This regular season, the Red Sox led the American League with 108 victories and the Brewers topped the National League with 96.
But by Pythagorean projection, which calculates how many victories normally would be expected with runs scored and allowed data, the leaders were their League Championship Series opponents.
The Astros projected to 109 victories instead of their actual 103 and the Dodgers to 102 victories instead of their actual 92.
This is the first time since 1977 and only the second since the leagues split into divisions in 1969 that both LCS featured matchups in which the teams with fewer victories had better Pythagorean records that their opponents.
The Pythagorean formula, which put the Red Sox at 103 victories and the Brewers at 91, is calculated by dividing runs squared by runs squared plus runs allowed squared to get a winning percentage.
The Dodgers scored 804 runs and allowed 610. Plugging those numbers into the formula yields a .63467 winning percentage that would translate to 102 victories in 162 games.
When there’s a sizable difference between actual and Pythagorean victories, a measure of good or bad fortune is involved (Baseball-Reference.com lists the difference in a ‘‘luck’’ column). A lights-out bullpen can move the victory total a bit. So can unusual runs of clutch hitting, for better or for worse.
Such things can lead to extraordinary records in one-run games. The Dodgers (22-22) and Astros (24-24) were .500 teams in one-run games this season, while the Red Sox were 25-14 and the Brewers 33-19.
RELATED
• Chili Davis ‘not going to blame myself,’ wishes next Cubs hitting coach better
• Cubs fire hitting coach Chili Davis year after seeking new voice, style
In 1977, the LCS teams were the Royals (102 actual victories, 98 Pythagorean) and Yankees (100, 99) in the AL and the Phillies (101, 98) and Dodgers (98, 101) in the NL.
The teams were nearly equal in real and Pythagorean records, but the Pythagorean leaders won both series. The Yankees and Dodgers advanced, with the Yankees winning a six-game World Series.
The disconnect between records and projections was larger this season than in 1977. Runs data suggest the Dodgers, with four fewer victories than the Brewers, projected as six victories better than their opponents. The Astros, who were five victories behind the Red Sox, projected as six victories better.
While we haven’t seen such Pythagorean reversal in both LCS in 41 years, there have been a number of individual series that fit the bill. In the 10 seasons ending in 2017, there were four Division Series, four LCS and two World Series in which the team with more Pythagorean victories had fewer actual victories than their opponents. In those series, the Pythagorean leader won six times.
The last time it happened was a 2016 ALDS in which the Rangers (95 victories, 82 Pythagorean) were swept in three games by the Blue Jays (89, 91). In the ALCS the previous season, the Royals (95, 90) topped the Pythagorean-leading Blue Jays (93, 102).
Both sides win some and lose some, but the results tell us it’s a mistake to assume overachieving or underachieving according to run data in the regular season will continue into the postseason.
