2026 Expected Points and All-Americans By Seed

[Wrestleknownothing]

I spent a little time yesterday folding the 2025 results into the model.

First, expected points, including bonus, per seed. Not a lot of big changes, but the hump in the 26/27 range has begun to flatten out a bit.

The other interesting thing is, despite the change in scoring there was not a noticeable change in bonus. While it is higher than the previous two years, those years were very down. I am looking at 11 years of data (the 16 and 32 seed era), and across that range 2024/25 bonus scoring does not stand out.

 

[Wrestleknownothing]

Next comes the probability of All-American by seed.

 

[Wrestleknownothing]

And a new one this year.

Taking a barbell approach, the most interesting results are often the extremes. So, here I have the probability of being in the final two versus the probability of going 0-2, BBQ.

 

[chasepollard]

As always, good stuff

 

[alephpoke]

I spent a little time yesterday folding the 2025 results into the model.

First, expected points, including bonus, per seed. Not a lot of big changes, but the hump in the 26/27 range has begun to flatten out a bit.

The other interesting thing is, despite the change in scoring there was not a noticeable change in bonus. While it is higher than the previous two years, those years were very down. I am looking at 11 years of data (the 16 and 32 seed era), and across that range 2024/25 bonus scoring does not stand out.

Just curious, do you get this data from web-scraping or did you manually compile it over the years? Awesome stuff.

 

[Wrestleknownothing]

Just curious, do you get this data from web-scraping or did you manually compile it over the years? Awesome stuff.

It started out with me getting bored during COVID and deciding to start building a dataset of NCAA tournament results. That was all manual. I mentioned on Intermat that I was doing this and I got an offer from someone to help. So the two of us manually input results for a bunch of years.

The first thing I did with that was start projecting numbers of AAs per team. But one thing led to another and I began converting that into advancement and placement points. Bonus points came along later.

Since that first data set I have had people reach out to me offering data they have assembled, but did not know what to do with. I have also found multiple things I can scrape with my limited skills. While others have asked for help QAing stuff they have done, and still others have reached out just to talk.

It is that last category that has been the most gratifying as I have made friends at the NWHOF, with national media figures, and with the inventor of the Hodge Trophy.

 

[proud2bpokes]

Awesome stuff! I bet it has been a lot of fun to put together.

 

[OSUMatFan]

Moneyball for wrestling? I love your work!

 

[chasepollard]

Moneyball for wrestling? I love your work!

Yeah, it’s really, really strong stuff! You can actually make money using his data.

 

[Cali_Nittany]

And a new one this year.

Taking a barbell approach, the most interesting results are often the extremes. So, here I have the probability of being in the final two versus the probability of going 0-2, BBQ.

This one is fascinating. I would've never guessed that the intersection of the odds between going 0-2 vs making the finals was at the 7 seed.

 

[OAMC99]

Great post. Thank you for the research. My initial impression is that they get the seeding pretty darn spot on.

 

[Wrestleknownothing]

Great post. Thank you for the research. My initial impression is that they get the seeding pretty darn spot on.

That is true. But for every rule there is an exception.

See my other post about the #5 seed this year. Rough.