Sports Analytics

Description

This book is a collection of applications of analytic techniques to a number of popular sports including baseball, basketball, hockey, Jai Alai, NFL football and horseracing. We focus on both the statistics of the sporting events and betting strategies on the events. The subject is fascinating as there are many twists and subtle complicated decisions. Sports analytics applies mathematical and statistical methods to important questions in the structure and performance of sporting activities using the same basic methods and approaches as data analysts in other disciplines. Sports games and events are a fruitful area for study and to evaluate betting strategies as there is extensive data and mean reversion. With prices changing continuously, risk arbitrage bets can be made. Moreover, little errors, like a penalty to a player or an error in a call by a referee, can change the score of a game and corresponding betting prices. The collection and analysis of in-game data can inform players, coaches and staff on effective decision making during sporting events. Novel features of the book include: an analysis of who were the greatest baseball batters; analyses of the players most important to team success (and they are not necessarily the best players) in basketball, NFL football and hockey; a tutorial on risk arbitrage and its applications to NFL football and NBA basketball; a discussion of many ad hoc decision rules by coaches and players and what was really optimal; in the racing section we discuss breeding, the analysis of various bets like the Rainbow and ordinary Pick 6, a discussion and betting on the most important races and a visit to the Breeders' Cup with Ed Thorp to demonstrate the place and show system in action. Contents Preface 1. Utility Theory and Preferences 1.1. Introduction Appendix: Technical Fundamentals for Utility Theory 2. Fortune’s Formula: How the Pros Wager Using the Kelly Capital Growth Investment Criterion? 2.1. Introduction 2.2. Unpopular Numbers 2.3. Good and Bad Properties of the Kelly Criterion 3. Rating Batters in Baseball 3.1. Introduction: Who is the Best? 3.2. Stochastic Dominance 3.3. Benchmarking 4. The Expected Utility of Performance: Dominant Batting Seasons in Baseball 4.1. Introduction 4.2. Aggregate Performance 4.2.1. Batting Impact 4.2.2. Dominance and Utility 4.2.3. Standardization 4.3. Comparison of Batters 4.4. Conclusion 5. Jai Alai Arbitrate Strategies 1. INTRODUCTION 2. THE PERFECT ARBITRAGE 3. RISK ARBITRAGES 4. FINAL REMARKS ACKNOWLEDGMENT REFERENCES 6. The Game Box Score in Basketball: Linking Statistics to Game Outcomes 6.1. Introduction 6.2. Performance Measurement: The Box Score 6.2.1. Event outcomes 6.2.2. Composite measures 6.2.3. Team value 6.2.4. The estimated player contribution 6.3. The 1989–1990 NBA Season 6.3.1. Team value 6.3.2. Player contribution 6.4. Discussion 7. Risk Arbitrage in the 2021 NBA Championship 7.1. What is Risk Arbitrage? 7.2. The NBA Playoffs 7.3. The NBA Finals 7.3.1. Game 1: Milwaukee Bucks vs. the Phoenix Suns in Phoenix 7.3.2. Game 2: Milwaukee Bucks vs. the Phoenix Suns in Phoenix 7.3.3. Game 3: Milwaukee Bucks vs. the Phoenix Suns in Milwaukee 7.3.4. Game 4: Phoenix Suns vs. the Milwaukee Bucks 7.3.5. Game 5: Phoenix Suns vs. Milwaukee Bucks in Phoenix 7.3.6. Game 6: Milwaukee Bucks vs. the Phoenix Suns at Milwaukee 7.4. The Road the Finals 7.4.1. Round 1 East 7.4.2. Round 1 West 7.4.3. Conference Semi-Finals: East 7.4.4. Conference Semi-Finals: West 7.4.5. Conference Finals Acknowledgements 8. Winning Hockey: Team and Player Impact in the NHL 8.1. Introduction 8.2. Winning Profiles 8.3. Team and Player Value 8.3.1. Event