Using xBABIP to Find Outliers – Players to Buy/Sell for the 2nd Half

Last season I ran a similar analysis using the now defunct site. Andrew Perpetua, the creator of xStats is now with the New York Mets, so he knew what the hell he was doing. Here are the two articles from last year covering the players overperforming based on their xBABIP and the players under-performing based on their xBABIP. Below is a table showing you each player’s current BABIP, their xBABIP, and their BABIP for the rest of the season.

2019 xBABIP - 2018 Recap.

Through Mid-June, 2018 ROS 
PlayersBABIPxBABIPBABIPRegression (Y/N)
Bryce Harper0.2160.2960.341Y
Johan Camargo0.2220.2930.352Y
Anthony Rizzo0.2270.2870.315Y
DJ LeMahieu0.3010.3460.294N
Trey Mancini0.2780.3220.289MEH
Ian Happ0.3850.2970.358Not enough
Matt Kemp0.40.320.29Y
Starling Marte0.3520.2910.3Y
Albert Almora0.3680.3120.31Y
Domingo Santana0.3680.308.500*N/A*
Scooter Gennett0.3890.3410.333Y
Nick Castellanos0.4110.3560.329Y
*Only received 35 plate appearances since June 11th

So, pretty decent results. I suppose if a player was carrying a .400 BABIP, there is really nowhere to go but down. However, in the case of Matt Kemp, Scooter Gennett, Albert Almora, and Nicholas Castellanos, they both fell well-below even their xBABIP. Likewise, we saw massive positive corrections for Bryce Harper, Anthony Rizzo, and Johan Camargo. All three were fantastic buy lows and owners who were able to buy them at a discount were rewarded in the second half of 2018.

This year, I don’t have the luxury of utilizing Luckily for me, Baseball Savant has a search tool where you can basically come up with anything you want based on the features and settings. It does take a little more leg work, but we are able to get it done. It’s important to note that expected statistics are not predictive. They are descriptive and merely show what a player’s expected numbers should be based on the quality of contact, launch angles, etc from his past performance. So knowing that we can find the player’s on the far end of each spectrum (the largest difference between BABIP and xBABIP). The probability of regression for these extreme cases is much higher than the rest of the group. That’s what I’ll be focusing on in this article.

Before we dive in, you’ll notice a bunch of Rockies on this list. The xBABIP equation does not account for Park Factors. Since Coors Field inflates BABIP as much as 20-25%, we can almost eliminate them from the regression list. If a Rockies hitter shows up on the positive regression list, that’s a completely different story. The other factor to consider is speed. xBABIP doesn’t include a speed component. So while guys like Elvis Andrus, Christian Yelich, and Tim Anderson show up the overperformers list, we need to consider that their speed could be playing a role that isn’t quantified. I won’t be expecting as much regression from those players with well-above-average speed. OK, enough rambling, here is the list of overperformers and I’ll discuss the negative regression candidates below.


2019 xBABIP Overperformers

Rhys Hoskins0.3080.242-0.066
Omar Narvaez0.3240.249-0.075
Charlie Blackmon0.3490.285-0.064
Brandon Lowe0.3890.314-0.075
Nolan Arenado0.3170.261-0.056
Eduardo Escobar0.3070.250-0.057
David Peralta0.3500.297-0.053
David Dahl0.4100.367-0.043
Miguel Cabrera0.3610.312-0.049
Trevor Story0.3610.307-0.054
Christian Vazquez0.3210.272-0.049
Gleyber Torres0.3190.266-0.053
Eric Sogard0.3190.272-0.047
Corey Seager0.3220.273-0.049
Elvis Andrus0.3490.294-0.055
Christian Yelich0.3280.295-0.033
Brian Goodwin0.3550.313-0.042
Marcus Semien0.2920.266-0.026
Austin Meadows0.3680.332-0.036
Tim Anderson0.3720.329-0.043
Jorge Polanco0.3490.320-0.029
Jeff McNeil0.3800.340-0.040
Byron Buxton0.3130.282-0.031
Adalberto Mondesi0.3520.322-0.030
Xander Bogaerts0.3280.301-0.027
Juan Soto0.3650.323-0.042
Joey Votto0.3260.296-0.030

