, as well as his or her background. MLB analytics departments are definitely not built in a day, but they have to start somewhere.

The lockout projected ZiPS Standings: National League Edition | FanGraphs

On the subject of expected outcomes, in this case wins and losses, FanGraphs Dan Szymborski has released his latest round of projections for the 2022 season on a team level. They are, predictably, not kind to the Rockies - sitting at 96 losses in the NL West basement, with a 0.1% chance of the playoffs. You’d be delusional if you thought the Rockies chances were even decent, but seeing the percentages still hurts.

★ ★ ★

Please keep in mind our Purple Row Community Guidelines when you’re commenting. Thanks!