Teams’ records in one-run games little indication of how good they are
BY JOHN GROCHOWSKI For Sun-Times Media June 11, 2012 7:56PM
Updated: July 13, 2012 6:20AM
Sixty games into the season, you pretty much can pick a
statistical category and find the Cubs are bad at it.
Among National League teams, only the Padres and Pirates have scored fewer runs and only the Mets, Brewers, Astros and Rockies have allowed more. The Cubs have come by their run shortage honestly, being weak at both getting on base and getting runners around the bases. Their .304 on-base percentage ranks 14th in the league and their .385 slugging percentage 12th.
The Pythagorean win-loss percentage (explained in the glossary below) sees the Cubs as a 24-36 team, based on runs scored and allowed. They’re actually four games worse at 20-40.
A big reason they’re worse is that they have been awful in one-run games. They have lost 12 one-run games in a row and are 6-16 for the season. That’s a .273 winning percentage, compared with .368 in games decided by more than one run.
Conventional wisdom tells us one-run decisions separate the good teams from the also-rans. But in sabermetrics, conventional wisdom often melts away under close examination.
Bill James did a piece of research in the 1980s that found the difference between good, so-so and poor teams is smaller in one-run games than in blowouts. When it’s close, it’s anybody’s game until the end. Good teams are more capable of putting the opposition away by bigger margins.
The Cubs being so weak in one-run decisions so far this season led me to look at the best and worst teams of the last five seasons.
From 2007 to 2011, the 20 teams that made the NL playoffs had a .565 winning percentage overall. In an average of 46 one-run decisions per team, that dropped to .543. In the American League, the difference was even larger. The 20 AL playoff teams had a .585 winning percentage overall, but it was only .544 in just less than 46 one-run games per team.
What about the poor teams? A look at the last-place finishers in each division from 2007 to 2011 found the 15 NL cellar-dwellers had an overall winning percentage of .404, but it was .434 in one-run decisions. In the AL, the percentages were .416 overall and .462 in one-run decisions.
A bad record in one-run games isn’t a reflection of good teams stepping up and poor ones stumbling. Rather, close games tend to narrow the differences between the good and the not-so-good.
GLOSSARY: The formula for Pythagorean win-loss percentage is runs scored squared/(runs scored squared + runs allowed squared). Most teams will come within few games of Pythagorean win-loss percentage for a full season.