Archive for May, 2008

That other sports statistic

Tuesday, May 27th, 2008

Over the years, I’ve wondered why some sports statistics hold a greater value than others.

In baseball, isn’t “on base percentage” more meaningful than “batting average?” Why did baseball get rid of the “Game Winning RBI,” but hockey still values the “Game Winning Goal?” Who really puts any value on football’s QB rating unless it supports their claim that a particular player is doing well or poorly (or that the statistic is meaningless)?

I am going to use the remainder of this entry to suggest a new or underutilized statistical measure for the big three of professional sports (football, baseball, and basketball).

I believe a very meaningful statistic to measure offensive prowess would be “Percentage of Innings/Possessions Converted into Points.” (A similar statistic could be created to measure a team’s defensive performance.)

In football, “red zone posessions” are analyzed in this way, but the offensive performance of the team in general is measured by “time of posession” and “third down conversion percentage.

I’m puzzled by basketball’s fascination with shooting percentage. It is a statistic inherently distorted by the fact that a shot attempt in which a foul was committed is only included in the shooting percentage if the shot was successful. The statistic is further flawed in that three-point and two-point field goal attempts are both counted when computing the percentage (unless the shooter was fouled and then missed the shot). A different set of complications come into play when “field goal percentage allowed” is used to assess a defense: the statistic is misleading unless rebounding and turnovers are simultaneously studied. A solid defensive team may have allowed their opponent to convert 48% of their shot attempts but created more than 20 turnovers and defended so tightly that every shot was a two-point attempt. Meanwhile, a team that held their opponent to a 39% field goal percentage could have been very ineffective if they allowed the opponent to collect the rebounds from most of those misses or if most of those shots were from beyond the three-point arc.

I see an overall “percent of possessions resulting in points” being a more meaningful and direct way of measuring offensive and defensive performance. If a team missed their first shot but gathered the rebound and scores on the second chance – that’s “1 for 1″ not “1 for 2.” Similarly, if a team lost the ball on a turnover, that is a missed opportunity – “0 for 1″ -whether they were able to take a shot or not.

As mentioned above, a similar stat is already used for special situations in football and in baseball to measure the percentage of plate appearances in which the batter reached base (“on base percentage”). Why not measure the offensive productivity of teams by studying the percentage of innings in which they score runs?

Just a thought…

Contributed by: Scott Copperman (Guest Author)