October Lies: What Baseball’s Playoffs Teach About Luck, Skill, and Judging Outcomes

Last November, the Toronto Blue Jays had the tying run ninety feet away in the eleventh inning of Game 7 of the World Series. Vladimir Guerrero Jr. stood on third. One out. A full building in Toronto and most of a country leaning toward the television.

Alejandro Kirk hit a ground ball to Mookie Betts. Betts stepped on second, threw to first, and the Dodgers were champions.

The Jays had a better regular-season record than Los Angeles. They out-hit the Dodgers fourteen to eleven that night. They led 3–0, then 4–2, and were inside the final three outs before Miguel Rojas, a utility infielder, tied it with a home run in the ninth. If one of a dozen small things goes the other way, the parade is in Toronto and we tell a completely different story about the 2025 Blue Jays.

That's the thing about October. It feels like the most honest part of the baseball season. It's actually the least.

The best team usually doesn't win

Since the wild card arrived in 1995, the team with the best regular-season record in baseball has won the World Series 8 times in 31 seasons, according to season-by-season tracking of the best-record team. That's about 26%. In 11 of those seasons, the best team in baseball didn't win a single playoff series. The 2001 Mariners won 116 games and lost in the American League Championship Series. The 2006 Cardinals won 83 games, roughly the thirteenth-best record in the sport, and won the whole thing.

This isn't a quirk of a few weird years. MLB's own history shows the same pattern going back to the divisional era. It's what the math predicts.

Baseball is a game where the better team loses all the time. A great team wins about 60% of its games. A bad team wins about 40%. Over a full season that gap is enormous, the difference between a parade and a fire sale. In any single game, it's barely more than a coin flip.

Run the numbers. If a team has a genuine 60% chance of winning each game against its opponent, which is generous against another playoff team, it wins a best-of-five series about 68% of the time and a best-of-seven about 71% of the time. String together the three rounds a top seed has to win, and its odds of a title land around one in three. Drop that per-game edge to a more realistic 55%, and the best team in baseball wins the World Series a little more than one time in five.

So when the best team loses in October, nothing has necessarily gone wrong. The system is working exactly as designed. It's just designed to produce drama, not to identify the best team.

What 162 games know that seven games don't

The regular season is a sample size. One hundred and sixty-two games, six months, injuries, slumps, hot streaks, bad bounces, and travel all washing out over time. It still isn't perfect. A team that's genuinely good enough to win 55% of its games will finish below .500 roughly one season in eleven just from noise. But it's the best evidence the sport produces.

The postseason throws most of that evidence away. It takes two teams that spent six months proving who they are, and decides their fate on a handful of games where one bloop single, one hung slider, or one ground ball at the right fielder can swing everything.

This year's version is already playing out. The Milwaukee Brewers finished 103–59, the best record in baseball. As I write this, they lead San Diego two games to one in the Division Series, and all three games have been decided by a single run. They've also lost 12 straight postseason road games, tying a major league record. Nobody thinks the Brewers forgot how to play baseball outside Milwaukee. That's just what small samples look like up close.

And the Blue Jays, the team that came within ninety feet? They finished this season 79–83. One year you're the second-best team on earth. The next you're watching from home. Which version was real? Probably both, and probably neither. That's the uncomfortable part about variance: it doesn't tell you when it's lying.

A .300 hitter fails seven times out of ten and goes to the Hall of Fame. In October, we judge him on four at-bats.

The case for October

The strongest counterargument deserves a fair hearing, because it isn't wrong.

The playoffs do measure something real. Short series reward a different kind of roster. A team with three dominant starters and a deep bullpen can shorten a game and lean on its best arms in a way that's impossible over 162 games. The Dodgers used Yoshinobu Yamamoto in relief in Game 7 on zero days' rest. You can't do that in June. Managers who understand leverage, players who stay composed under enormous pressure, and front offices that build for October rather than April are all showing real skill.

There's also something to be said for a system that gives everyone a chance. If the best record simply won the title, September would be the end of the story for most fan bases. The randomness is part of what makes the playoffs electric. That's a fair trade for a sport.

But notice what that argument concedes. October measures a narrower thing, over a much smaller sample, with far more noise. That's fine for entertainment. It's a terrible way to judge who was actually good.

The problem isn't that baseball has a lottery at the end. The problem is that we walk out of the lottery convinced it was a verdict.

We do this everywhere

The poker player and author Annie Duke has a name for this habit. She calls it resulting: judging the quality of a decision by the quality of its outcome. In Thinking in Bets, she points out that a good decision can produce a bad result and a terrible decision can produce a great one. Treating the result as the grade teaches you the wrong lessons.

Baseball just makes resulting easy to see. Once you notice it, you can't stop seeing it in everything else.

We fire the coach whose team ran into a hot goalie. Every Canadian has watched that one happen. We crown the founder who raised at the top of a bubble and call him a visionary. We pile into last year's best-performing fund, which is often just last year's luckiest fund. We promote the person who closed one huge deal, and quietly overlook the person who has delivered steady, unglamorous results for six straight years.

I've spent most of my career in SEO, where this is a daily hazard. A site gets lucky with a Google update and the team responsible becomes geniuses for a year. Another team does everything right and gets clipped by an algorithm change they didn't cause. The only way to know who's actually good is to watch the work over enough cycles that luck runs out of places to hide.

Politics works the same way. I've argued before that voters should vote like capital allocators, not believers, judging leaders on their record rather than their team colours. The same principle applies here. One election, one quarter, or one viral moment is a seven-game series. A decade of governing is a season.

Judge the at-bat, not the box score

There's an old idea in baseball that a hitter should judge himself by the quality of his at-bats, not his batting average on any given night. Did he swing at strikes? Did he take the walk? Did he square the ball up? A line drive caught at the warning track is a good at-bat. A broken-bat bloop that falls in is a lucky one. Both show up in the box score, and only one of them is repeatable.

That's the Stoic version of baseball, and it's a useful way to live. You control the preparation, the process, and the decision. You don't control the hop.

It's also why effort and process matter more than ever. In a world full of noisy results, the premium goes to the work that holds up over time: the habits, the verification, the judgment that keeps producing good decisions even when the outcomes wobble.

It's the same principle behind how I think about investing. PortfolioAI was built around systems that are tested across years of history, not one great trade, because one great trade tells you almost nothing. A strategy that wins over hundreds of decisions is showing skill. A strategy that won once is showing you a highlight.

Meritocracy needs a sample size

I care a lot about meritocracy. I think people should be rewarded for competence, effort, and real contribution. But meritocracy only works if you measure merit honestly, and honest measurement takes time.

A meritocracy that judges people on one result isn't a meritocracy. It's a casino with better branding.

The fairest thing you can do for the people you hire, the companies you back, and the leaders you elect is to look at the body of work. Ask what they did across enough chances that luck had room to even out. Be humble about short samples, including your own wins. Especially your own wins.

None of this makes October less fun. I'll still be watching, and if you're heading to a game, TheBallparkGuide will tell you where to sit and what to eat. Search interest in MLB is surging right now on DailySearchVolume, up more than 20% in the last month, because millions of people feel the same pull. The drama is the point.

Just don't mistake the drama for the verdict.

October tells you who won. April through September tells you who was good.