Looking to answer two questions regarding the relationships between Bat Speed, Squared-Up%, and batted ball outcomes (xwOBAcon)
What is the corresponding impact on Squared-Up% following increases/decreases in Bat Speed?
If a tradeoff exists, is there a “sweet spot” point or threshold that a hitter would ideally land between the two metrics for optimal performance outcomes (xwOBAcon)?
Performing time series modeling, results suggest there is a direct, albeit marginal, tradeoff between Bat Speed and Squared-Up%, where for every 1 MPH increase in bat speed from one month to the next, we expect to see a corresponding 0.7-0.8% decrease in Squared-Up%.
Some hitters still find success by swinging a fast bat even when not squaring up balls at a high rate, and some can also find success with a slow bat but maximizing exit velocities with a high Squared-Up%. So when a player falls into one of these buckets at either end of the spectrum, how might we determine, from a player development perspective, whether a hitter could benefit from training that would seek to improve the lesser metric of the two even though we would expect that doing so would have a negative tradeoff with the metric they excel at? With bat speed training being the big trend in player development currently, I’m particularly interested in whether there is a point when Bat Speed becomes “too high” and performance ends up suffering as a result of the sacrificed Squared-Up%.
Fitting a GAM model to predict xwOBAcon based on the relationship between Bat Speed and Squared-up% for players by season between 2024-2025, we do NOT see a point where xwOBAcon peaks and then subsequently declines. Rather, we see a continuously positive relationship between xwOBAcon and Bat Speed even when accounting for the tradeoff between Bat Speed and Squared-Up%. We can see this visually below, where our solid line, the total effect of Bat Speed and Squared-Up% on xwOBAcon, never crosses zero. If it had, it would have indicated there was a point where Bat Speed gains negatively affected xwOBAcon.
However what we do see is that as Bat Speed increases, the decline in Squared-Up% increasingly offsets gains in xwOBAcon. So theoretically, if you were to keep increasing Bat Speed without a ceiling, you would eventually reach a point where the loss in Squared-Up% completely cancels the gain from Bat Speed. But based on the scale of Bat Speed data, it never gets high enough where we see performance completely plateau.
So that then raises the question, do we ever get to a point where Bat Speed’s effect on xwOBAcon becomes diminished by a significant amount? It doesn’t appear so.
We don’t even reach a 10% loss of Bat Speed’s maximum effect until roughly 76 MPH of Bat Speed, which is quite high (single season max in my dataset is ~80 MPH). Graphing this “Cancellation Share”, or, how much of Bat Speed’s contribution towards xwOBAcon is cancelled out by the decline in Squared-Up%, everything on our scale is at least 20% or lower, signaling modest effects at best.
Evaluating the results of this study, is there anything we should take away from a player development perspective? Diving into this has allowed us to gather statistical evidence that the effects of declining Squared-Up% are not strong enough to warrant new consideration or process when training to increase Bat Speed. The known performance benefits of Bat Speed are enough to make it a focus in most player development plans.