Does a High Exit Velocity Make a Good Hitter?: An Examination of Exit Velocity in Comparison to Old and New Statistic Standards
Cole Hare
2023-03-21
0.0.1 Introduction
The game of baseball has seen a significant shift in the way organizations evaluate player performance. The old standard relied heavily on traditional metrics like batting average, but has since been challenged by a new age of thinking that places greater emphasis on on-base plus slugging (OPS). An article written by Zachary D. Rymer in 2014 describes why teams began to focus more on OPS as the new standard metric for determining hitter success. Rymer (2014) states:
You don't just see average, homers and RBI when baseball telecasts introduce
hitters. They now tend to include OPS, which is the most basic need-to-know
saber-stat in existence.... Crude as it is, it is a better reflection of a
hitter's talent than the traditional trio of average, homers and RBI.
Based on this determination, OPS will be the new standard statistic for evaluating hitter success, while batting average will be the old standard.
Additionally, Major League Baseball has been at the forefront of this statistical revolution, with teams employing advanced metrics to gain a competitive edge. One of the most significant developments has been the use of exit velocity. The MLB has begun incorporating exit velocity more and more into games in a number of different ways. One of the most prominent is through the use of Statcast, a tracking technology that captures data on every pitch and batted ball during games. Statcast measures exit velocity using high-speed cameras and radar equipment installed in every MLB stadium. Furthermore, national broadcasts of Major League games have incorporated exit velocity more, as broadcasters often cite a player’s exit velocity when discussing a certain at bat or their overall performance. This has helped to raise awareness of exit velocity among casual fans and has helped to further popularize the metric. It’s very apparent Major League Baseball and it’s fans are enamored with exit velocity numbers, but it does pose the question, does exit velocity matter when determining a good hitter?
0.0.2 Purpose and Hypothesis
The purpose of this research is to show that as standards for evaluating hitters have changed, exit velocity has begun to matter more when evaluating hitters. To prove this, linear regression models and correlations statistics will be calculated between exit velocity and OPS, as well as exit velocity and batting average. It is anticipated that the advanced metric, exit velocity, will have a stronger correlation with the new standards (OPS) in comparison to the old standard (AVG). By using linear regression and correlation statistics, insight into the direct relationship of the metrics used in the model will be provided. These statistics will help to explain why MLB organizations have shifted their focus to advanced metrics like exit velocity when evaluating hitters.
0.0.3 Data
The data being used is a list of the top 100 hitters in the MLB with the highest average exit velocity and their corresponding batting statistics. All data is the cumulative average of each player through the 162 game season in 2022 and is pulled from Baseball Savant.
0.0.4 Results
0.0.5 Batting Average
ggplot(data=velostats, aes(x=exit_velocity_avg,y=batting_avg)) +
labs(x = "Average Exit Velocity (MPH)", y = "Batting Average", title = "Average Exit Velocity vs. Batting Average") +
geom_point(aes(col=exit_velocity_avg)) +
scale_colour_gradientn(colours=c("blue","red")) +
geom_smooth(method = "lm")## `geom_smooth()` using formula = 'y ~ x'
Fig 1. This figure examines the relationship between hitter’s exit velocity and batting average.
batavg.lm <- lm(batting_avg~exit_velocity_avg)
stargazer(batavg.lm, type = "text")##
## ===============================================
## Dependent variable:
## ---------------------------
## batting_avg
## -----------------------------------------------
## exit_velocity_avg 0.002
## (0.001)
##
## Constant 0.092
## (0.106)
##
## -----------------------------------------------
## Observations 130
## R2 0.019
## Adjusted R2 0.011
## Residual Std. Error 0.028 (df = 128)
## F Statistic 2.487 (df = 1; 128)
## ===============================================
## Note: *p<0.1; **p<0.05; ***p<0.01Our first scenario focuses on the old standard in the MLB, which put a heightened emphasis on batting average. The test aims to determine if there is a significant linear relationship between batting average and exit velocity average. Because batting average is becoming outdated, there is reason to believe there will not be a strong positive relationship between the two metrics.
The regression analysis shown indicates that the estimated slope coefficient (i.e., exit_velocity_avg) is 0.002, which implies that, on average, an increase in exit velocity by one MPH is associated with an increase in batting average of 0.002 points. This means that even if a player works to increase their average exit velocity by five MPH, they’d only expect to see their batting average increase by a point; an increase that small wouldn’t convince an organization of improvement from that hitter.
The p-value for this coefficient is 0.117, which is greater than the significance level of 0.05. This means we fail to reject the null hypothesis that there is no linear relationship between the two variables.
It is worth noting that other exterior factors do impact batting average as well. However, for this research just examining these two variables makes sense because it allows us to have a straightforward comparison between these findings and the later findings of the relationship between OPS and exit velocity.
Based on the output and analysis it is fair to see why exit velocity wasn’t highly considered in the old standard of evaluating hitters. The correlation between a players batting average wasn’t dependent upon their exit velocities, therefore it wasn’t highly thought of during player evaluations.
