Baserunner Scoring Percentage (BSP): an analysis using play-by- play data William Knapp and Dr. Jason A. Osborne, Dept. of Mathematics and Dept. of Statistics, N.C. State University
Introduction What is BSP? Which teams have the best BSP? Effect of sacrifice bunts and stolen bases BSP as a predictor for runs scored and wins Methodology Linear and logistic regression Correlation
Data Acquisition Web-scraping and parsing with R and Java R functions: readLines() and regexp() Java: sort into arrays and change certain variables to binary
Analysis: Sacrifice and SB effects OutsBuntAttemptSBattemptEst. ProbStd Error 0No NoYes YesNo Yes No NoYes YesNo Yes No NoYes YesNo-- 2Yes --
Analysis: MLR on Runs and Wins Variables# of Predictorsrsqrpgrsqwlpct obp bsp hr lob obp bsp obp hr obp lob bsp hr bsp lob lob hr obp bsp hr lob
BSP vs Winning Percentage Scatter of all 30 teams over all 16 years Individual teams to show trends Houston needs some work