Skip to main content
ABC News
The Marlins Are Learning That It’s Better To Be Lucky Than Good

Last year, the Miami Marlins were the biggest surprise in baseball, overcoming modest preseason expectations — and an early season COVID-19 outbreak — to not only make the playoffs, but also win a postseason series. Along the way, they needed breakout performances from pitchers (like Pablo López and Sixto Sánchez) and hitters (like Miguel Rojas and Brian Anderson) alike, plenty of perseverance and, yes, an abundant serving of good luck. If not for MLB’s last-second decision to expand the 2020 postseason field from 10 to 16 teams, Miami’s 31-29 record would have left it out of the playoffs on a tiebreaker.1 Furthermore, the Marlins won a lot more than their negative-41 run differential would suggest they “should” have (more on that later). Those fortuitous factors combined to make Miami’s season a success, despite it becoming the first postseason team in MLB history that was outscored by at least a half-run per game during the regular season.

Because of this, though, it was also easy to predict that the Marlins would regress in 2021 if fortune no longer smiled upon them quite as brightly. And sure enough, Miami’s winning percentage has dropped by 64 points, leaving them under .500 (and seven games out of the National League’s final wild-card slot) as we approach the 60-game mark of this year’s schedule. But the drop-off in record has obscured the fact that these Marlins are actually playing much better than they did last season. Instead of putting up MLB’s seventh-worst runs-per-game differential, they now rank 11th-best according to that metric — and that might be underselling how good Miami has been. The main reason for the team’s decline is simply that its remarkable luck from 2020 hasn’t just regressed back to average, it has turned around completely to make the Marlins one of the unluckiest teams in baseball so far this year.

Looking at Miami’s runs scored (263) and allowed (304) in 2020 through the lens of the Pythagorean expectation, we would have predicted the Marlins to have a .434 winning percentage — 83 points lower than their actual mark of .517. Among MLB teams in completed seasons since 1901,2 only the 1905 Detroit Tigers (95 points) and 1981 Cincinnati Reds (87) had a bigger gap between their actual and expected winning percentages. And the gap is even wider (85 points) if we convert wins above replacement3 to an expected winning percentage (.432) and compare it with Miami’s actual record. That difference also landed Miami on the list of biggest gaps ever — so no matter how you look at things, the 2020 Marlins had historic luck on their side:

The 2020 Marlins were among the luckiest in history

MLB teams by biggest difference between actual and expected winning percentages, by run differential or wins above replacement, 1901-2020

🍀 Luckiest by Pythagorean winning percentage
Year Team W L Win% Pythag Win% Diff.
1905 Tigers 79 74 .516 .421 +0.095
1981 Reds 66 42 .611 .524 0.087
2020 Marlins 31 29 .517 .434 0.083
2016 Rangers 95 67 .586 .505 0.081
2008 Angels 100 62 .617 .542 0.075
2004 Yankees 101 61 .623 .548 0.075
1954 Dodgers 92 62 .597 .523 0.074
1984 Mets 90 72 .556 .483 0.073
1972 Mets 83 73 .532 .459 0.073
2017 Padres 71 91 .438 .366 0.072
2018 Mariners 89 73 .549 .478 0.071
1970 Reds 102 60 .630 .559 0.071
1981 Orioles 59 46 .562 .492 0.070
1917 Cardinals 82 70 .539 .469 0.070
1924 Robins 92 62 .597 .527 0.070
🍀 Luckiest by WAR winning percentage
Year Team W L Win% WAR Win% Diff.
1919 Reds 96 44 .686 .579 +0.107
1907 Cubs 107 44 .709 .607 0.102
1906 Cubs 115 36 .762 .663 0.099
2016 Rangers 95 67 .586 .490 0.096
1910 Cubs 104 50 .675 .581 0.094
1909 Pirates 110 42 .724 .631 0.093
1931 Cardinals 101 53 .656 .564 0.092
2007 Diamondbacks 90 72 .556 .466 0.090
2008 Angels 100 62 .617 .529 0.088
1908 Pirates 98 56 .636 .550 0.086
2020 Marlins 31 29 .517 .432 0.085
1949 Yankees 97 57 .630 .546 0.084
1981 Reds 66 42 .611 .527 0.084
1915 Robins 80 72 .526 .443 0.083
1904 Cubs 93 60 .608 .525 0.083

Pythagorean winning percentage is used to turn a team’s ratio of runs scored to runs allowed into an expected winning percentage.

WAR is measured using JEFFBAGWELL (Joint Estimate Featuring FanGraphs and B-R Aggregated to Generate WAR, Equally Leveling Lists), which averages the metrics found at and FanGraphs.

Sources:, FanGraphs

In the seasons that immediately followed, these teams tended to see their winning percentages regress back toward .500 faster than we would expect from an ordinary team with the same record.4 So in one sense, it’s not surprising that Miami would be just the latest victim to join that list. But there’s one big difference: According to both Pythagorean and WAR winning percentages, the other teams above stayed roughly steady in their true quality across seasons — in other words, the only thing driving their differences in winning percentage was changing luck. In Miami’s case, however, they have genuinely improved quite a bit since last year.