outcomes 8.3.2. Performance measure I: Elo 8.3.3. Performance measure II: The LOGIT 8.3.4. Connecting rating measures 8.3.5. Rating teams and players 8.4. NHL 2021 8.4.1. NHL teams 8.5. NHL North: Teams Values 8.6. NHL North: Player Effects 8.7. Discussion 9. NFL Analytics I 9.1. Expected Value of Field Position in Football 9.2. Combining Odds Ratios 9.3. Game Point Spread: Predicting Success 9.4. Aggregation of Performance Measures 9.5. Performance Measures: QB Rating 10. NFL Analytics II 10.1. Introduction 10.1.1. The jet lag effect 10.1.2. How to calculate the home field advantage 10.1.3. The decision on what to do on fourth down on the one yard line 10.1.4. The two versus one extra point fallacy 11. Stop Shy of the First Down 11.1. The First Down 11.2. The Choice 11.3. The Value of Possession, in Yards 11.4. Third and One 11.5. Second and One 11.6. Implied Value of Possession 11.7. Evidence of Poor Strategy 11.8. Conclusion 12. Efficiency in NFL Betting Markets 12.1. Introduction 12.2. Components of the Betting Decision 12.2.1. Odds and win probabilities in games 12.2.2. Point spread 12.3. Team Strength Ratings: The Elo System 12.4. Efficiency in the 2017 NFL Season 12.4.1. 538 Elo NFL Ratings for 2017 12.4.2. Probabilities and spreads from the Elo 12.5. Conclusion 13. National Football League: 2018–2019 Season 13.1. Introduction 13.1.1. 2018 Regular season: Teams 13.1.2. Players 13.2. Playoffs Teams in 2019 13.3. Wild Card Games 13.4. Division Champions 13.5. Conference Champions 13.6. Superbowl LIII 13.7. Awards 13.8. Conclusion 14. Review of the NFL 2019/20 Season, Playoffs and Superbowl 14.1. Introduction 14.2. NFL Regular Season: The Round Robin 14.3. The 2019–2020 NFL Playoffs 14.3.1. Division round 14.3.2. Conference round 14.4. Super Bowl 14.4.1. Awards 14.5. Conclusion: Looking Ahead 14.5.1. Scrambling 14.5.2. Prospects for 2020–2021 15. The COVID-19 NFL Playoffs and Super Bowl, 2020–2021 15.1. The NFL 2020–2021 Setting 15.2. NFL Super Bowl Betting 2021 15.3. The Evolution of the Betfair Odds Before and During the Playoffs and Super Bowl 15.3.1. Super Wildcard weekend January 9–10, 2021 15.3.2. AFC Games 15.3.3. NFC Games 15.4. Divisional Round, January 16, 17 15.4.1. The Conference Championship Games 15.5. The Super Bowl LV, February 7, 2021 15.5.1. Commentary 15.5.2. Final comments on betting 15.5.3. Epilogue 15.5.4. Appendix: The COVID-19 and injury effects 16. Team Composition: Are the Best Players on the Best Teams? 16.1. Introduction 16.2. Top Ranked Players 16.2.1. Number of top 400 players by team in 2017 by Sports Illustrated 16.2.2. NFL top 100 16.2.3. Comparison of SI and NFL rankings 16.3. Modeling the Dependence of Team Win Percentage on Top Players 16.3.1. Number of top players from Sports Illustrated 16.3.2. Weighting by position 16.3.3. Team performance: Predicted and actual 16.3.4. Model predictions 16.3.5. The Sports Illustrated list: Top 100 16.4. Conclusion 17. Primer on Dosage and the 2012 Triple Crown 17.1. Introduction to Dosage 17.2. The 2012 Triple Crown 18. An Application of Expert Information to Win Betting on the Kentucky Derby, 1981–2005 1. Introduction 2. The Racetrack as a Sequence of Markets 3. The Dosage Index and Performance Measures 4. Data Acquisition 5. Application of Breeding Information and Performance Measures to Refine Estimated Win Probabilities 6. The Betting Model 7. Conclusions Acknowledgements Notes References Appendix: Data Sources 19. Stochastic Programming and Optimization in Horserace Betting 19.1. Introduction 19.2. The Importance of Getting the Mean Right 19.3. The Favorite-Longshot Bias 19.4. Place and Show Optimization with Transactions Costs 19.5. The Place Pick All 19.6. Some Stochastic Programming Formulations 19.7. The Pick3 and Pick 4: Theory of Pricing the Bets 19.8. The Pick 4 19.9. Example of Pick 4 with embedded Pick3’s and Doubles: Travers Day at Saratoga 19.10. The Pick 6: Theory of Pricing the Bets 19.10.1. 