2023 NCAA Tournament Projection Based on Actual, Real, Live SEEDS

[Wrestleknownothing]

Oh, Joy. We have actual seeds.

No more messing around with ranks that are....I know, I know...NOT predictions (even though they are sometimes).

HEADLINE: THE GAP IS CLOSING

When last we gathered to perform this exercise immediately prior to the conference tournaments PSU was a solid 31 point favorite.

Now with guys like Max Dean (9th) and Aaron Brooks (3rd) seeded below their rank, PSU is coming back to the field and stands as an 18 point favorite. Still a clear favorite, but well within the bounds of volatility such that you have to give Iowa at least a puncher's chance (maybe more). Add to that PSU is not the bonus machine of years past and we could have a real barn burner of a tournament.

[ionel]

Actual Real Live Seeds:

[Wrestleknownothing]

And Actual Real Live Expected All Americans by Team:

For those of you who hate the decimal think of it as a rudimentary probability. So for PSU 5.6 means they have a 40% chance of 5 and a 60% chance of 6 AA's (I am simplifying a bit).

[ionel]

So you are still saying Penn St has a chance?  

[1032004]

It’s like Jimmy was on the seeding committee so he could tell us how PSU was gonna lose

[BerniePragle]

As always, very cool.  Now we see how good of a job the seeding people did (and the various ranking people).

I don't understand the scoring system much at all.  Will it be easy to separate out the bonus points going forward, or would it still require looking at every individual match in the tournament?  (Obviously a ton of work and not really sure of the value.)

BTW, 3 ACC teams in top 11.  Cool.

[Wrestleknownothing]

13 minutes ago, BerniePragle said:

As always, very cool.  Now we see how good of a job the seeding people did (and the various ranking people).

I don't understand the scoring system much at all.  Will it be easy to separate out the bonus points going forward, or would it still require looking at every individual match in the tournament?  (Obviously a ton of work and not really sure of the value.)

BTW, 3 ACC teams in top 11.  Cool.

It is not hard per se, just labor intensive. When you and I put together the data set I have been using for this we focused on endpoints (the #1 seed finished 2nd, the #2 seed finished 1st, the #3 seed finish round of 12, etc.). For bonus points we would also need to track each of their pathes to know what bonus they scored, and ideally where they scored it.

At one point last year, someone reached out to me to say they had a dataset that contained the detailed results of every NCAA match. They were looking to monetize this via Twitter and a website. We talked about providing visualizations and content, but he seems to have faded. I thought I could save some labor this way, but alas it was not to be.

[PortaJohn]

7 hours ago, 1032004 said:

It’s like Jimmy was on the seeding committee so he could tell us how PSU was gonna lose

Not going to lie. As much as I complain about him I'm looking forward to his annual "Why PSU is not going to win thread"

[Mr. PeanutButter]

3 hours ago, PortaJohn said:

Not going to lie. As much as I complain about him I'm looking forward to his annual "Why PSU is not going to win thread"

It will probably come after the 1st round, when some team like OSUw is in the lead