The AP ran an article on the Atlanta Braves penchant for losing games that are decided by one run. This season the Braves are 5-22 in 1-run games and 0-17 on the road. Which means that the Braves are 5-5 in 1-run games at home, which isn't all that bad. Overall, the Braves are 43-50. The Braves Pythagorean record is 49-44.

Chipper Jones theory on why the Braves are losing the close games:

Jones says the Braves would win more one-run games if they produced more one-run innings.

“You know, to win one-run games, pitchers have to get bunts down,” Jones said. “You’ve got to hit behind runners. You’ve got to put the ball in play with a guy on third and less than two outs. You have to do the little things to win those one-run games, because a run here and a run there is the difference.

“We don’t do it consistently.”

Hmm... interesting, but that's probably not the case. At this point, I'd come up with some sabermetric study showing how a team's record in 1-run games is not much different from its overall record. But, instead like Woody Allen pulling Marshall McLuhan out of nowhere in "Annie Hall," I just asked a noted sabermetrician (or as Rob Neyer called him "one of the top sabermetricians you probably never have heard of") Phil Birnbaum to explain why 1-run losses aren't all that much different than any other loss.

Suppose that team A is 3 runs a game better than team B (which is extreme, but what the heck). Normally they win 6-3 or something.

For team A to win by only one run (6-5), team B has to be two runs better than usual (or team A two runs worse, or some combination. But never mind for now). For team B to win by one run (6-7), it has to be four runs better than usual.

Those two numbers are fairly close -- two runs versus four.

But now, consider five-run blowouts.

For team A to win by five runs, team B has to be two runs worse than normal. (8-3)
For team B to win by five runs, team B has to be EIGHT RUNS BETTER than
normal. (6-11)

These two numbers are far apart -- two runs versus EIGHT.

Therefore, when there's a one run game, the chance that B wins it is pretty
close to the chance A wins it. But when there's a blowout, the chance that
B wins it is far from the chance A wins it. Therefore, B has a huge
advantage in blowouts.

Since the "typical" game is partway between a one-run game and a blowout, one-run games should be closer to .500 than typical, and blowouts should be farther from .500 than typical.

In games decided by 5 or more runs this season, the Braves are 16-12. And the Braves have outscored their opponents by a margin of 405-379.