Isiah Thomas wasn't really such a bad NBA executive. Phil Jackson raises the game of even superstar-laden teams. And kicking that field goal on fourth-and-goal from the three-yard line? Bad move.
These are just a few of the widely held sports beliefs economists David Berri and Martin Schmidt attack in their new book Stumbling on Wins, an analytical look at performance and decision-making in sports that is bound to provoke barstool arguments across America.
More important: A breakdown of wins and losses against individual stats shows that for all the access to data that's available to team executives, many continue to make the wrong decisions. Example: NBA players are disproportionately paid to score. Because so much of basketball success is tied to gaining (and maintaining) possession of the ball, players that rebound and commit few turnovers should be valued higher, they argue.
That's the basis for Berri's defending Isiah Thomas (sort of), who had a disastrous record running the New York Knicks from 2003 to 2008. The team was among the league's bottom-feeders despite fielding a high payroll. But because the market for the scorers Thomas acquired was already overvalued, his teams weren't "overpaid" relative to the industry. Blame Thomas for acquiring too many veteran scorers, perhaps, but most sports executives in New York are under pressure to deliver immediate results rather than rebuild. So they roll the dice.
"Any GM with the Knicks' budget would have done the same thing," says Berri.
Both economics professors, Schmidt at William & Mary and Berri at Southern Utah University, the two previously co-authored a book called The Wages of Wins, a look at how payroll and other statistics affect competitive balance and winning in sports. But it's their new research that's likely to attract attention. According to their findings:
• Hockey goalies are overpaid. Their goals against averages are determined as much by the quality of their teams as anything else. Most all stop about 90 percent of the shots they face. Put any goaltender on the New Jersey Devils and his numbers would likely resemble those of the renowned Martin Brodeur.
• LeBron James was not basketball's most valuable player last year. Adjusted for position, a breakdown of the basketball stats that go into winning shows that New Orleans guard Chris Paul narrowly beat out Cleveland forward LeBron James for most extra wins produced for his team in 2008-09 (the last season for which complete data is analyzed).
• Phil Jackson may actually be a genius. NBA players improve their stats when they play for Jackson, an indication that he's more than the beneficiary of having coached Michael Jordan and Kobe Bryant – he knows what he's doing.
• NFL coaches are still too conservative. Kicking an easy field goal from the two- or three-yard-line carries less expected value than going for the touchdown. Even if you go for it and miss, you've got your opponent backed up against its own end zone. That field position advantage is worth almost three points by itself.
• Final Four stars are overrated as NBA prospects. The evidence shows that college stars coming off Final Four runs tend to be drafted too high or too often. The 2005 champion North Carolina Tar Heels had four players taken in the first round, with mixed results for their pro teams. Historically for every star like Carmelo Anthony (Syracuse, 2003), there's at least one flop like Mateen Cleaves (Michigan State, 2000).
There's growing evidence that team decision makers are starting to get with it. The Moneyball era in baseball values on-base percentage more than it used to, while devaluing stolen bases and batting average. Pro basketball teams are employing stat gurus who look past traditional stats like scoring and rebounding while measuring the effectiveness of various lineup combinations. Still, there's a ways to go. Players in the NBA still get paid to score above all else, making optimal value inconsistent with optimal performance. Lineup combination stats still tend to vary from year to year, minimizing their predictive value.
"The NBA doesn't produce a lot of data, so you have to find your own," Dallas Mavericks owner Mark Cuban recently told an industry conference in Boston about finding competitive advantage. He's contemplating where there might be value, such as tracking how many passes a player deflects on defense. But he sees the biggest inefficiencies in an area the team can't control: "Referees," he says, noting that the varying styles – some refs are less likely than others to call three second violations, for example – throws some statistical analysis off track.
Other traditional sports stats have limited value too, according to Berri. There are "plus-minus" numbers in hockey (the measure of how many goals a player is on the ice for, both for and against), fumbles in football and earned run average in baseball. All vary so much from one season to the next that they aren't particularly useful for predicting future performance.
"If you see wild swings in the data," Berri says, "it's not helping you make good decisions."
The top five myths:
1. Isiah Thomas, clueless executive: Slideshow
2. A low ERA makes a great pitcher: Slideshow
3. Some NFL backs are "fumblers:" Slideshow
4. Judging a hockey player by plus-minus: Slideshow
5. The NBA player worth the most wins: Slideshow
• See more beliefs