109 Teams Have Been Exactly Where Boston Was on July 3. One Made the Playoffs.

Since 1998, 109 teams have sat where Boston sat on July 3. One made the playoffs. The Red Sox have gone 25–3 since — about a one-in-70,000 run — and are now 94% to reach October. This is not a story about belief being rewarded. It is a warning about every lesson drawn from the team that got lucky.

The question: the Red Sox were 37–48 on July 3 and are now 94% to reach October. What should we learn from that?

The answer: almost nothing. Since 1998, 109 teams have sat roughly where Boston sat. One made the playoffs.

Why that is surprising. The run is real, and the odds against it were about one in 70,000; that is exactly why it cannot teach us anything general.

On the morning of July 3 the Red Sox were 37–48, eleven games under .500, going nowhere.

Then they won fifteen straight, tying the franchise record set in 1946. They have not really stopped. Since that morning they are 25–3. They swept the Dodgers. They are 62–51, hold the second wild card by six games, and our model gives them a 94% chance of playing in October.

It is the best story in baseball, and you are going to hear about it for two months. It is also about to be used to teach a lesson it cannot support. So let us get the number on the table first.

The Base Rate

Go back through every season since 1998 — the whole 30-team era — and find every club that sat roughly where Boston sat in early July; a winning percentage between .415 and .455, about eighty-five games in.

Boston’s .435 lands in the middle of that band.

There are 109 such team-seasons.

One of them made the playoffs.

One caveat, and it is real. The playoff field has grown.

From 1998 to 2011 only eight clubs qualified, and none of these 109 made it. The field went to ten, then twelve in 2022. The single success comes from that recent era.

Look only at 2022 onward and the rate is one in eleven. Too small a sample to trust, but enough to say the honest modern figure is higher than one percent. Call it somewhere under one in ten. Still the long odds.

How Unlikely the Run Itself Was

Take Boston’s July 3 record at face value and ask what it said about the team.

Thirty-seven and forty-eight, run through the same Bayesian machinery this newsletter uses all year, says a club whose true quality was about .46. A bit below average, with a plausible range from roughly .41 to .52.

A team of that quality wins 25 of its next 28 games about once in 70,000 tries. The fifteen-game streak on its own is a one-in-50,000 event.

Those numbers are not an insult to the Red Sox. They are the point.

Something happened that the team’s first eighty-five games gave almost no reason to expect. Maybe the roster really was better than its record. That happens, and a .46 estimate carries real uncertainty.

But nothing available on July 3 justified predicting this. Anybody who says they saw it coming is describing a memory, not a forecast.

The Trap

Now watch what happens next, because it always does.

Boston will make the playoffs and a hundred columns will explain why. They never panicked. They bought at the deadline instead of selling. The clubhouse believed.

Every one of those explanations will be built on a single case that was chosen because it worked out.

That is survivorship biasi in its purest form.

The other 108 clubs also had clubhouses that believed. Some also bought at the deadline. Nobody will write about them, because nothing happened to them.

So the evidence that would let you judge whether belief or aggression actually works has been quietly removed from the record before the argument even starts.

The base ratei is the antidote, and it is almost never in the story. It is the denominator the highlight reel is missing.

What to take home: count the ones that vanished

None of this means Boston is a fraud. They won the games. Their 94% is real. If they win the World Series the trophy will not carry an asterisk.

Rare things happen; somebody has to be the one square, and this year it is them. You are allowed to enjoy that.

What it means is narrower and more useful: we cannot learn a general rule from the case that was picked for being exceptional.

Next time you read that a company survived by refusing to give up, or a fund returned 40% by ignoring the consensus, or a coach was right to go for it, ask the one question that makes the claim testable.

How many did the same thing and vanished?

If nobody counted them, you are not being offered a lesson. There is only a good story, which is a different product entirely.

The Red Sox are having a wonderful summer. Enjoy it as a rare event, which is what it is. And be careful with anyone who tries to sell it to you as a method.