Saturday, April 25, 2015

Playing Time

I wrote a few days ago about one idea I had recently about sports statistics. The only other good one I've ever had came to me more than ten years ago; I remember speaking with my son Dominic about this as we were playing darts in the Calgary house we then shared; it must have been around 2004 or 2005. The Vancouver Canucks' Sedin twins sparked the idea. They were both in the top 20 in scoring and I noticed that they were each being given no more than 15 or 16 minutes per game by then-coach Marc Crawford--while most of hockey's other leading scorers were playing 4 or 5 minutes more than that per game. I got out my calculator and figured out that if the Sedins had been playing 20 minutes per game, they would both (assuming the same rate of scoring) be among the league's top handful of players--not merely the top twenty. Why didn't hockey have a statistic that took time into account when measuring productivity?

Now it does, of course; that's been one of the statistics that have been developed in recent years as the sport has become much more sophisticated statistically.

The thought of those numbers and the Sedin twins came back to me this week. For the first few games of the Canucks series against the Calgary Flames the twins were being played for only 15 or 16 minutes. With the Flames were up 3 games to 1 current Canucks' coach Willie Desjardins changed tactics, and played them for roughly 20 minutes each in game 5; it worked, with Daniel scoring and the Canucks winning 2-1. And tonight in game 6? Daniel played for 20:36 minutes, Henrik for 22:18; neither scored and both were minus-2, with the Canucks season ending as the Flames won 7-4. Good as they are, the Sedins aren't the players they were ten years ago. In more than one way, time has to be taken into account.


Saturday, April 18, 2015

Counting All the Bases

During the Marlins-Mets game yesterday the announcer used a term that's familiar to anyone who follows the game--a "productive out." The Mets batter had hit the ball to the right side and been thrown out at first, but the Mets runner on second base had been able to advance to third on the play. Some "productive outs," of course, are taken account of in baseball statistics; if the batter is thrown out on a play that allows a runner on third to come home and score a run, for example, he is credited with an RBI. But in many, many cases no credit is recorded for runners advancing on the base paths as a result of what the batter has done.

This is not only a matter of what happens when the batter records an out; it also concerns bases advanced resulting from hits and walks. If a batter hits a single or walks with no one on base, he is credited with that single or walk. But no statistic measures the impact of a single or a walk that no only puts the batter on base but also advances one or more other runners. If there are already runners on first and second who advance to second and third respectively on a single or a walk, then the total impact of the batter's single or walk has been three bases advanced, rather than one. (The same would be true, of course, if the batter is hit by a pitch.)

Why not create a new baseball statistic to measure net bases advanced? By this measure that "productive out" which advanced a runner from second base to third base would add one to the batter's net bases advanced total. That single or walk which also advanced runners from first and second to second and third would add three to a player's net bases advanced total. A grand slam homer would add 10 to a player's net bases advanced total (4 for the batter, plus 3 for the player who had been on first, 2 for the player who had been on second, and one for the player who had been on third.) Stealing second base would add 1 to a player's net bases advanced total--but being caught trying to steal second would reduce a player's net bases advanced total by 1, since the play wipes out the effect of his having gotten on base in the first place. Striking out or flying out or hitting a ground ball out would not affect the total--but hitting into a double play would subtract 1 from a player's net bases advanced total, since the net effect would be to remove a teammate from the bases.

This is one of only two sports statistics ideas I have ever had; the other one can be the subject of another day's post. So too can be my reason for watching that Mets-Marlins game in the first place--the speedy and graceful and charming Dee Gordon.

Tuesday, April 7, 2015

Making the Poor Poorer: An Index

The government of British Columbia declared recently that BC would do as most other provinces do: index the minimum wage to the Cost of Living index. No longer will the working poor see their earnings decline in real terms year after year; this way they will finally be able to keep up.

Not so fast.

One problem with this reasoning has been pointed out by many commentators: since BC's minimum wage is far below a living wage, indexing it to the cost of living will guarantee it remains far below a living wage.

What has not been widely recognized is that indexing to cost of living will condemn the working poor not just to staying the same distance back of the average worker, but to falling further and further behind. The reason? Annual changes for the average worker reflect not just price inflation, but also economic growth; in most economies most of the time productivity keeps improving, and--over the medium or the long term--wages and salaries increase by considerably more than do prices. In the United States, for example, GDP per person increased by almost 50% between 1995 and 2013. Such gains deserve to be shared by all workers--including those making the minimum wage.

A much fairer way to index the minimum wage, then, would be to use that yardstick: GDP per person. Very rarely, the minimum wage might go down year-over- year using that approach. But most years it would go up--and go up by more than the cost of living. If we used that index, minimum wage workers would share more equitably both in the occasional economic downturn and in the much more frequent phenomenon of modest economic growth. Using prices as a means to set wages, on the other hand, is a recipe for making the poor slowly but steadily poorer and poorer and poorer in relation to the rich and the middle class.