2023 NCAA's in data

[Wrestleknownothing]

Let's start with the team race.

Expectations

We all know who won and by how much. But who made the seeding process look flawed? Well clearly that is American, Clarion, Drexel and Northern Colorado. They just need a few more wrestlers. Sometimes it is more about how you perform relative to expectations, than how you perform outright.

But seriously, the out-performers of the tournament were Ohio State and Michigan. Of course, there is more room to out-perform when you start with lower initial seeds, but Ohio State's rise was enough to take them from a pre-tournament predicted 9th place finish all the way up to an actual fourth place finish.

Penn State? Right on seed. For all the drama of a 2 for 5 Saturday night, really no drama thanks to guys like Bartlett and Van Ness beating their seed (and to a lesser extent Dean).

Missouri over-seeded? Not so much. Can we put that one to bed? No. Never. That was foolish of me. Sorry.

Bracketology

Most busted? You got it. 125 was the most busted, but not for that reason,. For the other reason. We can credit App State's five seed for the biggest contribution followed by Oregon State's seven seed. That Iowa kid was only the third biggest contributor.

On the flip side, 157 went closest to chalk. Seven of the top eight seeds finished on the podium and the won who did not, made it to the blood round (where he lost in SV to another top eight seed). For standard deviation though, it is actually 174. Take your pick.

 

But Why Is It Always About the Big Dogs?

After all, who doesn't love an under dog? The biggest out-performance always comes from the guys with the most to gain. Stand up and take a bow, gentlemen.

• Starting with Northern Colorado's Vinny Zerban who came into 157 as the 32 seed, but made the blood round for a Rocky Mountain High out-performance of 23 slots (I give him credit for ninth. Because). 
• Next we have Virginia Tech's Eddie Ventresca who tore up his bracket, screamed at the Wrestling God's and said, "seeds don't matter." He came in as a 27 seed, but left as an All-American (7th).
• Not to be outdone, West Virginia's Killian Cardinale said, "yes, please, and thank you. I will take some of that." As he came in as the 28 seed at 125 and finished finished on the podium (8th). (OK, he was slightly outdone in that Ventresca beat him 7-6 in the seventh place match, but come on, why ya gotta be like that?)
• And not exactly bringing up the rear (but coming from there) Cal Poly's Dom Demas exemplified the "chip and a chair" mentality. When given a second bite at the apple by being awarded the 33 seed at 149 after an injury to another wrestler, he made it all the way to the blood round's blood round (let it be a thing) for a healthy 20 slot jump.
• Last on my arbitrarily sorted list, but first in Sparty's heart is Caleb Fish. The 165 pounder was seeded 29, but made the blood round. That Fish is a keeper.

[goheels1812]

This was a fun read and really insightful. Thanks for the work you put into it!

[Wrestleknownothing]

Forgot to do Better, At Worse analysis for each team.

How it is organized:

• First three columns are the % of a teams wrestler that fit into each column (Better Than Seed, At Seed, Worse Than Seed).
• The second set of three columns is the average number of positions they were in the three categories.
• The third set of three is their average starting position (i.e. seed).

Sorted by % Better:

It may be a while before we here from @jajensen09

[goheels1812]

I’m still a little disappointed in UNC’s season (losing dual record, first ever dual loss to app state, only 5 qualifiers which is down from 9 a few years ago), but I can’t deny how well Coleman prepares his guys for March. 

5/5 wrestled at or above their seeds. Gavin Kane lost his first match and then ran the gauntlet to All-American. Even the most optimistic UNC fan wouldn’t have predicted Lachlan McNeil to place 4th this year. We now have 3 straight years with a wrestler competing Saturday night for a national championship, with 2 wins. 
 

UNC continues to average 3 All Americans per season and knocks off guys heavily favored against them in March. It’s a nice little thing going in Chapel Hill. 

[Wrestleknownothing]

1 minute ago, goheels1812 said:

I’m still a little disappointed in UNC’s season (losing dual record, first ever dual loss to app state, only 5 qualifiers which is down from 9 a few years ago), but I can’t deny how well Coleman prepares his guys for March. 

