Gil Put Me on the Spot. 15 Minutes Later, We Had Modeled the Entire Market.

Today on the show, Gil Alexander asked me a question I hadn't specifically modeled before:

Where is the value in the World Series Exact Outcome market?

This is the market where you're not simply betting a team to win the World Series. You're predicting both teams that will get there and which one will win.

For example:

Tampa Bay Rays to beat Philadelphia Phillies +7000

I was a little bit on the spot.

My initial answer was that I thought the interesting combinations would probably involve Tampa Bay and Philadelphia.

That wasn't the result of having an Exact Outcome model sitting in front of me. It was my read of what our existing numbers were telling me.

After the show, I decided to find out.

And this is where things got pretty cool.

From a Question to a Model in About 15 Minutes

We've spent years building the Basewinner data and modeling pipeline.

We already calculate probabilities for teams to win their leagues. We already calculate probabilities for teams to win the World Series. We have our underlying team ratings, simulations and postseason probabilities.

What we didn't have was a model specifically designed to price this particular Exact Outcome market.

With the existing Basewinner data pipeline — and with AI helping me write and connect the calculations — I had one running in roughly 15 minutes.

That's an important distinction.

I didn't build a baseball model in 15 minutes.

I was able to answer a completely new betting-market question in 15 minutes because the infrastructure to answer it was already there.

That's one of the things I think is becoming increasingly powerful about the combination of proprietary data, modeling and AI.

You can go from:

"That's an interesting question."

to:

"Let's actually calculate it."

very quickly.

How We Priced It

The Basewinner simulation already gives us two critical numbers for each team:

1. Probability of winning its league and reaching the World Series.

2. Probability of winning the World Series.

From those, we can calculate a team's conditional probability of winning the World Series given that it gets there.

We then transform those conditional championship probabilities into relative World Series strengths so that we can estimate a head-to-head series probability for each possible AL/NL matchup.

Finally:

Exact Outcome Probability = Probability Team A wins its league × Probability Team B wins its league × Probability Team A beats Team B

That gives us a model probability for every exact outcome.

We convert that probability into fair American odds and compare it with the sportsbook price.

Now we have a market.

And My Answer to Gil Held Up Pretty Well

This was the fun part.

When I sorted every Exact Outcome by expected value, the top of the board looked like this:

So before running any of this, I told Gil that Tampa Bay/Philadelphia combinations were where I thought we'd find value.

After actually modeling the market, Tampa Bay over Philadelphia came out No. 2, and Philadelphia over Tampa Bay came out No. 3.

I'll take that.

More importantly, though, the model tells us why.

Tampa Bay Over Philadelphia +7000

The model makes this exact outcome approximately:

2.215% probability

That translates to fair odds of approximately:

+4414

The market is offering:

+7000

At +7000, the implied probability is only about 1.41%.

Our model has it at 2.22%.

That's a meaningful difference in a market built around extremely low-probability outcomes.

Philadelphia Over Tampa Bay +5000

Interestingly, flipping the World Series winner doesn't eliminate the value.

The model gives Philadelphia over Tampa Bay approximately:

3.048% probability

Fair price:

+3181

Market:

+5000

So the Basewinner numbers currently suggest that the matchup itself may be mispriced, with value available on either side depending on which team you want holding the trophy.

That's exactly the type of thing an Exact Outcome market can hide.

Tampa Shows Up Everywhere

The No. 1 value on the entire board was actually:

Tampa Bay over Milwaukee +4500

Our number:

3.445%

Fair odds:

+2803

And another Tampa combination landed sixth:

Tampa Bay over Los Angeles +3000

Model probability:

4.448%

Fair odds:

+2148

So my initial Tampa observation wasn't limited to one opponent.

The Rays appear repeatedly near the top of the value rankings.

Philadelphia does too.

The Bigger Point

I love the betting analysis here, but what happened behind the analysis might be even more interesting to me.

Gil asked a question for which I didn't have a prepared table.

Fifteen minutes later, I had priced the market.

Not because the answer was sitting somewhere on the internet.

And not because AI magically knew the answer.

We had the underlying Basewinner probabilities.

We had the simulation infrastructure.

We had the data pipeline.

AI helped me rapidly turn those existing pieces into another layer of analysis.

That's where I think this gets really powerful.

For years, building another model or attacking another market meant writing everything from scratch, debugging it, connecting databases and spending hours — sometimes days — getting from an idea to an output.

Now, when the underlying infrastructure is already good, the distance between question and quantitative answer can become incredibly short.

Today it was about 15 minutes.

And, fortunately for me, the numbers also backed up the answer I gave Gil on the air.

Tampa Bay and Philadelphia were exactly where we needed to look.

Basewinner model probabilities and market comparison as of September 2, 2026. Exact Outcome markets involve long odds and low-probability events; model edge represents an estimate, not a guarantee of outcome.