MLB Free Agency Through Analytics

Every year, as soon as the MLB season concludes and a World Series champion is crowned, teams begin looking ahead to the next year. The winter months are labeled the offseason, but most players don’t really take time off from their profession. Players do everything from training in special high tech facilities to playing in professional winter leagues with the intention of improving their skill set for the next season. Front offices spend their offseasons trying to fulfill a similar goal; making their team better. For years, free agency has been instrumental in the construction of championship teams, and with all the talent available on the market this offseason, the 2020 season will definitely not be an exception. With the rise of analytics, teams are getting better and better at making free agent decisions. So what better way to examine this year’s free agent class than the way all 30 teams are actually doing it, analytically?

Stephen Strasburg: photo credits Laurie Shaull

The crop of pitchers that will switch teams this offseason is one of the best we have seen in recent years. At the top of the group, and arguably the entire class, is flamethrower Gerrit Cole. When a struggling Cole was traded to the Astros from Pittsburgh in 2017, he was instructed to ditch his awful sinker and throw his top-notch curveball more. He was also directed to throw his fastball more, and to locate it up in the zone. The Astros analytics staff told him that his high spin rate would make the ball to appear to rise to hitters, and generate more whiffs if he located it up. Immediately, Cole exploded into one of the best pitchers in the league, and has not looked back since. This past year, he posted a 2.50 earned run average (ERA), 2.64 fielding independent pitching (FIP) and 2.48 expected Fielding Independent Pitching (xFIP), all topping the AL. Even more impressively, he led the league with a 16.8% whiff rate (SwStr%) and an outrageous 39.9% strikeout rate (K%) en route to racking up 326 strikeouts (K). Cole had 7.4 wins above replacement (WAR) this year, and will likely have 6 or 7 more for many years to come. Despite those numbers, Cole was somehow outpitched in the World Series by Stephen Strasburg, who is also a free agent. The World Series MVP is no slouch either, as evidenced by his 3.32 ERA, 3.25 FIP and 3.17 xFIP. While he isn’t as good as Cole at generating swings and misses, he is still very good at striking out hitters and inducing swings at bad pitches. His 29.9% K rate and 37.2 % out of [strike] zone swing percentage (O-swing%) were 5th and 2nd in the NL respectively. Not only that, but when Strasburg allowed contact, it usually wasn’t very threatening, as he gave up 1.78 ground balls for every fly ball (GB/FB ratio). All in all, Strasburg tallied 5.7 WAR in 2019, and he is only getting better. Meanwhile, his ex-division rival Zack Wheeler is also looking for a new team this winter. The former Met had a breakout season in 2018 with a 3.31 ERA, but his peripherals regressed in 2019. Despite some counting stats showing worse performance, Wheeler managed to accrue 4.7 WAR this past season. His stuff is unquestioned; his arsenal is similar to Gerrit Cole’s and Jacob deGrom’s in terms of pitch type, velocity and movement. However, he tends to be less successful than that pair. A recent article on Fangraphs suggests that Wheeler either elevate his fastball more or throw his dynamic slider more in two strike counts to get more punchouts. While employing these strategies won’t fix all of Wheeler’s problems, it’ll certainly help. Just the notion that a pitcher could be so similar to the two best pitchers in the world, put up 5 WAR and still have room to improve shows Wheeler’s potential. Add in the fact that he was in the 90th percentile of opponent exit velocity (EV), and it is clear why Wheeler is such a prized free agent this year.

The fact that I haven’t even gotten a chance to discuss the National League’s best pitcher for the first four months of the season speaks to the depth of this year’s pitching class. Incheon, South Korea’s, Hyun-Jin Ryu was absolutely amazing for the Dodgers in the regular season. Ryu carried a 1.53 ERA into August, and despite being pathetic in that month, he still finished the year with a 2.32 ERA, way ahead of the rest of the MLB. His 3.10 FIP and 3.22 xFIP rank near the top of the league as well. He rode his excellent changeup to a 1.98 GB/FB ratio, and overall did a great job of preventing hard contact. However, where Ryu really opens eyes is in his control and command. Hyun-Jin Ryu walked 24 batters in 182 innings in 2019, good for a cold 3.3% walk rate (BB%). Overall, he posted 4.8 WAR this past season. There are some talented relievers looking for new teams as well this offseason, most notably Will Harris. Even though he’s probably known best for blowing the Astros’ lead in game 7 of the World Series and creating a deficit they would never recover from, Harris was actually a spectacular pitcher last year. His 1.50 ERA over 60 innings was the best in the American League, and his 0.93 walks plus hits per innings pitched (WHIP) and .235 weighted On Base Average allowed (wOBA) support the narrative. However, some stats indicate this cutter-curveball combo pitcher may not be able to sustain that level. His .245 batting average on balls in play (BABIP) appears to be set to regress, and his .269 expected weighted On Base Average (xwOBA) suggests he is giving up harder/more solid contact than his numbers reflect. Still, even if more balls drop in for hits next year, Harris will still be a well above average relief pitcher in 2020.

