hello, world

[wrestleknownothing]

Now that I am finally able to post on this board I thought I would start with something you might find interesting.

Let's start with the big picture. Given the current rankings from Intermat, Flo and Wrestlestat how many AA's and how many points should Oklahoma State fans expect?



For this I calculate the probability of AA and points by seed using the last ten years of NCAA results.

You should expect 84 to 85.5 points and between 5.4 and 5.5 AA's come March. You might reasonable ask "how can you have four tenths of an AA?" Well, you can't. So what that means is you have some probability of 5 and some probability of 6 (and other numbers) that adds up to 5.4.

So what does that look like?



Based on that the most likely outcomes in order are:

1. 6 AAs
2. 5 AAs
3. 7 AAs
4. 4 AAs
5. 3 AAs
6. 8 AAs
7. 2 AAs
8. 9 AAs
9. 1 AA
10. 10 AAs (0.018%)
11. 0 AAs (0.004%)

 

[chasepollard]

Now that I am finally able to post on this board I thought I would start with something you might find interesting.

Let's start with the big picture. Given the current rankings from Intermat, Flo and Wrestlestat how many AA's and how many points should Oklahoma State fans expect?



For this I calculate the probability of AA and points by seed using the last ten years of NCAA results.

You should expect 84 to 85.5 points and between 5.4 and 5.5 AA's come March. You might reasonable ask "how can you have four tenths of an AA?" Well, you can't. So what that means is you have some probability of 5 and some probability of 6 (and other numbers) that adds up to 5.4.

So what does that look like?



Based on that the most likely outcomes in order are:

1. 6 AAs
2. 5 AAs
3. 7 AAs
4. 4 AAs
5. 3 AAs
6. 8 AAs
7. 2 AAs
8. 9 AAs
9. 1 AA
10. 10 AAs (0.018%)
11. 0 AAs (0.004%)

Nice, but the math ain't mathing.

I've got us with 10 AA's, and my math has the DT and staff variable as well as the RW and Fish throw you on your head variable.

CY has just been farting around to avoid putting things on film and that variable is un-variable'able.

Jokes aside, good discussion info!

 

[johnstownsteel]

Ill be cheering for them to beat Iowa for second place..

 

[Pennstate1985]

Ill be cheering for them to beat Iowa for second place..

you and me both. If nothing more than to give me several days of feasting on their tears. Do tears have calories?

 

[Chasingthirty-five]

Now that I am finally able to post on this board I thought I would start with something you might find interesting.

You sir, are a gentleman and a scholar. Idk what back alley library you hail from but you’re numbers are welcome here.

Have you run the same numbers with PSU and Iowa?

 

[wrestleknownothing]

You sir, are a gentleman and a scholar. Idk what back alley library you hail from but you’re numbers are welcome here.

Have you run the same numbers with PSU and Iowa?

I have the top table for every team and the bottom graph for a handful of teams (it involves a few extra steps) including PSU and Iowa.

 

[chasepollard]

you and me both. If nothing more than to give me several days of feasting on their tears. Do tears have calories?

No, they’re the most fulfilling and health beneficial intake you can consume.

The health benefits are multiplied when consuming from the Iowa ducts.

 

[Chasingthirty-five]

I have the top table for every team and the bottom graph for a handful of teams (it involves a few extra steps) including PSU and Iowa.

Would love to see it!

 

[wrestleknownothing]

Would love to see it!

When showing a bunch of teams it works better to leave out the weight by weight detail. Here are all the teams expected to have at least two AAs by Intermat or Flo rankings (sorted by Intermat):



The PSU distribution looks like this. They have a 12% chance of 10 AA's. To put that in perspective, when Minnesota had 10 AA's in 2007, based on their seeds they only had a 1% chance of doing that. And their expected AA's was only 6.5. What Minnesota pulled off that year was a pretty extreme outlier, making it even more impressive than it already seems.



And the Iowa distribution looks like this.



And Oklahoma State again, so you can see them lined up against each other.

