The Do’s and Don’ts of Sports Metrics: The Tennis ATP Leaderboard
WAR, VORP, QBR, DVOA, PER might look like a cryptic code, but they are actually part of a growing language in sports. The language is statistics and these acronyms are just a sample from an increasingly large population of measures that are now part of the lingua franca of sport.
Often referred to as “metrics,” these numerical measures attempt to tell us something about —often complex—sport performance, like the quality of a batter’s performance or an individual’s contribution to his team. We can already point to examples where the adoption of an advanced statistic has reshaped a sport. Recognition of the offensive value of on-base percentage (OBP), for instance, was a large contributor to what eventually became the Moneyball phenomena in the MLB.
At the same time, the value of other advanced metrics have been heavily criticized. An example of this type is the Rating Percentage Index (RPI), which is used by the NCAA to measure the strength of collegiate basketball teams. Although many have expressed the concern that the RPI lacks statistical justification, it has been used for decades to set the seeding for the March Madness tournament, one of the biggest events in sport.
The highs and lows of metrics in sport show us that not all metrics are equally successful. How, then, does one develop a good metric in sport?
In this article, we present criteria for evaluating the quality of a sports performance metric. To make the use and application of these criteria more concrete, we apply them to the Association of Tennis Professional’s (ATP) Serve Leaders, a recently introduced statistic to assess the serve performance of elite tennis players.