The Drastic In-Season Turnaround of Nick Castellanos
The Phillies mainstay was among the worst players in the league. Now, he's a key component to Philadelphia's championship pursuit.
On September 1st, Nick Castellanos stepped to the plate in the bottom of the 11th inning of the Phillies Sunday Night Baseball showdown against their archrival Atlanta Braves. He would be the last batter to do so that night. Castellanos hit a walk-off single to center to give the Fightin’ Phils a 3-2 win and series victory over their NL East foe. But we’ve seen that script before. Castellanos’ four walk-off RBIs are tied for the most in Phillies history since RBI became an official stat in 1920.1 It is no wonder Casty has become somewhat of a fan-favorite in the city of brotherly love. But that love is likely a bit stronger now than it may have been earlier this season. In 2024, Castellanos has turned a season that was on track to be among the league’s worst into one that has played a pivotal role in the Philadelphia Phillies quest for a third World Series title.
May 24th, 2024
On May 24th, 55 games into the season, Castellanos was batting 0.198 with an OPS+ of 63 (37% below the league average) and had -1.2 WAR, which ranked second-to-last in all of baseball.2 In fact, among qualified batters over the entirety of the season, Castellanos holds the lowest WAR during that timespan of Opening Day to May 24th.3
Turning it Around
In the 92 games since that day in May, Castellanos has hit .267 with an OPS of 0.774 (giving him an OPS+ 23% above the MLB average at 123). His 1.1 WAR/162 is a whopping 4.6-points higher than his first 55 games. In fact, only eight players have turned around their season more drastically than Castellanos between those time periods. Only two of them, Randy Arozarena and Corbin Carroll, ranked in the bottom 15 in WAR during the Opening Day to May 24th period.
How Did This Happen?
While Casty’s performance statistics clearly took a drastic turn for the better, his underlying metrics - which examine how he played and his approach, rather than the outcome - remained relatively unchanged. Castellanos has a notoriously poor eye at the plate and chases pitches out of the zone at a rate far above average. However, this remained fairly constant between the two periods, with a 37.8% O-Swing% (percent of times a batter swings at balls outside the strike zone) from Opening Day through May 24th, and a 38.6% O-Swing% since then. If anything, his eye for the ball was actually a touch worse during his period of improved performance. Castellanos was not driving the ball considerably harder either, with his average exit velocity rising only a mere 0.6 MPH, good for just the 38th best improvement among 103 qualified batters between the two periods. His Barrel% - a Statcast metric aimed to quantify the frequency in which a batter produces an optimally hit ball based on exit velocity and launch angle - rose only 1.6% from 6.4% to 8.0%.4 This was again not significantly better than the rest of the league, ranking as just the 34th best improvement across the league. Similarly, his hard-hit rate - the percentage of balls hit 95 MPH or harder - was identical between the two periods at 37%. Lastly, Castellanos struck out only 2.2% more during his early season struggles. While that may be a good improvement, 31 qualified batters had better improvements in strike-out rate. So what exactly caused this drastic turnaround? Was it all just luck? I guess you could say that…
A Victim of Batted Ball Variance
The one statistic that really stood out between the two time periods for Castellanos was Batting Average on Balls in Play (a.k.a., BABIP). This is a recalculation of batting average only on balls put in play, meaning it excludes the "three true outcomes" of home runs, walks or strikeouts.5 If you have ever watched a baseball game, you have likely seen a player hit a rocket, but the ball just happened to go straight to the third baseman. A foot to the left or right, and its an easy single in the hole or double down the line. Similarly, you’ve probably seen a weak dribbler turn into a hit, just because the infield was playing back or the ball rolled just far enough from any fielder to allow the lucky batter to reach first safely. BABIP is a highly variable statistic based on quality of the opposing defense, field conditions, and just plain random luck. Batters and pitchers have little control over what happens once the ball is put in the field of play, so sometimes high quality contact can result in an out, while low quality contact results in a hit. Over the long term this will even out, but it takes a fairly large sample size of balls in play to do so.6 While a player may hold a comparatively high or low BABIP for a fair amount of time, the number tends to regress toward their career average over a growing sample size. Nick Castellanos had a BABIP 0.096 points below his career average over his first 55 games, this was the second-greatest difference in terms of BABIP below career average in the MLB during that timespan. Since then, his BABIP has regressed back toward his career number.
Castellanos’ BABIP regressed toward his career average at the 6th-greatest extent among all qualified players who saw positive regression, moving 0.076 points. Based on the stark difference between his career BABIP and first 55 game BABIP, we should have seen this coming. There is a jaw-dropping -0.86 correlation between the difference in career versus pre-May 24th BABIP and the change in BABIP a player saw between the first and second periods we’ve been assessing. This translates to period-one BABIP (Opening Day through May 24th) accounting for a whopping 73.2% of the variance in the change in BABIP a player would be expected to see in the following 3-plus month period.
Conclusion
Nick Castellanos has played in every one of Philadelphia’s 146 games this season. Even through his statistically lowest of low points, manager Rob Thompson maintained faith and kept writing his name in the lineup. A player with a BABIP that is substantially different from their career number usually is a sign of luck. The tricky part is, BABIP requires a large sample before it “stabilizes,” so you can’t be confident a batter has established a new level of BABIP without an extremely large sample.7 When we see a player with a batting average on balls in play over a small sample of games differ so greatly from what they’ve done over their entire career, we should expect that difference to gradually shrink and move closer and closer to that career number. Regarding Castellanos, he was never that bad, just incredibly unlucky. The Phillies, however, are very lucky to have him.
Shout-out to the great Sarah Langs for that nugget.
Only Andrew Benintendi of the lowly White Sox has a lower WAR with -1.4
Among those qualified during both the Opening Day to May 24th timespan, and May 25th to September 9th timespan.
Statcast defines a “Barrel” as a well-struck ball where the combination of exit velocity and launch angle generally leads to a minimum .500 batting average and 1.500 slugging percentage. Per MLB.com, a batted ball is “barreled” when it has an exit velocity of at least 98 mph. At that speed, balls with a launch angle between 26-30 degrees always attain “barreled” classification. For every mph over 98 degrees, the range of launch angle expands.
Good research here Sean!