Baseball by the numbers: Even top managers can’t alter the numbers
BY JOHN GROCHOWSKI For Sun-Times Media September 10, 2012 7:54PM
LOS ANGELES, CA - JULY 11: Former MLB manager Tony LaRussa arrives at the 2012 ESPY Awards at Nokia Theatre L.A. Live on July 11, 2012 in Los Angeles, California. (Photo by Frazer Harrison/Getty Images)
Updated: October 12, 2012 6:18AM
Two weeks ago, I wrote about Pythagorean won-lost records and pointed out the Reds and Cardinals had runs scored and allowed totals that in most years would put the Cards atop the NL Central standings. I also pointed to the Orioles in the AL East, with a record that’s 11 games better than their Pythagorean expectation.
That brought several emails asking if the managers could be raising their teams above their runs data or causing them to fall short. One example:
‘‘Is it coincidence that Tony La Russa is no longer manager and the Cardinals are short on Pythagorean? Could Baltimore’s record be a Buck Showalter effect? Maybe those guys have a knack for getting the best out of their teams. It seems to me that the manager could be an ‘X’ factor when a team is doing better or worse than expected.”
That’s easily enough checked. A manager who has a knack for getting his team to outperform its runs scored and allowed data should consistently have records better than the Pythagorean formula suggests. That formula is runs squared divided by (runs squared plus runs allowed squared) equals winning percentage. Most teams will be within a few games of Pythagorean in most seasons.
Showalter has managed 12 seasons before 2012. He was more than three games above or below his Pythagorean record only twice and both were on the negative side. His 1992 Yankees were four games below Pythagorean, finishing 76-86 with runs data that suggested an 80-82 team. And his 2006 Rangers were 80-82, six games worse than their Pythagorean record.
In total, Showalter teams were three games below the record suggested by run data, with seven seasons above Pythagorean, five below and one exactly on the mark. The differences are so narrow — a quarter of a win per season off expectation — that you can say Showalter’s teams have won right on the rate their runs data says they should have.
La Russa managed 31 complete seasons. Of those, 20 were within three games of Pythagorean, and no season was more than seven above or below expectation, reaching plus-7 in 1992 and 2007 and minus-6 in 1997. He had 15 seasons above Pythagorean, 14 below and two right on the formula. In those 31 years, La Russa totaled 23 wins above Pythagorean, or about three-quarters of a win per season.
Again, with a difference of less than a game a season and only one more season on the plus-side than on the minus, La Russa’s teams have won at a rate consisted with their Pythagorean data.
That’s not to downplay the manager’s role. Some are better at getting the most out of their teams, and that shows up in runs scored and allowed. But consistently outwin the Pythagorean data? That’s something no manager has been able to do.