Although Sánchez and Anderson have been limited by COVID-19 protocols and injuries this season, Rojas and López have been producing WAR at a better pace this year than last year, along with former All-Star righty Sandy Alcántara. (Rojas was also sidelined recently with a fractured finger, so the Marlins will be anxiously awaiting the return of their best position player.) Miami’s position players have gone from ranking 24th in WAR in 2020 to 20th in 2021 — and from 20th to fifth on defense in particular — with big boosts coming from sensational rookie second baseman Jazz Chisholm Jr. and an improved outfield featuring Starling Marte, Adam Duvall and Corey Dickerson. The pitching staff has made even greater strides, rising from 21st in WAR to second — including from 16th to eighth among starters and from 21st to sixth among relievers. Playing at a 6.1-WAR pace per 162 games, lefty starter Trevor Rogers has been the Marlins’ best player as a rookie in 2021, and MLB’s ninth-best pitcher overall.

With a more productive core in place, the Marlins should actually be improving on last year’s record. But in a strange twist of fate, Miami’s luck has gone from historically good in 2020 to historically bad in 2021. This year, the Marlins’ actual winning percentage is 61 points lower than its Pythagorean expectation and 69 points lower than its WAR-predicted winning percentage. If both numbers hold, they would give Miami a place among the unluckiest seasons in MLB history. And no team would have a more painful case of “luck whiplash” than the Marlins, especially in terms of the change in Pythagorean luck from one year to the next:

No team’s luck has collapsed like Miami’s

MLB teams by biggest negative year-over-year changes in luck, by run differential or wins above replacement, 1901-2021

Pythagorean Luck 🍀
Years Team Year 1 Year 2 Change
2020-21 Marlins +.083 -.061 -.144
1981-82 Reds +.087 -.036 -.123
1998-99 Royals +.051 -.069 -.120
1904-05 Cubs +.037 -.079 -.116
1934-35 Braves +.037 -.079 -.116
1917-18 Cardinals +.070 -.037 -.107
2020-21 Blue Jays +.048 -.058 -.106
1904-05 Browns +.031 -.068 -.099
1952-53 Giants +.041 -.058 -.099
1957-58 Redlegs +.039 -.057 -.096
WAR Luck 🍀
Years Team Year 1 Year 2 Change
1981-82 Reds +.084 -.074 -.158
2020-21 Marlins +.085 -.069 -.154
1934-35 Braves +.055 -.098 -.154
1935-36 Phillies +.032 -.086 -.117
1933-34 Senators +.054 -.062 -.116
1909-10 Red Sox +.053 -.061 -.114
1917-18 White Sox +.058 -.053 -.112
1969-70 Mets +.078 -.033 -.112
1917-18 Cardinals +.060 -.050 -.111
1953-54 Red Sox +.040 -.069 -.109

The difference between a team’s actual and expected winning percentages is considered its “luck.”

Pythagorean winning percentage is used to turn a team’s ratio of runs scored to runs allowed into an expected winning percentage.

Sources:, FanGraphs

(Interestingly, the Toronto Blue Jays of the past two seasons are on the Pythagorean list as well. Somewhat similar to Miami, Toronto is another case of a team that broke out last year and whose middling 2021 record hides a good roster underneath — which they showed the Marlins firsthand in their victory over Miami Tuesday night.)

Despite their complete reversal in fortune, the Marlins are still on the periphery of the playoff chase in the NL East. The injury-ravaged New York Mets are the only above-.500 team in the division, and Miami actually has a better Pythagorean record than any of its division rivals. Although our MLB forecast model gives them only a 4 percent chance to return to the postseason, the Marlins are just close enough to mount a run if they get back to full strength soon. All they’ll need to pull it off is, well, a little luck.

Check out our latest MLB predictions.


  1. Under a normal postseason format, the NL’s playoff teams would have been the Dodgers, Braves and Cubs as the three division winners, plus the two teams with the next-best records as wild cards. The Padres would have taken the top wild card, with the Cardinals, Marlins and Reds tied for the second spot. Since not all three teams played each other, intradivisional record would have been used as the tiebreaker; St. Louis was one game better than both Miami and Cincinnati in its divisional games, so the Cardinals would have made the playoffs as the second wild card.

  2. Excluding the short-lived Federal League in 1914 and 1915.

  3. Using our JEFFBAGWELL metric to blend WAR from and FanGraphs, for which you can download data on GitHub.

  4. The non-Marlins members of both lists above saw their winning percentages decline by 67 points the next year; a separate control group of teams (selected because their winning percentage matched the average winning percentage of the teams on the two lists) regressed by only 41 points on average.

Neil Paine was the acting sports editor at FiveThirtyEight.