2001 Breeders’ Cup Insurance bets for SCA 19.11. The One That Got Away: The Hitable $2 Million Pick 6 at the 2009 Breeders’ Cup 19.12. Professional Racetrack Betting Syndicates 19.13. Conclusion 20. A Walk into Greatness: June 2017 Visits to Tesio’s Horse Stables and the Sassicaia Winery 20.1. Seven Top Tesio Horses 20.2. The 10 Greatest Racehorses of All Time Acknowledgment 21. Horse Ownership: The Example of Honor Code 21.1. The Setting 21.1.1. Back to Honor Code: The Met Mile on the Belmont Stakes undercard, June 6, 2015 21.1.2. Honor Code at the Whitney, August 8, 2015 21.1.3. Honor Code at the Kelso Grade II Mile Race, October 3, 2015 21.1.4. The Breeders Cup Classic, October 31, 2015 21.2. After the Classic 22. The Pick 6 and the Rainbow Pick 6 22.1. Theory of Pricing the Bets 22.1.1. 2001 Breeders’ cup insurance bets for SCA 22.1.1.1. Insuring the Pick 6 22.2. The One That Got Away: The Hitable $2 Million Pick 6 at the 2009 Breeders Cup 22.3. A Big P6 Win 22.4. The Rainbow Pick 6 at Gulfstream Park 22.5. The Pick 6 at the Del Mar Futurity 22.6. A Pick 6 at Del Mar Acknowledgment 23. A Pick 6 Tale 23.1. Background 23.1.1. Example 1: An interesting Pick 6 23.2. Remarks on How to Win at Racing 23.3. Postscript 23.3.1. Comments 23.3.2. Conclusion Appendix 24. The Triple Crown and Major US Three Year Old Races, 2019 24.1. The 2019 Florida Derby 24.2. The Kentucky Derby 24.3. The Preakness 24.4. The Belmont Stakes 24.5. Haskell Invitational, July 17, 2019 24.6. The Run Happy Travers, August 24, 2019 24.7. The Pennsylvania Derby at PARX racetrack, September 21, 2019 24.8. The Jockey Club Gold Cup at Belmont Racetrack, September 28, 2019 24.9. The Breeders’ Cup, Santa Anita, November 2, 2019 24.10. A Forecast of the Top Horses of 2019 by Gehrke (2020) Acknowledgments 25. The Pegasus World Cup III: Accelerate vs. City of Light 26. The Big Money Older Horse Races: Pegasus, Saudi Cup and Dubai World Cup in 2020 26.1. The Pegasus 26.2. The Saudi Cup 26.3. The Dubai World Cup 26.3.1. Racing is in trouble Appendix: Past Performances for the Saudi Cup Acknowledgment 27. The COVID-19 Triple Crown, 2020 27.1. The Belmont Stakes 27.2. The 146th Kentucky Derby 27.3. Preakness 27.4. Breeders’ Cup Acknowledgment 28. Dr Z’s Place & Show Racetrack Betting System at the First Breeders’ Cup 28.1. Transactions Costs 28.2. Three to Beat the Breeders’ Cup by Bruce C. Fauman 28.2.1. First Race: The Juvenile, one mile, for two year old colts and geldings, purse $1,00,000 28.2.2. Second Race: The Juvenile Fillies, one mile, for two year old fillies, purse $1,000,000 28.2.3. Third Race: The Sprint, six furlongs, for three year olds and up, purse $1,000,000 28.2.4. Fourth Race: The Mile, one mile on the grass, for three year olds and up, purse $1,000,000 28.2.5. Fifth Race: The Distaff, 1 1/4 miles, for fillies and mares, three year olds and up, purse $1,000,000 28.2.6. Sixth Race: The Turf, 1 1/2 miles on the grass, for three year olds and up, purse 1,000,000 28.2.7. Seventh Race: The Classic, 1 1/4 miles, for three year olds and up, purse $3,000,000 28.2.8. Eighth Race: Fleet Nasrullah Stakes, six furlongs, for two year olds, purse $60,000 added 28.2.9. Ninth Race: Seabiscuit Claiming Stakes, 1 1/2 miles on the grass, for three year olds and up, purse $100,000 added 28.3. Postscript by Ziemba 28.4. Off Track and Other Track Betting 28.4.1. Teletrack in New Haven, Connecticut Bibliography Author Index Subject Index