Think about it. Line drives hit right at fielders, mis-hit balls dying in the infield. Fly balls barely caught and barely missed. Balls called strikes and strikes called balls. Even the best batters make twice as many outs as hits. Even the best teams lose more than a third of their games.
This season, the Houston Astros had the highest win percentage (66.0%) in baseball, yet they lost two out of six games to Baltimore, which only won a third of their games—not because Baltimore was the better team, but because Baltimore was the luckier team in those two games.
The Astros are one of the 10 best teams this season (along with the Yankees, Tampa Bay, Minnesota, Cleveland, Oakland, Atlanta, Washington, St. Louis, and the Dodgers), but who would win a 7-game series between any two of these teams? Your guess is as good as mine—perhaps better—but it is still only a guess.
Remember the 1990 Oakland A’s, with league MVP Rickey Henderson, Cy Young winner Bob Welch, and Cy Young runner-up Dave Stewart? Their reliever Dennis Eckersley had an 0.61 ERA, with 73 strikeouts and 5 walks (one intentional) in 73 innings. Oh, they also had Mark McGwire, Jose Canseco, Carney Lansford, and a half dozen other stars.
Playing in the American League west, the toughest division at the time, they won 103 games during the regular season, the third year in a row that they led the league. On the eve of the World Series, award-winning sports journalist, Thomas Boswell wrote in the Washington Post, “Let’s make this short and sweet. The baseball season is over. Nobody’s going to beat the Oakland A’s.”
The A’s lost the World Series in four straight games to the Cincinnati Reds, who had only won 91 games during the regular season. Chicago writer Mike Royko said that it happened because the A’s had three ex-Cubs on their roster. No, it happened because pretty much anything can happen in a 7-game series between two good teams.
How about the 1969 New York Mets? They entered the league in 1962 and lost a record 120 games. Over their first six years, they averaged 54 wins and 108 losses, a .333 winning percentage. They improved to 73 wins and 89 losses in 1968 and then got a flukey 100 wins in 1969. In the World Series, they faced the Baltimore Orioles who had won 109 regular season games with All-Stars everywhere, including Frank Robinson, Brooks Robinson, and Boog Powell. They had pitchers Mike Cuellar (23-11 with a 2.38 ERA), Jim Palmer (16-4 with a 2.34 ERA), and Dave McNally (20-7 with a 3.22 ERA). Relievers Eddie Watt, Pete Richert, and Dick Hall had ERAs of 1.65, 2.20, and 1.92.
Behind Cueller, the Orioles breezed through the first game of the World Series just as expected, winning 4-1. Then they lost the next four games. For the next 14 years, the Mets were a distinctly mediocre team, winning 46% of their games; but for one amazing season, they were the Miracle Mets.
What are the chances that the better team, by all objective measures, will lose the World Series? Surprisingly high. In a game like baseball, where luck is so important, a brief 7-game series tells us very little.
A team’s win probability varies from game to game along with the starting pitcher and other factors, but we can get a pretty good estimate of the overall probability that a team will win a 7-game series by simply assuming a constant win probability for each team. Suppose that the Astros have a 60% chance of winning any single game against the Nationals and the Nationals have a 40% win probability. These probabilities are generous for the Astros because they only won 66.0% of their games during the regular season, playing against average teams, and the Nationals won 57.4% of their games. But we will use 60% to give the Astros the benefit of the doubt.
Even with this generous assessment of the Astros, it turns out that there is a 30% chance that the Nationals will be World Champions. If the Astros have a more plausible 55% chance of winning each game, the Nationals have a 40% chance of popping the champagne.
We can also turn this question around, and ask how much our assessment of the Astros will be affected by how well it does in the World Series. Internet companies use a formula called Bayes’ Rule to estimate and revise the probability that you will like a certain product or service based on information they collect about you. We can do the same thing here, revising our 0.60 estimate of the Astro’s win probability based on how well they do in the World Series.
The answer is not much at all. The Astro’s regular-season record doesn’t guarantee that they will win a 7-game series against the Nationals, and winning or losing the World Series doesn’t tell us much about how good they are. Even if the Astros lose 4 straight games, our best estimate of them beating the Nationals in another game only drops slightly, from 0.600 to 0.594. If the Astros win 4 straight, their probability increases slightly to 0.604. Anything in between has even less effect. No matter how the World Series turns out, it should barely affect our assessment of the Astros.
That is the nature of the beast. In a game like baseball, where so much chance is involved, a 7-game series is all about the luck. The best team often loses, and winning the World series tells us very little about which team is really better.
Enjoy the World Series; I know I will. But notice the luck and remember the 1969 Mets and the 1990 A’s.
If you enjoyed Gary Smith’s take on luck and the World Series, check out his explanation of a fundamental rule of high-level sports:
The paradox of luck and skill Why did Shane Lowry win the British Open golf championship? Because someone had to. In any competition including academic tests, athletic events, and company management where there is an element of luck that causes performances to be an imperfect measure of ability, there is an important difference between competitions among people with high ability and competitions among people of lesser ability.