0.0.6 On Base Plus Slugging (OPS)
ggplot(data=velostats, aes(x=exit_velocity_avg,y=on_base_plus_slg)) +
labs(x = "Average Exit Velocity (MPH)", y = "OPS", title = "Average Exit Velocity vs. OPS") +
geom_point(aes(col=exit_velocity_avg)) +
scale_colour_gradientn(colours=c("blue","red")) +
geom_smooth(method = "lm")## `geom_smooth()` using formula = 'y ~ x'
Fig 2. This figure examines the relationship between hitter’s exit velocity and OPS.
OPS.lm <- lm(on_base_plus_slg~exit_velocity_avg)
stargazer(OPS.lm, type = "text")##
## ===============================================
## Dependent variable:
## ---------------------------
## on_base_plus_slg
## -----------------------------------------------
## exit_velocity_avg 0.022***
## (0.003)
##
## Constant -1.179***
## (0.265)
##
## -----------------------------------------------
## Observations 130
## R2 0.296
## Adjusted R2 0.290
## Residual Std. Error 0.071 (df = 128)
## F Statistic 53.750*** (df = 1; 128)
## ===============================================
## Note: *p<0.1; **p<0.05; ***p<0.01Our second scenario emphasizes the new standard, using OPS and examining its relationship with exit velocity. This test aims to determine if there is a significant linear relationship between on-base plus slugging (OPS) and average exit velocity. Because OPS is becoming more of the standard benchmark for hitter success, there should be a strong positive relationship with exit velocity.
The regression analysis provided shows that the estimated slope coefficient implies that, on average, an increase of one MPH in exit velocity will cause an increase in OPS of 0.022 points. This is incredibly significant when looking at ways to improve hitter success. If a player can see an uptick in their OPS from say, .680 to .700, by working on increasing their average exit velocity by just one MPH this could play major dividends for the team. This increase is also much more significant than the one seen in the batting average model.
The p-value for this coefficient is much smaller than the significance level of 0.05. This indicates strong evidence to reject the null hypothesis and accept the alternative hypothesis that there is a significant linear relationship between the two variables.
Based on these results, there is strong evidence to suggest that there is a significant linear relationship between on-base plus slugging and exit velocity average. An increase in exit velocity average is associated with an increase in OPS, and the regression model can explain a significant amount of the variation in OPS. These findings align with the newer standard of thinking in baseball as exit velocity is a direct measure of how hard a ball is hit, which has a direct impact on the ball’s trajectory and how far it travels. This is reflected in OPS, which combines a player’s on-base percentage and slugging percentage, both of which are heavily influenced by how hard a player hits the ball. Therefore, exit velocity is considered to be a valuable metric in evaluating a player’s hitting ability and is used in conjunction with other advanced statistics to get a more comprehensive understanding of a player’s performance.
0.0.7 Conclusion
stargazer(batavg.lm, OPS.lm, type = "text")##
## ===========================================================
## Dependent variable:
## ----------------------------
## batting_avg on_base_plus_slg
## (1) (2)
## -----------------------------------------------------------
## exit_velocity_avg 0.002 0.022***
## (0.001) (0.003)
##
## Constant 0.092 -1.179***
## (0.106) (0.265)
##
## -----------------------------------------------------------
## Observations 130 130
## R2 0.019 0.296
## Adjusted R2 0.011 0.290
## Residual Std. Error (df = 128) 0.028 0.071
## F Statistic (df = 1; 128) 2.487 53.750***
## ===========================================================
## Note: *p<0.1; **p<0.05; ***p<0.01As the game of baseball continues to evolve, it is likely that newer approaches to evaluating hitters will focus more on exit velocity as it becomes increasingly aligned with OPS which has become the new standard when evaluating hitters. These findings support the hypothesis that the old way of thinking, which relied heavily on traditional metrics like batting average, will continue to be challenged by a new age of thinking that places greater emphasis on advanced statistics like exit velocity, along with other statcast metrics.
0.0.8 Sources
Statcast custom leaderboards. baseballsavant.com. (n.d.). Retrieved March 23, 2023, from https://baseballsavant.mlb.com/leaderboard/custom?year=2022&type=batter&filter=&sort=4&sortDir=desc&min=q&selections=xba%2Cxslg%2Cxwoba%2Cxobp%2Cxiso%2Cexit_velocity_avg%2Claunch_angle_avg%2Cbarrel_batted_rate%2C&chart=false&x=xba&y=xba&r=no&chartType=beeswarm
Rymer, Z. D. (2017, October 3). Sabermetrics for dummies: How-to guide for MLB fans to learn the ropes. Bleacher Report. Retrieved April 17, 2023, from https://bleacherreport.com/articles/2040748-sabermetrics-for-dummies-how-to-guide-for-mlb-fans-to-learn-the-ropes#:~:text=The%20Best%20Ways%20to%20Evaluate%20Hitters&text=For%20those%20who%20don’t,Hitters%20exist%20to%20score%20runs.