It’s been a tough couple weeks for the Braves. After jumping out to the majors’ best record, thanks to excellent pitching and Justin Upton’s 12 home runs, the team lost three games in a row to the Pirates, saw a rare blown save from Craig Kimbrel against the Rockies, and got swept in embarrassing fashion by the Tigers. I’ve detailed before in this space how the Braves pitching, although talented, has gotten pretty lucky so far this season. Well, in the last week and a half, that luck has run out.

Down With the Pitching Stats

A quick note on the difference between regression to the mean and gambler’s fallacy: The gambler’s fallacy is expecting a coin to come up heads after a long run of tails. The chances the next coin flip will come up heads are 50/50, same as always. The coin does not “owe” anyone a run of heads to “even out” the tails flips that preceded it. In baseball, the gambler’s fallacy would look like this: expecting a batter with a .330 BABIP to suddenly tank down to a .270 BABIP to “even out” his good luck. (For the sake of that example, assume the hitter had a career .300 BABIP.)

Regression to the mean, rather, is simply returning from a high or a low point towards a career average. In the example of a hitter with good luck on balls in play, it is more appropriate to expect that he will post about a .300 BABIP for the rest of the season. Obviously there is a lot of variance in both directions; players are not machines, some make adjustments, some have breakout years, some deal with off-the-field stuff that outside observers have no chance of knowing. That’s why it’s important not to be a slave to the numbers. Regression to the mean gives you an idea of what is likely to occur; it is not ironclad.

Back to Braves pitching. In the last week, the staff has gone from excellent and fairly lucky to pretty bad and unlucky. After the first two weeks of the season, the Braves pitchers allowed a .235 BABIP which led to a 1.83 ERA. In the last seven days, the staff has surrendered a .345 BABIP and a 5.74 ERA.

To illustrate the significance of luck in ERA, look at the Braves FIP. Fielding Independent Pitching factors out BABIP to eliminate the luck variable and attempt to paint a clearer picture of pitchers’ actual talent. In that first stretch, where the Braves posted a 1.83 ERA, they had a 3.25 FIP. In the last week, with that ghastly 5.74, the team’s FIP was 4.53. A jump of a full run is not insignificant, to be sure, but the difference in FIPs is a whole lot smaller than the difference in ERAs.

Think of FIP as ERA’s more stoic cousin. When you get super excited over a great ERA, or panicked over a terrible one, look at FIP to get a sense of how the team is actually pitching. On the year, the Braves ERA is 3.15, second in the league be .01 and their FIP is 3.51, fifth in the league. The difference indicates the Braves have gotten slightly lucky so far, but not outrageously so. (The team’s still incredibly high strand rate probably accounts for a lot of that luck.)

Going forward, don’t fall for the gambler’s fallacy. Expect the Braves to put up numbers around 3.50, for both ERA and FIP, and treat any exceptionally hot or cold stretches as outliers.