Negative Regression 

Brandon Lowe (2B – TB)
Surprise, surprise. No, not really. Lowe has by all accounts been a pleasant surprise for fantasy owners this season. He’s hitting for average, power, and chipping in with some speed. Anyone can look at Lowe’s BABIP and expect regression but what is interesting is that xBABIP is still .314. That means his batted ball quality has been great. His barrel rate is fantastic and he hits a ton of line drives and high-quality fly balls. It’s going to be difficult to keep up that quality of contact but even if he does, the expected metrics drop his BABIP by .075. That means his average goes from .279 to around .230. He’s a clear sell candidate but try and get a top 100 player for him. He’s ranked 71st on the Razzball Player Rater so it should be possible.

David Peralta (OF – ARI)
Peralta had an unexpected breakout last season at age-30 in terms of power. As we peek at his player page, we can see that his barrel rate, average exit velocity, and hard hit% are all down this year compared to 2018. On the plus side, his batted ball profile is similar to last season and while his exit velocity on line drives and fly balls (LD/FB) is down, it’s still pretty strong at 94.3 MPH. I don’t think Peralta is a complete lost cause, but there’s just no way he can maintain his .350 BABIP given his quality of contact. I don’t think he falls to the .250 range, but something around .270 with moderate power is what I expect going forward.

Eduardo Escobar (SS/3B – ARI)
Wait, did the Diamondbacks remove the humidor this year? What is going on? Escobar also showed up as a potential negative regression candidate on my home run per barrel (HR/BRL%) article earlier this month. His over-performance was largely due to a significant portionof his home runs were not barreled, aka lucky homers. It also appears he’s due for some BABIP regression. It’s not that his actual BABIP is that high but his quality of contact is awful. His hard-hit rate is just 29.2% and has just a .316 xwOBA. I would not be surprised if he hits around .250 going forward. In addition and as previously mentioned, he’s still vastly outperforming in terms of home runs. Just regressing his barrels to league-average HR/BRL% (I know, that’s lazy but hear me out), he should have between 10 and 11 homers. Sell him immediately for anything inside the top 200. He might very well hit .250 with 10 HR from here on out. You can find that on the waivers.

Joey Votto (1B – CIN)
Votto’s decline continues. As bad as he’s been, his xBABIP thinks he should be worse. Everything is out of whack with the future Hall of Famer. His strikeout rate is up, his walk rate is down, and the power is once again diminishing. While his .319 BABIP is right in line with his previous two seasons, his quality of contact is down. Votto has never been an elite Statcast guy except for the xwOBA metric. Since Statcast’s inception in 2015, Votto has finished the season with an xwOBA in the top two percent in every season except 2019. He’s not even close. His xwOBA of ,.326 is right at league-average. That’s not what we are used to seeing with Votto. The good news for owners is that his hard contact has been up in the month of June while his strikeout and walk rates are much closer to a one-to-one ratio this month. At this point, I’m not selling, you can’t get anything for him. I’m riding it out and this given the properties of the ball, he could partially turn his season around. I think Votto can finish around .275 with 20 homers. That’s not what you drafted him for, but is useful ROS.

Rhys Hoskins (1B/OF – PHI)
Whoa, looks like we are in for some steep regression with Hoskins. When looking at his metrics, he’s essentially the same player he was a year ago. The only differences are he’s hitting the ball a little harder, walking more, AND hitting more infield fly balls. So while hitting the harder with help with his BABIP and home runs, the increased popups will hurt his BABIP. In other words, his BABIP should mirror last season’s .272 BABIP. That’s a steep drop and I’ll take the under on THE BAT’s .259 BA going forward. He’s still a great source of power and RBI and of course a hold in OBP leagues but I’d sell him in BA leagues to someone who thinks he’s a third-round value.

Quick hit: Jeff McNeil has been so impressive in his brief career thus far. He’s carrying a .370 BABIP thus far in his career over 526 plate appearances. That’s not exactly a small sample. It’s hard to see how he’s able to maintain such elevated marks without the elite quality of contact and foot speed. Don’t get me wrong, his quality of contact is good and because of his very low strikeout rate, he’s also a threat to hit .300 but I can’t envision a .380 BABIP going forward.


