January 23, 2018

Statcast analysis of the xwOBA statistic as a predictive tool for fantasy baseball value. Matt Wilkes finds potential risers and fallers for 2018 drafts.

With spring training on the horizon, fantasy baseball draft season is in sight, as well. That means it’s time to pore over rankings, go through countless mock drafts, and do plenty of research trying to find that diamond in the rough other owners are undervaluing.

Luckily, more statistics are available than ever before to determine which players to target and which to avoid. Statcast has opened up fans and teams to a plethora of new and useful data, bringing statistics like exit velocity and sprint speed to the forefront of baseball conversations and analytics.

One such stat is xwOBA, otherwise known as expected weighted on-base average. It's another tool that fantasy baseball managers have at their disposal to guide them before and during 2018 drafts. Let's take a closer look at the statistic itself and how you can use it to find another advantage.

• BvP matchups data (Batter vs. Pitcher)
• PvB matchups data (Pitcher vs. Batter)
• Who should I start? Fantasy baseball comparisons
• Who should I pickup? Fantasy baseball comparisons
• Fantasy baseball trade analyzer
• MLB Props finder tool
• Fantasy baseball closer depth charts

Defining xwOBA

First, what is wOBA? Quite simply, it is used to establish the value a player brings per plate appearance, accounting for unintentional walks, hit by pitches, and all base hits. Because some events (e.g., home runs) are more valuable than others (e.g., walks), it uses a weighted scale to determine the given player’s output. It’s scaled like on-base percentage, so a .400 wOBA is elite while a .275 mark is dreadful.

(You can read more about the math behind it here and search the leaderboards on Baseball Savant.)

Expected wOBA is calculated using exit velocity and launch angle of each ball put in play, and it attempts to predict what a player’s production should have been based on these factors, independent of defense. If Kris Bryant crushes a line drive into the gap, but Billy Hamilton runs it down and makes a diving catch, regular ol’ wOBA counts that as a goose egg. On the other hand, xwOBA will calculate what should have been the likely result based on Bryant’s hit probability. This can give a better indication of Bryant’s true skill.

In short, xwOBA can be used to determine whether a player was unlucky because he consistently hit the ball hard right at a defender or lucky because he routinely snuck weak grounders or bloopers past the infield.

If you doubt the statistic’s viability, here are the top 10 players in xwOBA last year (min. 400 at-bats):

No huge surprises here. These are some of the most feared hitters in the game, and xwOBA backs up that notion.

But you already knew that. So how can this metric help fantasy owners? It’s also useful for predicting a player’s future production. By subtracting actual wOBA from expected wOBA, it’s easy to see who did and did not have luck on their side and set your expectations accordingly.

That being said, it’s important to note that while xwOBA is helpful, it — like any other stat — should not be used alone. For example, elite players often outperform their expectations (e.g., Mike Trout). Others are held back (Albert Pujols) or helped (Dee Gordon) by their baserunning ability, which allows them (or doesn’t) to beat out softly hit groundballs and turn singles into doubles. And it's always smart to look at a player's age and past performance to get appropriate context.

With all that aside, let’s look at the top-10 underperformers and overperfomers per xwOBA in 2017 and what that means heading into 2018.

On-Base Underperformers

Does this mean you can expect these players to bounce back this year? Not necessarily. None of these players are exactly burners on the basepaths, so that holds them back a bit. For instance, Pujols had the slowest recorded sprint speed by Statcast, while Kendrys Morales was eighth-slowest. They’re not beating out infield singles and likely aren’t stretching any singles into extra-base hits. Additionally, the xwOBA of Pujols and Alex Gordon were still poor. So while they weren’t as bad as the numbers indicated last year, neither was rosterable for fantasy purposes even if they performed up to expectations.

The other players on the list present more interesting cases. If you want proof of xwOBA’s predictive ability, exhibit A is Manny Machado. Big things were expected from the Baltimore third baseman heading into his fifth full big league season in 2017, but he struggled mightily for the first half of the year, hitting a disappointing .230/.296/.445. His wOBA was a disappointing .319, but his xwOBA (.355 — just a bit below his actual outputs in 2015 and ‘16) was far higher because he was still crushing the ball.

