IU Research & Creative Activity Magazine

Page 21

Looking for the edge In basketball, the depth of statistics is not as great, so Winston and Sagarin use an adjusted plus/minus system to determine player values. Plus/minus stats in hockey, a fixture of that sport for four decades, are grade-school simple: player on the ice gets a +1 when his team scores a goal, and a -1 when his team allows a goal. But that doesn’t take into account the quality (or lack thereof) of teammates or competition. For basketball, Winston and Sagarin devised an adjusted plus/minus rating system that, for the 2006 –07 season, revealed Kevin Garnett to be the best player in the NBA. He played on a bad Minnesota Timberwolves team that finished 32-50 and failed to make the playoffs, but without him the Timberwolves would have been far worse. His point rating was +19, meaning that over a 48-minute game, he improved his team’s performance by 19 points as compared to an average NBA

going against the norm in running a team. “There were two primary values. First, by the time the playoffs came along and we had played a team four times, Wayne’s analysis could tell us what lineups worked. The second value was that it could identify player performance trends over a period of years. There were specific markers that showed up in Wayne’s system that would deteriorate over time. That was important to us.” Analysis also helped Cuban find a new coach for the 2008 – 09 season, Rick Carlisle. Previously with the Indiana Pacers and Detroit Pistons, Carlisle turned losing teams around. Bill Parcells was a similar kind of coach in the NFL. “They use guys in the spot where they’re the best and that improves their talent level,” Winston says. “That’s all a coach can do.” The next frontier In Mathletics, Winston also lands mathematically backed jabs at college basketball’s Ratings Percentage Index, or RPI (“you can win a game and go down, or lose a game and go up”), the NFL’s overtime system (“the coin-flip winning team wins 60 percent of the time; it’s only going to change if the Super Bowl is decided in overtime and the other team doesn’t get the

“I was looking for any edge I could get. No other team was using statistical analysis, and Wayne was the best stats guy I knew,” says Mark Cuban, the Dallas Mavericks’ owner. “There were specific markers that showed up in Wayne’s system and that would deteriorate over time. That was important to us.” player. His offensive rating was +7, and his defensive rating was a league-best -12 (meaning he prevented 12 points per 48 minutes, hence the minus). Another measure of Garnett’s greatness that season was revealed in measuring how the team played with him in the lineup and without him. Predictably, without a man who averaged 22.4 points per game and a league-best 12.8 rebounds per game, the team was horrendous. Without Garnett, every Timberwolves player was 14 to 25 points worse. Before the 2007 – 08 season, Garnett was traded to the Boston Celtics and in his first year, the team won an NBA championship, making Garnett well-deserving of the highest paycheck in the league, at more than $20 million per season. Winston and Sagarin worked as consultants for the Dallas Mavericks for nine seasons starting in 2000 – 01, using lineup analysis and player-win shares to determine which players the team should pursue and retain and the best combinations of them to use. In the 10 years before, the Mavericks were perennial losers and never in the playoffs. Since 2000 – 01, the Mavericks have won at least 50 games and qualified for the postseason each year. “I was looking for any edge I could get. No other team was using statistical analysis, and Wayne was the best stats guy I knew,” says Mark Cuban, the Mavericks’ owner known for

ball”), and the bane of college football fans everywhere, the Bowl Championship Series (“it doesn’t count the scores of the games  — that’s stupid”). Ideally, the next realm of sports math would help find better answers to such systems, in turn allowing winners to be settled more fairly on the field. Every weekend of televised sports fuels discussion and launches spreadsheets among Winston’s students at the Kelley School of Business. Projects last semester included analyzing NFL teams using two-point conversions and when basketball teams should foul at the end of games when leading by three points. As computer systems become more advanced, Winston says the day will come when high school coaches can type in their basketball lineups and instantly glean the most winning combinations. In football, a system will analyze plays on a point-per-play basis rather than a yardage basis. Using math to follow the money in sports isn’t for everyone yet, but it is becoming harder to ignore.“Some don’t like it because it gives an unambiguous answer, and people want to have an opinion,” Winston says. “You should be able to look at the game and have your opinion. But the math is one of piece of the equation you should look at.”

Research & Creative Activity | S P R I N G 2 0 1 0

Photo by Ann Schertz, courtesy of the Kelley School of Business at Indiana University

just as important, who isn’t worth what their agent is asking? “I’m always amused when teams pay players totally out of kilter with what our numbers say they should. It’s usually that they’re paying for the sizzle, not the steak,” says Jeff Sagarin. “A lot of players’ public images are better than they really are.”

John Schwarb is a freelance journalist. He lives in Indianapolis.

19


Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.