2019 xBABIP Underperformers

Yonder Alonso0.2000.2790.079
Justin Smoak0.2320.3040.072
Kyle Schwarber0.2680.3250.057
Robinson Cano0.2700.3230.053
Jason Kipnis0.2720.3220.050
Franmil Reyes0.2520.3000.048
Maikel Franco0.2010.2420.041
Nick Markakis0.2890.3290.040
Evan Longoria0.2660.3050.039
J.T. Realmuto0.3100.3460.036
Jose Ramirez0.2370.2730.036
Mike Trout0.3170.3510.034
Dansby Swanson0.2890.3230.034
J.D. Martinez0.3110.3430.032
Enrique Hernandez0.2350.2650.030
Lorenzo Cain0.2900.3200.030
Niko Goodrum0.3120.3420.030
Joc Pederson0.2040.2320.028
Albert Pujols0.2150.2430.028
Amed Rosario0.3110.3380.027
Marcell Ozuna0.2740.2990.025
Anthony Rendon0.3200.3450.025

Positive Regression 

Kyle Schwarber (OF – CHC)
Looking at the bottom of the list, you’ll notice a theme. It’s speed or lack thereof. So while I’m expecting quite a bit of positive regression for Schwarber, I don’t think he will manage a .325 BABIP going forward. That being said, he’s absolutely killing baseballs this year. His hard-hit rate is over 50% which ranks inside the top two percent of Major League Baseball. He doesn’t waste his balls in play as his soft contact rate is third lowest to only J.D. Martinez and Justin Turner, and right in front of Matt Carpenter. I mention Carpenter because I think Schwarber could have a stretch similar to what we saw from Carp last season. All the metrics are pointing to elite numbers but so far the surface stats are a little bit pedestrian. Schwarber’s limited by the shift but I wouldn’t be surprised to see him hit .260 with 16-18 homers the rest of the way.

Franmil Reyes (OF – SD)
Coming off a two-homer game, Franmil is not someone you will likely to be able to buy low on. Reyes reminds me of a right-handed Schwarber but without the elite walk rate. His power metrics are off the charts and the results have been there. In his 551 plate appearances to start his career, he has already racked up 38 home runs. That’s Aaron Judge/Cody Bellinger territory. OK, this isn’t about his power, it’s about BABIP. He’s hitting .247 but based on the xBABIP, he should be closer to .280. His strikeout rate near 28% will limit his batting average upside, but I’ll lean closer to .260-.265 the rest of the way with 40-homer power. That’s some good stuff right there.

Anthony Rendon (3B – WAS)
Wow, really? Rendon is already hitting .314 but based on his xBABIP, he should be closer to .340! I’ve discussed Rendon at nauseam because I think he’s an MVP candidate and doesn’t get enough love. He showed up on my HR/BRL underperformers from about a month ago and is still underperforming. This is a guy who could honestly his .350 with 20 home runs the rest of the way and I wouldn’t be surprised. There’s not more I can say, he’s great!

Jose Ramirez (2B/3B – CLE)
Ugh, I just traded Jose Ramirez, Domingo Santana, and Chris Paddack for Nolan Arenado and Kyle Gibson in a 12-team league. I thought it was more important to get an elite player for a bunch of mid-tier options in a shallower 12-team league. Part of me wanted to hold on to Ramirez to see if he could turn it around. He’s been better of late but even if he improves to his actual talent this year, he’s still a .250 hitter. I am a believer in that his power will increase with the weather heating up. If he hits .250 with 10 homers and 12-15 SB, owners should take it. I would have as well but with Arenado dangling, I couldn’t resist. 

I won’t go into too much detail with the top two names: Yonder Alonso and Justin Smoak. Both have been very disappointing and typically underperform based on their xBABIP but not to this extent. Neither player is fleet of foot so I wouldn’t expect full positive regression from either. Still, both players have good power and if healthy could hit around 15 home runs in the second half. If their BABIPs come up 30 or 40 points, both are useful in 12-team leagues and a CI or utility spot. In 15-team leagues, I’d look to acquire them (Smoak over Alonso) as a throw-in to a bigger deal.

