Sports Illustrated writers were at it again with misinformed analysis. In Ted Kethi's Power Rankings, he says that the Dodgers are a dangerous playoff team because they have the best record in one run games of any team in the National League. I have already explained why UZR does not correlate to record in one run games (sorry the formatting is messed up; the article was transferred from my old site) and now it is time to explain why this latest comment makes no sense.
First of all, a team's winning percentage in one run games has very little bearing on a team's total winning percentage. If you run a regression on the two stats, the r-squared is only somewhat significant at .488. In the Dodgers case, if you apply the formula predicted by the regression, you get a result that undershoots there actual winning percentage by .045. As you can see, it is not realistic to project a team's actual winning ability based solely on their record in one run games.
Overall this year, the top ten teams in the league based on the winning percentage this year have a winning percentage of .578. Their winning percentage in one run games is .555. They have actually underperformed by .023, while the Dodgers have underperformed their total winning percentage by .012. If we believe that record in one run games is actually due largely in part to luck, the Dodgers have actually artificially inflated their overall winning percentage more than the other top teams. I would argue that this makes them LESS dangerous in the postseason. However, that does not make a cute little tidbit of information in the Power Rankings like Mr. Miracle Andre Ethier.