Wednesday, March 29, 2006

Numbers lie, part two.

"The cold, hard numbers have been tabulated. And in what surely is the game's least surprising development since Angels owner Arte Moreno was applauded by fans for lowering concession beer prices, home runs are running for cover now that steroid testing is a regular part of the game.

Last season, the first with a relatively tough steroids policy in place, major leaguers swatted fewer homers than in any summer since 1997.


The decline was not dramatic. And a slight change in one season still could turn out to be as much aberration as trend."

Not to get all MENSA on Scott Miller, but anything that occurs in just one season can not be a trend.


On to the numbers and a most simplistic statistical analysis:

Measurements: 2.05, 2.08, 2.28, 2.34, 2.25, 2.09, 2.14, 2.25, 2.06.

Number of measurements: 9.

High: 2.34.

Low: 2.05.

Average: 2.17.

Standard deviation: 0.11.


Conveniently enough, the 2005 homerun measurement is exactly one standard deviation off the average. One standard deviation is not a lot.

Or just look at the numbers as they move from left to right, from 1997 to 2005. Do you see a trend? Can you guess which year steroids ended?

I'd say steroids ended after the 2001 season when the homeruns per game dropped from 2.25 to 2.09. (It's a trend! The numbers are in and the steroid era ended in 2002!)


I am not blind to the effects of steroids, though I'm not exactly sure why Palmeiro and Bonds are pariahs while Ivan Rodriguez and Clemens get off scot-free.

I also know that homeruns are hardly running for cover. The modern game has not fundamentally changed because homeruns dropped to "only" 2.06 per game.

No comments: