100,000 Leagues Under the PECOTA: 2025

Image credit: © Reinhold Matay-Imagn Images

Flip a coin 10 times and you’ll probably come up with a combination of heads and tails. Probably. But not always. On rare occasions, you might get heads 10 times in a row. The odds against it are 1,024 to 1. Your chances of getting 10 consecutive heads are small, but not vanishingly small. It’s entirely possible. You want to know what’s almost impossible? The odds of winning Powerball are one in 292,201,338. Your chances of getting heads 28 times in a row are better. But people still play Powerball. They’re not dissuaded by the odds.

That’s because sometimes really unlikely events occur. Here at BP, we calculate projected standings using PECOTA, our proprietary model. But PECOTA, like all projection systems, has error bars around it. You can see them if you scroll down past the numbers on the standings page. When we produce projections for a team’s win total, or a player’s OPS, we’re not saying they’ll hit it exactly. Rather, we’re saying that’s the likeliest outcome. Each estimate has an implied plus-or-minus factor. Stack a few of those pluses and minuses atop one another, and you can get some pretty big outliers.

I got the idea for this analysis from this 2016 article by Sam Miller, in which he ran a million full-year simulations and noted some of them. All of them had, by definition, at least a one-in-a-million chance of occurring. For the sixth year, I’m doing the same here, using 100,000 simulations. That’s a lot of projections*! Let’s look at some of the most noteworthy ones.

Play It Again: Sim 29992

Last year, the Yankees, Guardians, Astros, Phillies, Brewers, and Dodgers won their divisions. The Orioles, Royals, Tigers, Braves, Mets, and Padres were the wild card teams. There were only 14 simulations that wound up with the same division champions, none that had the same 12 clubs playing in October. But Sim 29992 featured the same postseason combatants in the Senior Circuit.

On Second Thought, 2024 Wasn’t All That Great: Sim 16292

There isn’t a sim in which none of the 12 postseason teams repeat. That’s not surprising; when 40% of all teams make the playoffs, the odds of all of them dropping to the lower 60% approach zero. But in sim 16292, only the Dodgers and Braves get into the playoffs.

Cue The Salary Cap Diatribes: Sim 19637

Per Cot’s Contracts, the 10 teams with the highest payrolls are the Dodgers, Mets, Phillies, Yankees, Blue Jays, Rangers, Astros, Braves, Padres, and Angels, in that order. Sim 196374 sees that as money well-spent. Probably too well-spent for some.

The playoffs would consist of the 10 teams with the highest payrolls, averaging about 93 wins each, plus two Central Division champions. If the Cubs and Twins exit early, the narratives will be excruciating.

In Fact, They’ll Be Really Loud: Sim 57073

The two teams that garnered the most headlines this winter were the Mets, for signing Juan Soto, and the Dodgers, for signing everyone else. Sim 57073 adds fuel to the rein-them-in fire. (Of course, if you limit their payrolls without also reducing their revenues, they’ll just make even more money.)

The Vanishing Middle Class: Sim 34530

Eleven teams with 100 wins or losses (italicized).

The Burgeoning Middle Class: Sim 92829

No team with over 93 wins or 95-losses. All but one American League team within 10 games of .500.

Urban Dominance: Sim 9738

The three largest metropolitan areas in the U.S. are New York, Los Angeles, and Chicago. In sim 9738, they all do pretty well. This is not a small task, given that Los Angeles and Chicago each have one pretty bad ballclub.

Only the White Sox would miss the playoffs.

Urban Non-Dominance: Sim 98716

The big markets fare a lot worse in Sim 98716.

The Dodgers are, of course, like Thanos, inevitable. But none of the other large-market teams have a winning record, much less make the playoffs.

Break Up the Small Markets: Sim 49741

The smallest metropolitan areas in MLB are Milwaukee, Cleveland, Kansas City, Cincinnati, Pittsburgh, and Sacramento. If that’s a disadvantage, you wouldn’t know it by sim 49741.

Depending on how the AL tiebreaker goes, four or five of the six small-market teams would make the playoffs. Only three of the big-market teams would.

Jerry Reinsdorf Makes History, Version 1: Sim 396

No team has ever gone from 120-plus losses in one season to a postseason berth in the next. Until now.

Jerry Reinsdorf Makes History, Version 2: Sim 77481

No team has ever gone from 120-plus losses in one season to an even worse record in the next. Until now.

To Which the Rockies Laugh Derisively: Sim 17920

You could say, “Well, they play in the same division as the Dodgers,” except the sim has the Dodgers failing to win 100 games for only the third time since 2017 (excluding 2020). A 28-134 record, a .173 winning percentage, would be the worst since the formation of the American League in 1901. The only full-season teams worse were the 20-134 1899 Cleveland Spiders and the 23-113 1890 Allegheny City (Pittsburgh) club, which was so bad it didn’t get a nickname.

