Weekend Odds and Ends: Over/Unders, “Short Hops” vs Strategy
Vegas Watch posted MLB win totals in February, and with several projection systems out now, it’s always fun to see where the major discrepancies lie. Last year, several teams projected win totals and their Vegas over/under figures (and remember, over/unders aren’t designed to project actual wins, they’re designed to produce an equal amount of bets for both sides) differed markedly. Matthew Carruth’s piece last year argued that the Vegas line was biased towards the previous year’s results, or more accurately stated, the Vegas line was set to account for the fact that most bettors are heavily influenced by last year’s results.
What’s interesting, at least to me, is how much they agree this year. Last year, six teams had over/under projections that were at least six wins above or below the average of their CHONE and CAIRO projected win totals. This year, well, I can’t use CHONE as its creator was snapped up by an MLB team, and thus the system isn’t public anymore. But CAIRO’s here, and Baseball Prospectus playoff odds report is here. Compare them to the over/unders here, courtesy of Vegas Watch. BP’s odds came out after Chris Carpenter’s injury, but even so, they’re within a win or two. CAIRO and the Vegas lines are surprisingly close; there just aren’t outliers here. The only one that stands out is the Angels, who have an over/under of 85, but 77 wins in both CAIRO/BP’s projections. What would account for this convergence? Are bettors more aware of regression? Do big trades/signings help bettors put aside previous year’s results? If so, were there really that many more impact moves this off-season than last? This would all be more fascinating if the projections for the M’s weren’t so uniformly terrible.
Sabermetrics has always had critics – people who assume that measurement and comparison of events and players must necessarily come at the expense of enjoying them. This belief is never really explained, and the more you think about it, the more bizarre it becomes. Michael Lewis’ Moneyball occasioned a wave of condescending non-sequitors disguised as critiques in part because the book argued that using data was a way to improve actual roster construction/MLB strategy (and thus more than a way to improve fantasy drafts for lonely losers). I’d love to read an intelligent, evidence-based argument arguing that improved scouting would produce a better return on investment, or that differences in player development programs make looking at raw minor league numbers problematic. Will Alan Hirsch give us that argument?
“Rather than celebrating baseball’s delightfully spontaneous quality, sabermetricians deny it or rebel against it.”
Huh. Doesn’t look like it.
I haven’t read the book, so it’s tough to judge how it proves its anti-Moneyball case, but Alan Hirsch spells out his primary problem with what he calls sabermetrics in this article:
‘Rather than celebrating baseball’s delightfully spontaneous quality, sabermetricians deny it or rebel against it…Sadly, this distorted perspective has spread to other sports. The NBA’s Houston Rockets employ Sam Hinkie as “head of basketball analytics.” In an article in the New York Times Magazine lauding Hinkie and the Rockets, Michael Lewis quotes Mr. Analytic as follows: “I care a lot more about what ought to have happened than what actually happens.” This staggering statement captures the impoverished perspective of sabermetrics: Sports as a plaything for social scientists, a laboratory to manipulate probabilities and chart results in order to assist the next simulation. The game itself conveys no beauty or meaning.’
Keep in mind that these quotes come from an employee of a particular team. He was hired, presumably, because he can help isolate estimates of true talent from the chaos of a game or season – he was hired specifically to strip out luck in order to help his team find more talented players. Hirsch’s problem doesn’t seem to be with sabermetrics, advanced statistics or Mr. Hinkie, it’s the entire enterprise of trying to separate talent from luck. He can’t seem to imagine why any team would have an employee who watched the game differently from a fan.
“Every basketball fan understands the frustration when good defense is thwarted by bad luck, but surely such “near-random events” are phenomena to be embraced, not wished away. Hinkie doesn’t see it that way. To him, a wild shot is the sort of thing that happens but ought not, and thus undermines rather than enhances the sport.”
Remember: Hinkie is an employee of a team, a team which just lost thanks to a wild shot from a bad player. Hirsch damns Hinkie for not giggling in child-like wonderment when the team that employs him loses on a lucky play. The last quote gives some insight into Hirsch’s problem with Moneyball: it was about strategy. He doesn’t seem to be promulgating a different, better strategy, he appears to be arguing that strategy itself is the enemy of the anarchic beauty of sport. The job of a general manager is to assemble the best possible team, not because doing so will eliminate luck from the equation, but because talent is something the GM can control. None of this, and I seriously can’t believe the idea didn’t occur to Hirsch, means fans and analysts hate luck – it just means that a team is better off identifying and acquiring the most talent they can given fiscal constraints than it is to ask fans to renew their season tickets for the chance to see a whole bunch of random occurrences.