Mike Trout, LOL

Dansby Swanson (SS – ATL)
We are in the midst of Swanson’s breakout. If you missed out, that’s OK because I don’t think he’s being fully appreciated. Maybe it’s prospect fatigue and the fact that he didn’t bust out in his first couple seasons. I don’t know, either way, I think there’s more upside here. He’s still just 25 years old, already has a career-high in home runs and has more barrels through the first half of this season that he has in his last two seasons combined! In addition to huge gains in hard contact, he’s swinging at pitches outside the zone less often and smoking line drives and fly balls. There’s no reason his batting average should be in the .250s. I think he will comfortably sit around .275-.280 going forward with good power numbers and a prime spot in the Braves offense. Don’t sleep on his speed either, 10-12 SBs plays in today’s fantasy game.

Amed Rosario (SS – NYM)
This one snuck up on me a little bit. After a decent audition in the second half of 2017, he was pegged as a potential breakout in 2018. He didn’t quite live up to the hype but was serviceable, especially for a 22-year-old. He’s already matched his home run total from 2018 with nine but his batting average is right in line with last year. He’s improved his exit velocity by 2.5 MPH on average and is a few more line drives and fly balls while hitting fewer popups. That will boost ones BABIP for sure yet his current BABIP matches what he did last season. He makes enough contact and has great speed so I’d expect something closer to his xBABIP for the second half. He also has an outside shot a going 20-20 which is rare in today’s game.

Follow me @FreezeStats. Check out my work at FantasyPros and Pitcher List.

Photo Courtesy of the Sun Times


Outfielder Analysis Using Statcast – Get with This or Get with That?

Let’s kick it back to the old school. It’s been a while since I went back to my 90s hip hop roots and tied in fantasy baseball. If you don’t get the reference, sit back, relax and kick it with Black Sheep. I’m going to compare a set of hitters who, on draft day, were over 100 spots apart. I will show two tables. The first will go over each players’ standard fantasy categories and the second will show the Statcast data. I’ll follow up with a discussion on how I feel about both players going forward in 2019. Let’s keep in mind, I do like both players and am not devaluing either of them going forward. I’m showing how impressive these slow starters actually are.

Rhys Hoskins or Franmil Reyes

Rhys Hoskins0.273209250
Franmil Reyes0.23598140

OK, so both had eight home runs (Hoskins hit his ninth last night) but Hoskins has the clear advantage in runs and RBI as well as batting average. Is this really all that close? Hoskins actually has more than 30 plate appearances than Reyes which explains the advantage in R + RBI. The fact that both are tied in home runs tells me that Reyes has some massive power. But, he doesn’t play every day because the Padres have five outfielders. Hoskins, on the other hand, plays every day and hits cleanup for the Phillies. That’s a huge advantage for Hoskins. What if I told you that from this point forward, these two would provide similar value? Let’s check the Statcast numbers.

Rhys Hoskins23.895.15.50%377 Ft0.344
Franmil Reyes15.997.314.30%402 Ft0.441

Depending on how you interpret these Statcast numbers, Reyes gets the clear advantage in all metrics. Some might argue that Hoskins gets the advantage in terms of launch angle but a portion can be attributed to an elevated 12.7% popup rate. Meanwhile, Reyes has done a great job of avoiding the near-automatic out by popping up just 3.1% of the time. In fact, Reyes’ fly ball rate per BaseballSavant is actually two percent higher than Hoskins (34.4% for Reyes and 32.4% for Hoskins). Regarding approach and plate discipline, Hoskins is by far the better option. In OBP leagues, Hoskins is the clear winner. However, if Reyes and Hoskins were to receive the same number of plate appearances going forward, there’s no doubt Reyes would have more home runs. Unfortunately, there’s the playing time issue is San Diego. I am hoping this plays itself out in the upcoming months and the Padres trade at least one of their current outfielders. Either way, here my thoughts on Reyes.