ESPN Quietly Re-Ups for Sunday Night Baseball: Sim 3516

That Sunday night timeslot doesn’t have a lot of competition, and we know how ESPN will fill it. Over and over and over.

Juan Soto Has Regrets: Sim 5716

Or Maybe He Made the Right Call: Sim 51013

And All the Teams are Above Average: Sims 28152

There were 7,909 sims that had every team in a division above .500. Here’s the best.

And All the Teams are Below Average: Sim 42334

Congratulations to American League Central champion Kansas City Royals!

The Royals would play the loser of the tiebreaker between the 86-76 Orioles and Astros in the wild card round, which would be watched at home by the 84-win Rays and 80-win Angels.

Once Again, the Battleground States Determine the Outcome: Sim 3424

The battleground states in the 2024 Presidential election were Wisconsin, Michigan, Pennsylvania, Arizona, Georgia, Nevada, and North Carolina. The latter two don’t have MLB teams. In Sim 3424, representatives of the other five states win their division.

I’ll bet you didn’t have the Pirates as the Pennsylvania team that wins its division. And, yes, the Brewers and Pirates tied, and who knows how the tiebreaker with the Cubs would work out. But a swing state team can plausibly claim to have won two-thirds of divisions.

The Aesthetically Pleasing Sims 87105, 98100, and 30773

Can you detect the pattern? Here is sim 87105:

The teams are in alphabetical order, by city. The NL Central configuration occurred 1,654 times and the AL East one 421 times. There were only six occurrences of a Houston-Los Angeles-Sacramento-Seattle-Texas AL West, and only one each of Chicago-Cleveland-Detroit-Kansas City-Minnesota and Atlanta-Miami-New York-Philadelphia-Washington. It seems that 2025 is not the year for Arizona-Colorado-Los Angeles-San Diego-San Francisco; it didn’t occur. This should not surprise you.

This is sim 98100:

Alternatively, check out sim 30773:

In those sims, the teams finish in alphabetical order by nickname. That AL East configuration occurred 770 times, AL Central 728, NL West 418, NL Central 274. There were no Angels-Astros-Athletics-Mariners-Rangers or Braves-Marlins-Mets-Nationals-Phillies finishes.

Will They Take Down the Billboards? Sim 69364

Probably not.

The Absolute Best Sim I’ve Ever Seen: Sim 61389

In past years, I’ve highlighted sims in which there were four-way ties for first in divisions. There were several this time as well: 19 in the AL East, four each in the AL West and NL Central, three in the AL East. But I’ve never seen this before:

Please let this come true. This is the sixth year I’ve done this, and this is absolutely the best outcome imaginable.

So You’re Telling Me There’s a Chance: Sims 65736, 60860, and 11585

When you run 100,000 simulations of a season, you get a lot of outcomes. But not every one possible! There are two outcomes that didn’t occur in any of our simulations: Dodgers last and Rockies first. That doesn’t mean they won’t happen (though they won’t). It just means that, per PECOTA, the chance of them happening is less than 1 in 100,000.

But there were three results that exactly hit that 1 in 100,000 threshold. Here are the results that have a 0.001% chance of occurring:

Sim 65736: Rockies finish second

Sim 60860: Braves finish last

That race for third will really capture the public imagination.

Sim 11585: Marlins finish first

This could really put a crimp in the plans of all those teams expecting to trade for Sandy Alcantara before the deadline.

As Predicted by My Birdfeeder

There were 6,895 simulations in which the Blue Jays had more wins than the Cardinals and the Cardinals had more wins than the Orioles. As anyone who owns a birdfeeder frequented by those three species knows, that’s how it goes in the avian world as well. Cardinals, while shy and gentle, have a size advantage over orioles. Blue jays are much larger birds and, frankly, big jerks who don’t like sharing the feeder.

Odds and Ends

The five most common outcomes were Rockies last (99,162 occurrences), Dodgers first (96,069), White Sox last (92,043), Cubs first (84,136), and Marlins last (81,382). The hardest team to predict is the Rays: 14.0% first, 18.1% second, 21.2% third, 23.9% fourth, 22.8% fifth. The single most common divisional finish was Dodgers, Diamondbacks, Padres, Giants, Rockies, which occurred in 34.6% of simulations. The toughest-to-predict division was the American League East, where the most common finish (Orioles, Blue Jays, Yankees, Rays, Red Sox) occurred only 2.1% of the time.

Thanks to Robert Au for running these simulations, and thanks to Baseball Prospectus for footing the bill for the computing time necessary to complete them.

*Yes, I know, the Jules Verne science fiction novel whose title I’m stealing describes 20,000 leagues. We’re giving you five times as many leagues for your subscription!