Anthony Rendon: photo credits Laurie Shaull

While the free agent crop on the other side of the ball isn’t as deep, it still has some household names. First is the NL MVP candidate Anthony Rendon, who was outstanding in literally every way last year. His Fangraphs statistics page literally makes one roll their eyes at his greatness. Last year, his triple slash line (batting average (AVG)/on base percentage (OBP)/slugging percentage (SLG)) was .319/.412/.598. Additionally, his wOBA was a fantastic .413, and his OPS was an all-world-like 1.010. Normally, I wouldn’t bring up batting average because of its poor predictive value. In other words, a player’s batting average in a year says very little about his batting average and overall performance in the next year. However, I brought it up here because Rendon’s statcast contact metrics back up all of his regular rate stats. His expected Batting Average (xBA) was .319, the exact same as his AVG, and his expected Slugging Percentage (xSLG) was only one point off at .599. Most importantly, his xwOBA was .418, also lining up with his wOBA. Not only that, but the ex-National rarely swings and misses, as his 5.1% SwStr% was the best in the NL. He also had the highest wFB/C in the NL, making him the best fastball hitter in the league last year. Altogether, Rendon accumulated 7.0 WAR, and will probably keep doing so in the next decade. Rendon’s division was very deep in third basemen. While of course being headlined by himself, Atlanta’s Josh Donaldson is not only the next best third baseman in the 2019 NL East, but also the second best free agent infielder in the MLB. Where Donaldson excels is in his ability to hit the ball very hard. In fact, he did so on just under 50% (hard hit %) the time, ranking top 10 in the league. He had a barrel rate of 15.7%, and a very high average EV as well. Not only that, but the 2015 AL MVP is also very patient, walking 15.2% of the time, good for top 10 in the NL. Donaldson’s 4.9 WAR marked a great bounceback from a 2018 season where he was between multiple teams looking for a job after an injury. Another National League third baseman is probably the third best infield free agent in baseball. Brewer’s Mike Moustakas is a pure slugger who is perfect for anyone looking for home runs (HR). With 38 HR in 2017 and 35 in 2019, nobody is questioning the Moose’s reputation as a slugger. Just for good measure, I’ll add that the 2015 World Series champion and three time all-star has a 45% FB rate. In other words, in order to hit home runs, you need to hit the ball in the air, and Moustakas does a lot of that. However, Moustakas isn’t amongst the top free agents in terms of defense or speed according to statcast metrics. A team signing him knows they are just getting him for the dingers.

Yasmani Grandal (sorry about the old photo)

Yasmani Grandal, who actually signed with the Chicago White Sox (4 yr/$73 mil, $18.25 mil AAV), is the best catcher in this year’s class, and it isn’t very close. He had a 121 weighted Runs Created plus (wRC+) in 2019. Also, his 17.2% walk rate was on par with Alex Bregman and behind only Mike Trout. Even though Grandal was actually the worst cutter hitter in baseball by wCT/C, that clip can be dismissed due to small sample size or due to the fact that few pitchers throw cutters. Also, he hit every other pitch at an average to above average clip. Finally, Grandal is one of baseball’s premier pitch framers. He has consistently been at the top of the framing runs saved leaderboard for several years, and he will look to transfer his 5.2 WAR in 2019 to even greater things in Chicago’s south side. Next is a player who has a qualifying offer (QO) from the team on Chicago’s north side, Nicholas Castellanos. The long time Tiget got hot last year afte being traded to the Cubs at the deadline, logging a .646 SLG and 1.002 OPS in that time. Despite his rough start, he had a career high .525 SLG in 2019, and also had very good contact stats. Castellanos, a good fastball hitter, also put up 121 wRC+, and at only 27 years old, that number is likely to keep improving. However, the outfielder is very impatient, as he has a 6.3% walk rate and a 40% O-swing%. Also, his fielding is significantly below average. However, Castellanos is still the best outfielder in the 2019 free agent class. Close behind is his division rival, ex-Cardinal Marcell Ozuna. While Ozuna had a down year, especially in terms of batting average (he hit .243, a career low), this season seems like a fluke. The outfielder was suffering from some poor batted ball luck, as he had a .259 BABIP, the worst of his career by about 50 points. On top of that, his average EV of 91.8 mph, xwOBA of .382 and xSLG of .548 all were in the 91st percentile or higher. Most impressively, he had a hard hit rate of 48.1%, putting him in the top 10 of all qualified MLB players. Additionally, Ozuna showed the baseball world last year that he is getting more patient at the plate, as he posted career bests in walk rate and O-swing%. Even though his fielding may bring his value down a little, Ozuna is still a great fit for any team that needs a big bat in the outfield.

Of course the players analyzed above aren’t the only available players who can make positive impacts on their new teams. Brett Gardner, Madison Bumgarner, Craig Stammen, Avisail Garcia and Wade Miley are only a few of the other quality players moving on to the next stages of their careers. Also, while it is important to use analytics in decision making and player evaluation, one can’t rely on them to do the entire process. Analytics are a tool to help better understand player’s true talent levels and better predict future player’s performances. Considering that this tool is not only being more widely used but is also constantly improving and getting more efficient, it is imperative to understand them and apply them in real situations, just as we have done right here with MLB free agency.

This article was originally written for For The Win Magazine (FTW) at the Horace Mann School. However, with the permission of FTW Editor in Chief Robbie Werdiger, it was published on Medium prior to the release of the issue of FTW containing it.

— Jake Federman

Horace Mann School ’21

New York, NY

Email me at or

Data from Fangraphs, Baseball Savant and Baseball Reference

Follow my Twitter @federmanjake

Check out my other articles

All rights reserved