Machado’s average exit velocity was 92.1 mph, 12th in all of baseball. He was due for a rebound, right? After all, his .239 batting average on balls in play was ridiculously below his career mark, which sat at .310 coming into 2017. As expected, he became the Machado everyone knew in the second half, hitting .290/.326/.500 with a .360 xwOBA and .352 wOBA. In other words, don’t expect Machado to hit .259/.310/.471 again or perform like anyone other than a top-tier third baseman.

Another intriguing case is future Hall of Famer Miguel Cabrera, who sits atop the table. While the statistics didn’t reflect it, Cabrera continued to mash the ball last season. His 91.1 mph average exit velocity was 12th among all hitters with 100 or more at-bats and his line-drive percentage was a career-high 27.3 percent. That led to the second-largest discrepancy in baseball between wOBA (.322) and xwOBA (.382). Injuries and his lack of speed played at least a partial role in that number, but luck certainly wasn’t in his favor, either. If he continues to hit the ball with that kind of authority, he isn’t going to hit .249/.329/.399 again.

A similar case can be made for many players on this list. Despite his age, Robinson Cano is still hitting the ball hard and making frequent contact, so it’s not time to expect a decline yet. Nicholas Castellanos had a breakout season and potentially still underachieved. Shin-Soo Choo’s on-base skills and power still make him a viable fantasy outfielder. Mitch Moreland may be a sneaky late-round draft pick at a deep position if his luck turns around.

On-Base Overperformers

Now that’s a fascinating group of names. There’s an MVP winner (Altuve), an MVP second runner-up (Ramirez), a fifth-place finisher (Blackmon), and a bevy of other notable players. Let’s start with the MVP candidates.

First, Jose Altuve will be fine. Only four players in baseball had an xwOBA higher than his actual wOBA, so it’s not a number many can put up. Plus, he’s not big in stature, so expecting him to hit the ball as hard as, say, Aaron Judge is unreasonable. Moreover, his xwOBA was still respectable at .349 and he also has speed on his side to help him leg out softly-hit balls. He likely won’t put up numbers quite as good as he did in 2017, but he’s a safe bet to be a top fantasy second baseman again.

Charlie Blackmon and Jose Ramirez are in similar situations. They hit above expectations last year, but their projected wOBA was still strong and they bring speed to the table. Expect some regression, but they should be fine. Dee Gordon will also be OK because he can beat out so many of his soft hits. Ender Inciarte and Eduardo Nunez also swiped over 20 bags last season and had above-average sprint speed, so they’re more likely to overcome a poor xwOBA, as well.

It’s unlikely everyone on this list will escape regression, though. Scooter Gennett might be the player to worry about the most. He brings no speed to the table, he doesn’t walk much, he strikes out at an above-average rate, and his power surge (27 home runs) came out of nowhere. While a change in swing mechanics (i.e., launch angle) or a juiced ball may have helped, he’s unlikely to sustain a 20.8 percent home run-to-fly ball ratio or a .374 wOBA with an 86.0 mph average exit velocity.

Didi Gregorius has similar concerns due to poor exit velocity and walk rate, so a dip in his triple slash numbers might occur. He does have above-average speed (28.3 mph vs. 27.0 mph league average) to beat out some hits, however, as well as a short porch in Yankee Stadium to put up 15 or more home runs. Late bloomer Marwin Gonzalez made real strides with his plate discipline, nearly doubling his walk rate, but it's unlikely will he hit for the same average or power. Zack Cozart’s projected wOBA was still solid thanks to his suddenly high walk rate and increased launch angle. But he probably won't produce the same way in a bigger ballpark with no speed to speak of and an average exit velocity of 85.8 mph.

What to Make of xwOBA

Getting an edge on fantasy opponents involves more than looking at the categories your league uses. xwOBA is one of the many useful metrics available to determine what to expect from players in the coming year. It should be taken with a grain of salt in some cases, but when taken in full context, it provides excellent insight into player performance that you can utilize to your advantage.

More Sabermetrics and Advanced Stats Analysis