After much anticipation, the MLB Statcast team has begun to release bat tracking data, including metrics such as bat speed, swing length and “squared-up rate.” This development is particularly exciting as it enables public analysts to explore the underlying skillsets of hitters at a level never before possible. Tracking data is especially useful as it tends to reflect player ability in very small samples (even two swings, as per Tom Tango). This piece will be a bit different from my typical posts as instead of going in-depth on one player or topic, it will instead contain several quick-hitters I noticed during a first pass through the new data.
Bat Speed and Swing Length
One immediate trend that’s noticeable in the bat tracking data, pointed out by Justin Choi and others on Twitter, is that players with higher bat speed tend to have longer swings…
This does not necessarily mean that higher bat speed results in a longer swing (or vice versa) though. As David Gerth noted, there is significant potential for selection bias in this case. Bat speed is an indicator of a player’s power and swing length is potentially a measure of contact-related ability. Players that struggle in both metrics aren’t likely to be big leaguers, so they are excluded from this sample, which may cause these metrics to seem correlated when they are not. However, there is an alternative physics-based explanation for this phenomenon. Longer swings give more time for the bat to accelerate, which can allow for a higher bat speed at the point of contact. Interestingly, this relationship holds up on the play level, not just the player level, even when both bat speed and swing length are measured relative to the player’s seasonal average…
This suggests that the physics-based explanation potentially has merit, even if there may be some selection bias in the player-level sample.
Bat Speed Variation
Giancarlo Stanton swings the bat extremely hard. That is clear to anyone with eyes and is also evident from the bat tracking leaderboards, where Stanton leads MLB in average bat speed by a large margin…
But that’s not the only thing that makes Stanton unique. Notice that his Fast Swing Rate, at 98%, leads MLB by an even larger clip. Why is that? Stanton doesn’t only swing a fast bat, he also swings a consistently fast one. Compare his bat speed distribution with the player second on the average bat speed leaderboard, Oneil Cruz…
Quite a difference! Stanton has a very tight distribution of bat speeds, while Cruz’s is significantly more variable. This difference can be measured using the coefficient of variation of a player’s bat speed (standard deviation/mean)…
So, not only is Stanton the hardest swinger on average, but he also has the least variation in his swing speed. Others, like Cruz and Harrison Bader, have tended to vary their swing speed much more. Note that more or less consistent bat speed is not necessarily a positive attribute. It is a stylistic difference that may be positive or negative for different batters. More research is certainly needed in this area.
Swing Length Variation
This same analysis can be applied to a player’s swing length. Here are the players with the most consistent and inconsistent swing lengths, as measured by their coefficient of variation…
Jose Abreu’s swing length has been particularly inconsistent in 2024. Perhaps this has something to do with his tremendous struggles so far this season.
Another interesting observation regarding swing length variability is that players with shorter swings tend to exhibit more variable swing lengths…
I’m not particularly sure why this would be the case. Maybe short-swinged players can afford to lengthen their swing on occasion, while longer swingers are hard-capped at their length, as a longer swing would not be able to hit MLB velocity.
Performance vs Velocity
Discussion of swing length inevitably turns towards a player’s ability to hit high velocity. With this in mind, I chose to group players into 3 swing length groups and assess their performance versus high and low velocity…
These results seem to align with expectations. Players with longer swings (who also tend to have higher bat speed, see part 1), perform extremely well versus low velocity fastballs. However, they have a staggering dropoff in both contact rate and contact quality metrics versus higher velocity. In contrast, the short-swinged players (who also tend to have lower bat speed) don’t perform particularly well against high or low fastball velocity, but have a much smaller dropoff when compared with the long-swinged group.
Swing Length and Contact Point
When analyzing a player’s swing length, it is crucial to keep in mind contact point. Balls met out in front of the plate will naturally have a longer swing length due to their contact point. These “in-front” contact points typically result in pulled contact. As a result, there is a significant correlation between spray angle and swing length, as pulled balls will typically have a longer swing length…
Bat Speed and Launch Angle
Bat Speed (and resultantly exit velocity) can have a tremendous impact on quality of contact. However, this impact does not affect all batted balls equally. The value of fly balls is far more sensitive to exit velocity than grounders. Finally, we can (roughly) reproduce this Connor Kurcon plot using bat speed rather than the proxy of max exit velocity…
As expected, the difference in quality of contact between bat speed groups is magnified in the “line drive/fly ball zone” where home runs tend to occur. Note that because this plot uses a limited sample of exclusively 2024 batted data (when bat speed is available), I’ve used smoothed curves, which may gloss over finer distinctions in performance by launch angle.
Bat Speed and Exit Velocity
Unsurprisingly, there is a strong link between a particular swing’s bat speed and the exit velocity of a batted ball resulting from that swing. Perhaps more astonishing is that this relationship seems to level off at ~80 mph(?)…
I’m not particularly sure why this seems to be true. I surmised that this was perhaps due to high bat speed players (who typically swing >80 mph) potentially having worse “barreling” skills. However, this effect seems to hold up even more strongly when comparing a player’s bat speed on a given batted ball to their seasonal average…
Furthermore, it seems to apply not only to average exit velocity, but high-end EVs as well…
There is a relatively small sample of swings at these higher thresholds, so it’s hard to make any firm conclusions. It’s possible that I’ve made a mistake in my code or missed something in my analysis. However, this point does seem worthy of future research. One potential rationale for this effect could be that players have difficulty squaring up the ball on swing speeds significantly higher than their average rate. This would result in diminishing returns in exit velocity for an increase in swing speed.
Public access to bat tracking data is a further step in increasing our understanding of baseball. Discerning the the fundamental physical skills of players is key to analyzing the process that underlies their success, rather than just the result. I’m excited for the metrics, articles and insights yet to come in this new frontier.
Thank you for reading this post. Public bat tracking data is extremely new and all elements of this piece are food for thought and future research, rather than firm conclusions. Thanks to Tom Tango, Mike Petriello and the rest of the MLBAM team for their work to make bat tracking data public. I can be reached on Twitter here.