5/5 wrestled at or above their seeds. Gavin Kane lost his first match and then ran the gauntlet to All-American. Even the most optimistic UNC fan wouldn’t have predicted Lachlan McNeil to place 4th this year. We now have 3 straight years with a wrestler competing Saturday night for a national championship, with 2 wins. 
 

UNC continues to average 3 All Americans per season and knocks off guys heavily favored against them in March. It’s a nice little thing going in Chapel Hill. 

Yeah, but that guy who only wrestled to seed needs to do much better. And don't start with the excuses like he was the #1 seed.

[goheels1812]

3 minutes ago, Wrestleknownothing said:

Yeah, but that guy who only wrestled to seed needs to do much better. And don't start with the excuses like he was the #1 seed.

I’ll put in a phone call to the UNC room and let Austin know he needs to step it up because the message board doesn’t have time for underachievers. 

[Wrestleknownothing]

14 minutes ago, goheels1812 said:

I’ll put in a phone call to the UNC room and let Austin know he needs to step it up because the message board doesn’t have time for underachievers. 

Feel free to keep the source confidential

[ionel]

1 hour ago, Wrestleknownothing said:

Feel free to keep the source confidential

Hey ... AOC ... know nothing on line 3!!

[Wrestleknownothing]

Another way to think about performance is performance vs expectations. We do this all the time. It is what makes us feel good about a 33 seed making a run to the quarter or semi finals. Using my AA probability metric (% of times a given seed achieves AA between 2010 and 2022) we can look at what the pre-tournament expectations were versus what actually happened. The good news is that no one massively under-performed. The biggest miss was by one. But this also smooths over high-end upsets as long as the upset still AAs.

[Offthemat]

Who is the third expected AA for OKST?

[goheels1812]

Loving all of this data. It’s fun to look through. Curious the metric you used to say UNC was expected to have 2 AAs? Was your 2nd McNeil? I was pretty confident he’d AA as long as he got a reasonable blood round draw, but he was a 10 seed so not technically projected to finish top 8 using seeding. 

[Wrestleknownothing]

1 hour ago, Offthemat said:

Who is the third expected AA for OKST?

 

58 minutes ago, goheels1812 said:

Loving all of this data. It’s fun to look through. Curious the metric you used to say UNC was expected to have 2 AAs? Was your 2nd McNeil? I was pretty confident he’d AA as long as he got a reasonable blood round draw, but he was a 10 seed so not technically projected to finish top 8 using seeding. 

The metric I use is to sum the empirical results by seed from 2010 - 2023 and use that to come up with a percentage expectation that a given seed will AA. For example, during that time period only 49% of #8 seeds achieved AA status, while 36% of #9 seeds did as well. So counting an 8 seed as an AA and a 9 seed as not ignores this reality. Even a 1 seed is not a sure thing (but close at 98.5%). The probabilities I use look like this:

As for OkSt and UNC, I saw it like this:

In both instances it was not about predicting a single wrestler to make it, but between two wrestlers with a roughly one-third shot of making AA you would expect a decent shot of at least one of them achieving it.

[goheels1812]

9 minutes ago, Wrestleknownothing said:

 

The metric I use is to sum the empirical results by seed from 2010 - 2023 and use that to come up with a percentage expectation that a given seed will AA. For example, during that time period only 49% of #8 seeds achieved AA status, while 36% of #9 seeds did as well. So counting an 8 seed as an AA and a 9 seed as not ignores this reality. Even a 1 seed is not a sure thing (but close at 98.5%). The probabilities I use look like this:

As for OkSt and UNC, I saw it like this:

In both instances it was not about predicting a single wrestler to make it, but between two wrestlers with a roughly one-third shot of making AA you would expect a decent shot of at least one of them achieving it.

Love this. Thanks for the explanation!

[Idaho]

My counselor warned me that I may have to talk about these stats...