Rejoice, sports fans: Nate Silver is back on our team. After three years as the New York Times’ in-house statistical warlock, Silver has signed a new deal with ESPN, one that will allow him to indulge his interests in noninaugural balls.
Before he launched FiveThirtyEight, the one-time economic consultant developed a system to forecast the performance of baseball players—a tool he designed to help him win his fantasy league. But Silver isn’t just a baseball guy. In 2009, he helped ESPN devise the Soccer Power Index, an alternative to the hilariously inaccurate FIFA world rankings. (Brazil, which demolished Spain to win the Confederations Cup, is currently ranked an implausible No. 9 by FIFA compared to Spain’s No. 1. In the SPI, Brazil is No. 1 and Spain is No. 3.) He’s also done posts for the Times on why it’s better to be a No. 15 seed than a No. 8 in the NCAA Tournament, the flukiness (or lack thereof) of Jeremy Lin, and why Canadian teams don’t win the Stanley Cup.
What will ESPN’s new man do on his first day on the job? Though he’s (probably) not a warlock, Silver will have an army of researchers at his disposal and will be given the freedom to explore whatever sports questions he finds most fascinating. In a podcast with Bill Simmons last November, Silver said that, among other things, he was interested in the “contract-year phenomenon”—whether players do in fact perform better when they’re shooting for a big free-agent deal. That’s interesting and all, but I’ve got some other ideas. Here are eight more projects to get Silver started.
Create college football rankings that actually make sense. After this coming season, the BCS will be destroyed and shot into outer space, where intelligent beings on other planets can gaze upon it and learn from humanity’s mistake. Starting in 2014-15, there will be a four-team college football playoff, with the combatants chosen by a selection committee. Though it doesn’t take a genius to come up with something superior to the intentionally hobbled BCS computers, Silver has the authority and the platform to come up with some smart standings that fans, pundits, and the sport’s new selection committee will take seriously.
Fix the tennis rankings. Next to FIFA’s abomination, the points systems used by the ATP and WTA tours do the worst job of reflecting a sport’s real-life pecking order. This year, Rafael Nadal actually fell in the rankings after repeating as French Open champion. As Howard Bryant explained, the only way for a player to improve his standing is to get a better result—something that’s not possible if you’re repeating as champion of the same tournament year after year. That’s dumb. Should we create different rankings for different surfaces? Should you get bonus points for knocking off the No. 1 player? You tell me, Silver!
Figure out how NBA players move. In arenas across the land, there are ceiling-mounted cameras tracking every athlete’s every move. The mountains of data these SportVU cameras have dredged up are a rich bounty for sports analysts—just take a look at Zach Lowe’s Grantland story on how the Toronto Raptors are now able to determine exactly how their players should rotate on defense. For sports fans and researchers, the problem here is that teams usually treat these findings as if they’re nuclear launch codes. Just a few researchers, including Grantland contributor Kirk Goldsberry, have access to these SportVU goodies. Given ESPN’s huge investment in the NBA, I’m guessing that Silver, too, will be given the launch codes. Since the rest of us are in the dark, we’re going to depend on those in the know to figure out how the NBA ticks and to pass that knowledge on to the rest of us ignoramuses.
Create a new two-point conversion card. Football coaches are way too risk-averse, typically opting for the point after touchdown when a two-pointer is a sound strategic option. Most coaches use a simple card that helps the arithmetically challenged determine that going for two makes sense when you’re trailing by five points. A better card, like the one seen here, shows that going for two is the better probabilistic play in many, many different situations. This is a case where much of the work has already been done. Silver just needs to help create a small, easy-to-understand card, perhaps one featuring explanatory cartoons.
Kill the RPI. This one’s an easy target. Everyone agrees that the Ratings Percentage Index is terrible. It’s a metric that the NCAA uses to help sift college basketball teams for tournament selection, but it’s worse than many other measures at predicting outcomes. As Ken Pomeroy pointed out in Slate, “three-quarters of the RPI is determined by a strength-of-schedule component. That means who you play is often more important than whether you win or lose.” And yet the NCAA, in its infinite nonwisdom, continues to use the RPI as part of its March Madness toolkit. ESPN’s the Basketball Power Index, masterminded by Dean Oliver, is superior to the RPI in every way. As a company man, Silver will be deployed to stump for it and perhaps even improve it, making it clear that the RPI is bad for America and that its defenders must be destroyed.
What baseball players are good at defense? Figure it out! There are a bunch of different systems that purport to measure how well baseball players field their positions: Range Factor, Ultimate Zone Rating, Revised Zone Rating, Defensive Runs Saved, Plus/Minus Runs Saved. These systems don’t often agree with one another. I’d like to know which of these systems will tell me, in the most accurate terms, precisely how bad Derek Jeter is at playing shortstop.
Build a better NBA. This one’s more theoretical than practical. The NBA salary cap is one of the sports world’s most restrictive, forcing teams to trade away emerging stars and making it monstrously difficult to keep a winning roster together. The league’s economic model also encourages tanking, with teams finding it makes long-term sense to purposefully become terrible and load up on high draft picks. There has to be a way to incentivize decent teams to stay decent, to discourage tanking, and to allow teams like the Oklahoma City Thunder to keep their cores intact. Perhaps the solution is to force LeBron James to play three games a year for every NBA team. But I’m guessing Silver can come up with something better.
Use stats to figure out who’s doping (and who’s totally clean). Is it possible to determine if an athlete is cheating simply by looking at his numbers? There’s a fascinating debate going on right now about whether such a thing is possible, and Tour de France champion Chris Froome is at the center of it. A bunch of statistical sleuths believe Froome’s climbing times are suspect, as they’re uncomfortably close to those of a doped-up Lance Armstrong. Baseball pundits, too, are asking whether Orioles slugger Chris Davis’ home-run-heavy first half should raise our suspicions. These accusations, though, could be totally unfair. It’d be worth Silver’s time to do a deep dive on the stats of known dopers across a bunch of sports and determine whether there are statistical techniques that could transform our steroid witch hunts into informed steroid witch hunts. And if there aren’t statistical techniques that can do the job, it’d be great to know that, too—and to have someone pass that message along to Rick Reilly.