 

[chasepollard]

Good stuff @wrestleknownothing , thank you!

But your data is missing a variable about Iowa under-performing in March. So, be sure and knock some of that probability down.

 

[wrestleknownothing]

Good stuff @wrestleknownothing , thank you!

But your data is missing a variable about Iowa under-performing in March. So, be sure and knock some of that probability down.

While that is true, you may not want to look at this table too closely. It is 2010 - 2024 (The Sanderson Era) performance relative to seed, sorted by "Worse Than Seed" for select teams.



I know, I know. New regime. And now we have a baseline to compare to.

This table shows the performance as averages rather than buckets, sorted by average deviation from seed. Neither one is perfect. Starting points matter. But combined they tell a pretty accurate story.

 

[chasepollard]

Do Iowa the last 4 years only?

 

[wrestleknownothing]

Do Iowa the last 4 years only?

Now you are on to something.

 

[chasepollard]

Nice!

 

[west2east]

Good stuff @wrestleknownothing , thank you!

But your data is missing a variable about Iowa under-performing in March. So, be sure and knock some of that probability down.

Can't the same be said about Oklahoma st over the last 10 years 🤷😜

 

[chasepollard]

Can't the same be said about Oklahoma st over the last 10 years 🤷😜

Yup

 

[west2east]

Yup

That won't continue with the new staff. I expect Oklahoma St to peak in March from now on

 

[chasepollard]

That won't continue with the new staff. I expect Oklahoma St to peak in March from now on

I expect it as well. But, look at how well Cael has improved them, 30% below seed to 18% the last four.

That’s tough to do when you get 10 seeded to AA.

 

[Chasingthirty-five]

Thanks @wrestleknownothing for sharing! Excited to see how we finish to seed this year.

 

[cowboyfan05]

I think comparing NCAA finish with Intermat ranking could be more meaningful. Carter Starocci was undefeated (two inj defaults) and coming off of 3 consecutive NCAA championships and was given the NINE seed. I don't think him placing 8 spots above his seed is a meaningful statistic

 

[Chasingthirty-five]

I don't think him placing 8 spots above his seed is a meaningful statistic

That’s fair and I would love to see that too. However you probably just picked the biggest outlier when the guy has years of data. There will always be outliers.

 

[wrestleknownothing]

I think comparing NCAA finish with Intermat ranking could be more meaningful. Carter Starocci was undefeated (two inj defaults) and coming off of 3 consecutive NCAA championships and was given the NINE seed. I don't think him placing 8 spots above his seed is a meaningful statistic

While it is an interesting idea, it has problems of its own.

1. Not all of the final 33 Intermat ranked wrestlers even make the tournament. There would be incomplete data.
2. And there is not archive of intermat rankings to analyze. With seeds you can go back as far as is practical and find the data.
3. The seeds are also deterministic in a way that rankings are not. The seeds determine who you wrestle, not the rankings.
4. And then you are talking about replacing all 3300 observations because of a disagreement with one observation.

All that said, I think I have three years of intermat rankings, so I will see who did better, the intermat staff or the seeding committee.

 

[pokepelli]

But honestly…cool info. We should have a math/stat thread for random stats both true and untrue.

 

[wrestleknownothing]

But honestly…cool info. We should have a math/stat thread for random stats both true and untrue.

Its OK. I know who I am.

 

[chasepollard]

Its OK. I know who I am.

I’ll ban him for rough talking the stat guy!

 

[wrestleknownothing]

I think comparing NCAA finish with Intermat ranking could be more meaningful. Carter Starocci was undefeated (two inj defaults) and coming off of 3 consecutive NCAA championships and was given the NINE seed. I don't think him placing 8 spots above his seed is a meaningful statistic

I took a quick look at 2024, and so far Intermat is better than the seedings. While Intermat had 18 ranked wrestlers who did not make the tournament, the 18 who made it in their place only scored 17.5 points (1.2% of the total). So, not very consequential.

