Perhaps one of the most intriguing aspects of major league baseball in comparison to other popular sports is the amount of tracked statistics and sabermetrics for every player in the league. With the recent introduction of Statcast tracking technology across all 30 major league ballparks, the amount of elements of the game that can be tracked has exploded. These new measured statistics from Statcast range from the spin rate of every pitched ball, to the probability that a fielder will make an out on a play. Given that most all MLB teams today play "Moneyball" to maximize their on-field performance and prioritize analytics when making decisions about what players to add to their roster, this begs the question: Do MLB teams value these new statcast statistics when determining a players value to the team, and thus, do these new metrics play a role in determining a players salary?
During the 2020 -2021 off season, it was revealed that there was rampant cheating among many pitchers in the MLB, as they were creating and using foreign "sticky" substances to increase spin rates on pitched balls. This increase in spin rate makes fastballs drop less, and increases the break on breaking and off-speed pitches. This makes the ball harder for a batter to hit, thus increasing a pitchers performance. Because spin rate has been one of the more controversial new metrics, the foreign-substance cheating scandal helps this study to show whether or not these new metrics mater in determining a pitchers salary, and whether pitchers who cheated were rewarded financially at a statistically higher rate than pitchers who did not use foreign substances.
The data used in this series comes directly from Statcast via Baseball Savant. The more "traditional" metrics of pitcher performance (such as earned runs, strike outs and walks) were obtained from Baseballreference.com, which is the largest resource for baseball data. Because Statcast was introduced in 2015, a pooled cross section of over 2200 pitchers was sampled for the period 2015 - 2020. Because the goal of this study is to determine whether teams value these new metrics, a starting pitchers salary is used as the dependent variable in all created models, as the more valuable to a team a pitcher is, the higher they will be financially compensated.
Model 1: A basic OLS regression model was created and estimated to test whether the values and significant for the "traditional" metrics matched what was found in the previous literature on this topic.
Model 2: Model two added the new Statcast metrics to test whether they were statistically significant in determining a starting pitchers salary, and to quantify how much the new metrics impact a starting pitchers salary.
Intertemporal Models: For the new statcast variables that were found to be statistically significant in model two, intertemporal models were estimated to determine what year MLB teams started valuing these new metrics, as well as to provide evidence that starting pitchers who cheated during the 2018 - 2020 seasons gained a measurable increase in compensation over pitchers who did not cheat.
Model 1: It was shown that all of the "old" metrics displayed the expected signs and statistical significance to what was estimated in the cited literature. This is good for this study, as it shows that the date present shows congruent results to other studies on this topic.
Model 2: It is determined that the new Statcast variables "Fastball Spin Rate," "Average Fastball Speed," "Slider Spin Rate," and "Expected OPS against" are all positive and significant (at a 5% confidence level) at explaining a starting pitcher's salary. For example, the model shows that a one percent increase in the average spin rate for fastballs that a pitcher throws leads to a .798 percent increase in salary.
Intertemporal models: The intertemporal model for "Fastball spin rates" shows that there is a positive and statistically significant increase in the relationship between fastball spin rates and starting pitchers salaries in the years 2015, 2017 and 2019, with values for 2019 being the highest in player compensation (when foreign substances were the most heavily abused.) The coefficient for fastball spin rates in 2019 shows that one percent increase in the average spin rate on pitched ball increases a starting pitchers salary by 1.37% which is almost twice as large compared to the estimate shown in model two.
* It is important to note that while these percentages seem small, the league minimum for players in the MLB is $550,000. Thus, a pitcher making league minimum increasing their salary by only .798 percent is a total increase of $4,389. Pitchers accused of cheating via foreign substances often increased spin rates by over 10%, leading to a big payday given by the models estimated here!
Given what the above models and intertemporal estimations show, some important conclusions can be made. First, it is obvious from these results that MLB teams do indeed value these new analytics made available by Statcast tracking technology. Insights about the proactiveness of MLB teams can be seen from the intertemporal models, where it is determined that teams were placing value on new metrics like spin rate in the first year that they were made available to teams.
These intertemporal models also provide supporting evidence that pitchers who cheated gained measurable financial compensation over pitchers who did not cheat via foreign substance usage. In 2019 when the cheating was reported to be at it's peak, average spin rates of pitched balls were valued higher than any other year where this statistic was measured. Given that it was proven that MLB front offices knew about pitchers cheating, compensating players at a higher rate for increasing average spin rates proves that pitchers who cheated gained significant financial compensation over pitchers who did not cheat. The value for this estimation shows that during this year, a pitcher who increased average fastball spinrates by one percent was rewarded with a 1.38% increase in pay, which is over twice the average for previous years. Because many players increased spinrates by 10% or more, this means that pitchers who cheated were most likely increased compensation by tens of thousands of dollars by doing nothing else but increasing the spinate of their pitches.
In conclusion, this study advanced the current literature on this topic by proving that MLB teams place monetary value the new Statcast metrics, and that any team that fails to use these new metrics when determining a players worth is not properly optimizing the overall performance of the team. It is also discovered here that the cheating scandal was not insignificant, and that pitchers who cheated saw measurable and significant returns to their salary by abusing foreign substances.
A CDFI (Community Development Financial Institution) is a government run partnership with a financial (i.e. credit unions and banks) that provides these institutions funds to lend to communities and businesses in traditionally credit underserved areas. This program is important as it allows financials to directly support credit underserved areas with funds to help improve the local community.
1) How do various social factors (i.e. such as whether the credit union is minority or woman owned, whether the credit union is in a rural or urban area, etc.) effect the amount/rate of CDFI funds lent out?
2) Do rural and urban CDFI Credit Unions lend CDFI funds differently?
The data set used consists of 217 variables across 48 different credit unions for the years 2003 through 2015
The independent variables of interest that were selected from the dataset are:
The dependent variables of interest are:
The model estimated was a pooled OLS model, where the dependent variable was CDFIloanratio.
While this model was not significant as a whole (shown by the insignificant F-statistic,) some interesting findings are still shown.
1)Minority owned Credit Unions lend CDFI funds at a rate of 43% more than non-minority owned Credit Unions.
2) Women owned and Faith-based Credit Unions did not lend CDFI funds at a significantly different rate.
3)The poverty and unemployment rate of the areas the Credit Union is in do not appear to affect the level of CDFI funds lent out.
4) Rural and urban credit unions did not appear to lend CDFI funds at statistically different rates.
Taylor Overton
Copyright © 2022 Taylor Overton - All Rights Reserved.