Sunday, 12 June 2011

Scorecasting by L. Jon Wertheim and Tobias J. Moskowitz

From The Week of April 17, 2011

Any rational fan of professional sport understands that mythology plays a significant role in their love for their favorite pastime. Home-court/field/ice advantage, the hot hand, the hitting streak, momentum... We have welded these and dozens of other axioms onto the basic narrative of sport because, well, the human brain is a powerful machine designed to find patterns in the chaos. In other words, our beloved teams don't just lose because a few, random bounces went against them in a contest between two teams of roughly equivalent skill. No, "the refs were against us," or "that Smith couldn't miss," or "their best players were their best players." Meaning out of chaos... Scorecasting is a systematic attempt by Mr. Wertheim and Mr. Moskowitz to demolish the most widespread of these sporting conceits, subjecting them to the rigor of statistical analysis. And though they largely succeed in this mission, I'm left to wonder if they haven't missed the point of sport.

Scorecasting assembles a collection of statistical oddities in sport and explains them in one of two ways: you're not seeing what you think you're seeing, or you're not properly interpreting what you're seeing. An example of the former is streaky hitting in baseball. The authors conclude that these fabled runs of excellence are essentially meaningless. If a baseball player hits for an average of .300 in a given season, that doesn't mean he will hit successfully three out of every ten times. He may get ten hits in a row before making outs in his next 15 at-bats. This does not make him streaky; a coin flipped ten times may come up heads ten times, but flipped a thousand times, it will come up heads roughly half the time. The sample size is everything... An example of the latter is home-arena advantage. In an era of standardized stadia, it shouldn't make any difference to the players where they are contesting the match, but yet home teams generally win more often than they lose. Why? Because referees tend to favor the home teams over the visiting ones, calling penalties against road teams they don't call against home teams. The authors again conclude that the fans themselves play almost no role. From punting in football to hot-off-the-bench shooting in basketball, most of the popular sports suffer the wrath of precisely this analysis and, in virtually every case, the narrative loses out to cold, hard truths.

Scorecasting will be a handy weapon in the arsenal of stat geeks, a club with which to pummel other fans obsessed by the narrative and mythology of sport. Yet it completely misses the point of why people love sport. It is not to tease out every variable, dispel every illusion, capture every nuance. It is to experience the joy of victory and the despair of defeat. It is to ride the glorious highs of winning streaks and championships and suffer through the doldrums of winless months and trophy-less decades. Passion is what creates the bond between a person and a team. It's a bond built on irrationality, not logic. If we subjected every aspect of sport to logic, no one would pay $100 for a ticket to see grown men act out a child's game. A fan would not buy a jersey for a team whose players don't even know he exists. But people do buy tickets and they do buy jerseys because they want to be part of the narrative. And there's no room for that in statistics.

This is a well-researched and enlightening piece. For some, it will surely strike a cord. But for me, it only succeeded in taking away the last vestiges of magic leftover from childhood. Looking for good, statistical arguments, a must-read. Wanting to believe in the magic of sport? Steer clear. (3/5 Stars)

No comments:

Post a Comment