Overview

         This book describes a new baseball statistic I call GPA, or Gross Productivity Average. The first third of the text describes how GPA was created; the rest of the book applies GPA to settle long-standing controversies in baseball. The information in this book will allow everyone from baseball professionals to average fans to better understand the game.         Growing up a Red Sox fan, I learned to live with memorable disappointments. The Red Sox lost Game 7 of the 1975 World Series to the Reds thanks to Joe Morgan's bloop single in the top of the ninth inning to score Ken Griffey with the go-ahead run. Bucky Dent's three-run home run over the Green Monster gave the Yankees the lead for good in the seventh inning of a one-game playoff at Fenway Park for the 1978 American League East title. The Red Sox were one out away from winning the 1986 World Series when the Mets staged a historic rally which culminated in Mookie Wilson hitting a slow ground ball under Red Sox first baseman Bill Buckner's glove to win Game 6. The Red Sox went on to lose the World Series two days later.         Those painful losses have stayed with me my whole life. I have continued to think and rethink about the poor strategic decisions I felt were made by Red Sox managers over the years. What was I seeing that they couldn't see? Could a statistic be invented that would enable teams to employ more effective strategies in key game situations and get better production out of their players?         Jim Rice was elected to the Hall of Fame in 2009 for a stellar career that lasted from 1974 to 1989 and included 382 home runs and a 0.298 batting average. During the 1984 season he had a 0.280 batting average with 122 RBI and 28 home runs.         What bothered me about Rice's 1984 season was that he grounded into a then-record 36 double plays and struck out 102 times while walking only 44 times. I found it hard to believe that the benefits of having him hit in the heart of a potent lineup were not outweighed by the rallies he killed and runners he left on base throughout the year. There were no statistics at the time that could definitively determine how much of a benefit Rice was to his team that year. I began to think about creating a baseball statistic that would more accurately reflect a player's productivity than the oft-cited, traditional statistics batting average, RBI, home runs and slugging average.         Bill James defined sabermetrics as "the search for objective knowledge about baseball." James pioneered the in-depth analysis of baseball statistics when he published his Bill James Baseball Abstract beginning in 1977. His first edition presented statistics compiled from box scores from the 1976 baseball season.         James was frustrated because it was difficult for him to obtain detailed play-by-play information from past baseball seasons. When he contacted the Elias Sports Bureau, the official American and National League statistician since the 1920s, James was denied the detailed play-by-play statistics he wanted for his Baseball Abstract.         James did not accept Elias's rejection. Instead, he started Project Scoresheet in the 1980s to gather and share detailed play-by-play data from scorecards created by fans at the ballpark and from fans watching the game on TV or listening on the radio. Project Scoresheet eventually folded, but Dave Smith, who had worked on Project Scoresheet during its final years, then formed Retrosheet (retrosheet.org), which today provides a free, comprehensive play-by-play database of almost all major league baseball games played from 1947 to the current season.         The GPA statistic was created from the play-by-play data of all games from 1997 to 2009. The data is adjusted so that average GPA is equal to the average batting average of all major league players from 2005 to 2008. It works out well that the average GPA and batting average of all major league players from 1952 to 2012 are virtually identical at 0.259. This sets a modern standard for major league players untainted by steroids. Reporting GPA on a scale similar to that used for batting average makes it easy for people to understand.         Baserunning can significantly affect a hitter's or pitcher's production for his team. How much did Rickey Henderson's record 130 stolen bases add to his productivity for the 1982 Athletics? Which players had the most and least productive single seasons on the base paths? How much production does a knuckleball pitcher lose from all the stolen bases he allows and wild pitches he throws? How has baserunning changed over the years? These questions can be answered using a baserunning correction that is applied to GPA.         The ballpark can have a dramatic effect on a hitter's or pitcher's production. How much did playing at Coors Field help the average hitter or hurt the average pitcher? This book describes a ballpark correction that allows a player's production as measured by GPA to be accurately adjusted for the mix of ballparks and the era in which he played. Corrections are provided for every major league ballpark used from 1952 to 2012.         Offensive production has varied greatly from era to era and to a lesser extent from seasonto season. Over the years rules changed, different stadiums came into use, baseballs were manufactured by different companies and to different specifications, managers employed new strategies and players grew bigger, stronger, and even more athletic. Because all these changes occurred over time, it is difficult to look at any one statistic from 1952 to 2012 and get a clear picture of the change in offensive production taking place.         How was offensive productivity affected during the pitching-dominated years from 1970 to 1984 and the Steroid Era (defined here as the 1994-2004 seasons)? During what year did steroid use most affect the game? How much of the increased offensive productivity seen from 1994 to 2004 was due to steroids? Gross Productivity Average can answer these questions using a technique to filter out the other changes occurring over time.         Some baseball strategies have been endlessly debated without any consensus reached s to how or when they should be applied--or whether they should be applied at all. Where should the best hitter bat in the lineup? When should a hitter be intentionally walked? When should a sacrifice bunt be attempted? How often does a runner need to be successful when advancing a base in order to justify the risk?         This book will describe a computerized game simulation, developed independently of GPA, that allows these questions to be answered and definitive guidelines created. As the simulation demonstrates, the number of runs a team scores and the number of games it wins are determined by the GPA of the players in the lineup. The simulator shows that every player with the same GPA is equally productive for the team no matter how many home runs or singles he hits, no matter how many times he strikes out or walks, no matterhow many double plays he hits into.         With GPA, it is also possible to look back at every season from 1952 to 2012 and determine which player was the most deserving of the Cy Young and Most Valuable Player awards.         My background is primarily in computer science, not statistics. It is not necessary to have a background in statistics to understand GPA or the information presented here. This book avoids complex statistical methods, relying instead on the computer to produce a comprehensive and easily understood evaluation of the play-by-play data in the Retrosheet database.         My hope is that both baseball professionals and the average fan will come to accept GPA as the new standard for evaluating a major league player's productivity.