The Boston Red Sox sit atop the American League East, with a half-game edge on the Tampa Bay Rays. The New York Yankees, the expected division leaders, are four-and-a-half games behind, in third place. That’s surprising to almost everyone outside of Beantown. And it might not be sustainable. Based on run differential, Boston has been the second-luckiest team in baseball thus far, and should be concerned about fortune changing directions.

Let’s start with Boston’s record. The Red Sox are 43-29 with a run differential of plus-34. We would expect a team with that run differential to be 39-33 (Boston’s Pythagorean win-loss record), meaning the Red Sox have four more wins than expected, and in reality should be about level with the Yankees.

If we take it a step further and adjust for the sequence in which their batting events occurred — also known as third-order wins — we would expect the Red Sox to be 36-36, a .500 club. Why take sequencing into account? If a team gets four walks in a row during an inning, a run will almost certainly score. If those same walks are spread out over four innings, it is likely they wouldn’t produce a run. It’s the same number of events, but the sequence in which they occur has a big impact on the scoreboard.

Lucky teams have a checkered postseason record, both with advancing to the playoffs and with succeeding once there. From 2010 to 2019, the last 10 full MLB seasons, almost half of the teams that had five or more wins more than expected (16 of 37) failed to qualify for the postseason. A dozen more lost their first series of the playoffs, whether a one-game wild-card showdown or during the divisional round. Three went on to win the World Series.

So what’s causing Boston to win more games than expected?

The Red Sox have gone 12-6 in one-run games, giving them the third-highest win rate in these situations. Good teams win games by larger margins, so when you see a squad padding its win total in close games, that’s usually a sign of concern. From 2016 to 2019, the last four full seasons, teams annually finishing in the top three in win percentage in one-run games have either missed the playoffs entirely (five out of 12) or lost in the first round they played (five out of 12). The two exceptions are the 2017 Chicago Cubs, who lost in the NL championship series, and the 2018 Red Sox, who won the World Series.

Before Red Sox fans start crowing about how their team can defy the odds again, consider this: That team’s lineup was batting in line with expectations. The Red Sox batted .268 with a .453 slugging percentage that year, and based on the launch angle and exit velocity of each ball put in play we would have expected a .266 average and .452 slugging percentage, almost exactly the actual results. This year, however, Boston is batting .258 with a .441 slugging percentage, both higher than we’d expect based on launch angle and exit velocity (.240 and .430).

One of their biggest bats, J.D. Martinez, is among the most prominent overachievers. Martinez is batting .309 with a .552 slugging percentage, yet based on the launch angle and exit velocity of each ball put in play we would expect him to be batting .266 with a .493 slugging percentage. His batting profile is suspect, too. His line drive rate is down from last year (from 34 percent to 28 percent), his pop-up rate has doubled (from three to six percent) and his rate of hits on the sweet spot of the bat, also known as barrels, has been flat for three seasons.

According to MLB Statcast data, Martinez’s batting profile is similar to that of Paul Goldschmidt, who is batting .255 with a .419 slugging percentage. Five other Boston hitters with enough at-bats to qualify for the batting title, including Xander Bogaerts, are also exceeding expectations in average, slugging or both.

These are slight differences to be sure, but this early in the season those advantages add up and it’s been enough to help put the Red Sox into a surprising division lead. It has also pushed their chances of winning the AL East from nine percent before the season to 44 percent as of Thursday morning, per FanGraphs. If regression starts to creep in, that could yet mean a tough road to the playoffs for Boston.