Franmil is one of the most dangerous hitters in the game. He’s in the top 6% in exit velocity, xwOBA, and xSLG. He’s in the top 10% in hard hit% and more surprisingly, in the top 10% in xBA. Given his current .233 average, one has to wonder what’s going on? It’s the BABIP which sits at a lowly .200. He has increased his FB% by a whopping 17% (nearly 11 degrees on his average launch angle since 2018). It’s a valuable increase in elevating the ball as he hasn’t hit an infield fly ball (per FanGraphs) and the aforementioned 3.1% PU rate.

This approach should provide Franmil with plenty of home runs. I’d put his ceiling at 45 HR given 600 PA this year. No, that’s not a joke but given the depth in the Padres outfield and his poor defense, he’ll probably reach somewhere around 475-500. That should still yield close to 35 homers for big Fran. He’s also changed his approach. He’s swinging at everything. His swing rate is up 10% and he’s swinging outside the zone 38.8% of the time up from 31.7%. It’s not all bad because his zone-swing percentage is up 11% To 81%! His overall contact rates remain relatively steady. The aggressive approach has helped cut his K% because he isn’t getting himself into as many deep counts. That’s great for his overall production but pitchers will adjust. Unfortunately, luck has dug him into a batting average hole but I think he can hit .250-.260 with huge power upside. If he finds himself in an everyday role, he could be a top 75 player this year.

David Peralta or Jesse Winker

Daivd Peralta0.315194200
Jesse Winker0.234178130

I know what you’re thinking, why are you choosing these boring hitters without any speed? In fantasy baseball, the players with power + speed tools are always scooped up while the Peralta-types get left on waivers and are devalued. I’m of course talking about Peralta prior to last season’s breakout. Thus far in 2019, Winker has shown more power but Peralta gets the clear edge elsewhere. Both players are hitting in top third of their respective lineups but the Diamondbacks have gotten off to a better start. In both the preseason and likely right now, most fantasy owners would prefer Peralta to Winker. Let’s take a look at the Statcast metrics.

David Peralta8.194.83.70%407 Ft0.297
Jesse Winker10.3945.60%384 Ft0.381

Honestly, they aren’t all that different until you get to the last row. Winker’s expected wOBA is nearly .090 points higher! Walks are factors into wOBA, so Winker gets the edge there but he’s only walking 4.6% more often than Peralta, so there’s something else at play here. Oh, there it is. Peralta’s BABIP is currently an unsustainable .391. He’s also been a little more aggressive this year but mostly on pitches outside the zone. That’s caused his strikeout and walk rates to go in the opposite directions. Peralta is still a very good contact hitter and has shown that he can muscle up with some power. I just think Winker is the guy that should be hitting .315 while Peralta should be closer to .275. Going forward, Peralta is a bit of a sell and of course you know Winker is a buy.

Here are my thoughts in Winker from the preseason. Here are my thoughts now.

Winker is trading some patience for power this year. He’s seen his walk rate dip by four percent while his strikeout rate has increased by about the same amount. He’s only increased his swing rate by four percent and it’s all going to pitches inside the zone (which is great). His swinging strike rate and contact rates are nearly identical to 2018. I’d expect Winker’s walk rate to rise a little based on this approach. He’s also stinging the ball with hard contact rate of near 50% and his line drive rate is a healthy 26%. Even with his poor foot speed, you’d expect an elevated BABIP. Instead, his BABIP sits at a measly .200.

Winker is being shifted on more in 2019, up to 40% of the time this year compared to just 19% in 2018. His batting average and wOBA have taken a hit thanks to the shift. What doesn’t make sense is his 0.063 BABIP on fly balls and an extremely low .368 BABIP on liners. For reference, his career BABIP on line drives is .681 which matches closely with league-wide average. If we only regress his BABIP on line drives back to his career rate, he’d be hitting .293, LOL. That doesn’t factor in the unlucky BABIP on fly balls. Look, Winker is about to go nuts, he’s already proven that the power is real, although he won’t maintain anything near a 33% HR/FB going forward. Winker likely ends the year near .300 with 25+ homers with a possibility of 30 long balls. Isn’t that what you were hoping for from Corey Seager this year?

Follow me @FreezeStats. Check out my work at FantasyPros and Pitcher List.

(AP Photo)