Image courtesy of Wikipedia
Tiger Woods is back. After months of revelations and speculations about his personal life, Woods returned to professional golf at the 2010 Masters Tournament in early April. Although he didn’t win or meet his own expectations, Woods ended up tied for fourth. A week earlier in the latest Distinguished Lecture at the MAA's Carriage House, statistical scientist Scott Berry had posed the question “Is Tiger Woods a Winner?”
Berry’s focus was not on the golfer's personal life, but rather on whether Woods’ winning record was a direct consequence of his superior golfing skill or included an additional component—the extra boost provided by a player who, when near the lead, does whatever it takes to win.
“It's arguable, but I think it's true: Tiger is the best golfer of all time,” Berry said. “He dominates golf to a level that I would have said was impossible years ago." A renowned sports statistician, Berry has written dozens of articles on such topics as “Does Defense Win Championships?” and “The Cold-Foot Effect.” His popular "A Statistician Reads the Sports Page" columns for Chance magazine were a regular feature from 1999 to 2006.
Woods “matches guys like Babe Ruth and Wayne Gretzky in that they just did freak statistical things,” Berry noted. “Tiger’s stats are really that good."
It's common practice in the world of sports for fans and commentators to offer reasons for phenomena that are actually random occurrences, Berry said. They invoke character traits such as "heart," "chemistry," and "drive" and apply terms like "clutch hitter," “hot hand,” or "streak” to account for surprising strings of successes. "What the person is really saying is that they can't explain the randomness, so [here’s] a word that explains it," Berry said.
Berry developed a mathematical model to determine whether Woods' winning record is a consequence of an ability to win more often than his natural skill dictates that he should. For his model, Berry used the round-by-round scores for each tournament on the PGA tour from 1997 to 2004, which included the results from Tiger Woods and 2,003 other players participating in 352 tournaments—a total of 147,154 rounds and 10,509,253 shots.
"The golf score of any particular round is normally distributed with a mean that is the ability of the player," Berry said. "There's also an additive effect due to the particular round."
Berry’s computer simulations strongly suggested that Woods does not have a special ability to win beyond his skill level. His careers wins to date were very consistent with his skill level, compared to the abilities of the other players on the PGA tour.
Assuming that Woods plays to the same level for the next 10 years, Berry’s model predicted that he is almost certain to win at least 19 major tournaments, breaking the all-time record. Indeed, the data suggest that the record is most likely to fall at the 2013 U.S. Open, Berry said.
"Of course, whatever happens going forward—if [Woods] does better, if he does worse, or if he does the same—[commentators] will say it's because of his latest marital mishaps,” Berry said. “It will be the explanation regardless of the data."
Listen to the full lecture (mp3)
Paper: “Is Tiger Woods a Winner?” (pdf)
This MAA Distinguished Lecture was funded by the National Security Agency.