The Orioles' September magic

Rookie Manny Machado has been a big part of Baltimore's September push. Greg Fiume/Getty Images

Baltimore's victory over the Boston Red Sox on Friday carried with it a significance that likely went unnoticed except by the most dedicated and exacting of baseball nerds: In defeating the Red Sox 9-1, the Orioles were owners of a positive run differential for the first time since June 24, when they beat the Washington Nationals 2-1 at home. For the three-plus months and 84 games in between, however, the O's were a team that, despite winning games, couldn't pile up more runs than their opponents.

The practical implications of Baltimore's negative run differential were minimal, of course. Despite a midseason lull that found the Orioles just two wins above .500 and 10 games behind the New York Yankees in the middle of July, the Orioles did more winning than losing for most of the summer. After that June 24 win against Washington, Baltimore stood at 41-31 (a .569 winning percentage) and 2.5 games out of first place in the American League East.

Going into today's action, the Orioles' record is 91-67 (.576) and they're tied for first place in the East. Their odds of making the postseason are better than 96 percent.

Trailing the Orioles for much of the summer, however, were constant questions regarding the sustainability of their performance. Since Bill James' work on Pythagorean (or expected) record, it's been pretty well established that the correlation between run scoring/prevention and winning games is extraordinarily strong -- and that a team's ability to score runs while preventing them is more predictive of a team's future record than present wins and losses.

Instead of excellent run scoring and prevention, much of the Orioles' success was due to their almost impossible fortune in one-run games. While an excellent bullpen and/or particularly skilled manager can provide a slight edge in such instances, it's overwhelmingly the case that a team's record in one-run games will gravitate toward the .500 mark.

Baltimore, on the other hand, has posted a 28-9 record in one-run games this season, giving it roughly 10 more wins than one might reasonably expect by that measure. While the odds suggest Baltimore would ultimately fade away, something different has happened this September: Rather than seeing their record regress back to what we'd expect given the Orioles' poor seasonal run differential, what's happened is that their underlying performance has improved dramatically.