Interestingly though, Starocci was only the fourth biggest miss by the seeds (he was the biggest miss relative to Intermat). Vito Arujau, Braeden Davis, and Richie Figueroa were bigger misses for the seeding committee because they scored more bonus than Starocci.

The absolute error was also larger for the seeding committee than Intermat (2.8 per seed vs. 2.6).

That said, it is hard to overcome the lack of historical data in Intermat rankings.

 

[cowboyfan05]

I took a quick look at 2024, and so far Intermat is better than the seedings. While Intermat had 18 ranked wrestlers who did not make the tournament, the 18 who made it in their place only scored 17.5 points (1.2% of the total). So, not very consequential.

Interestingly though, Starocci was only the fourth biggest miss by the seeds (he was the biggest miss relative to Intermat). Vito Arujau, Braeden Davis, and Richie Figueroa were bigger misses for the seeding committee because they scored more bonus than Starocci.

The absolute error was also larger for the seeding committee than Intermat (2.8 per seed vs. 2.6).

That said, it is hard to overcome the lack of historical data in Intermat rankings.

Figs wasn't mis-seeded at least. He had a mid regular season and then brought it together in the tourney. Vito was seeded too low and Davis was probably seeded too high iirc

 

[Pennstate1985]

Figs wasn't mis-seeded at least. He had a mid regular season and then brought it together in the tourney.

Same thing happened to Quentin Wright for PSU in 2011. Average regular season but caught fire at B1Gs and then won nationals from the 9 seed. They are just statistical anomalies which in a way prove the methodology.

 

[tom_v31]

That won't continue with the new staff. I expect Oklahoma St to peak in March from now on

Depends on our seeding. Ohio state actually performed better than I thought according to their seeds.

 

[Pennstate1985]

Depends on our seeding. Ohio state actually performed better than I thought according to their seeds.

yeah, I would have guessed Cornell myself. I wonder if Ohio State is underperforming at B1Gs and thus having an abnormally lower seed at nationals and thus "overperforming".

 

[west2east]

yeah, I would have guessed Cornell myself. I wonder if Ohio State is underperforming at B1Gs and thus having an abnormally lower seed at nationals and thus "overperforming".

Ohio st has had unbelievable recruiting classes and shouldve been way higher at NCAA then theyve been

 

[west2east]

https://rokfin.com/article/17509/Ten-Years-Of-Recruiting-Rankings

In the last 10 years Ohio st has had 9 top 10 recruiting class and that is the most. Penn st, Oklahoma st, Nebraska and Iowa have had 7 classes in the top 10. Penn st has had 6 classes in the top 5. And Oklahoma st and Ohio st 5 in the top 5. Shows how teams are developing guys

 

[wrestleknownothing]

Depends on our seeding. Ohio state actually performed better than I thought according to their seeds.

People tend to remember the high seeds who underperform, but forget the low seeds who outperform.

In 2021 Ohio State had D'Emilio seeded #30. That is expected to be an 0-2, but he went 2-2. Similarly, Rocky Jordan won 3 on the backside as a #24 seed. But the big winner was #21 seed Tate Orndorf who pulled the first round upset, won 3 more on the backside and finished 8th.

And each of the other four years they had one guy outperform his seed by double digits. The biggest was 2022 197lber, Gavin Hoffman. Seeded #21, he had three straight upset wins on the front side to make the semis. He eventually finished 6th, 15 spots over his seed.

 

[Pennstate1985]

https://rokfin.com/article/17509/Ten-Years-Of-Recruiting-Rankings

In the last 10 years Ohio st has had 9 top 10 recruiting class and that is the most. Penn st, Oklahoma st, Nebraska and Iowa have had 7 classes in the top 10. Penn st has had 6 classes in the top 5. And Oklahoma st and Ohio st 5 in the top 5. Shows how teams are developing guys

I saw a stat the other day. In the past 10 years, PSU has hauled in 18 top ten pfp recruits. Ohio State has had 14, the next highest by far. So they may be over performing against seed but they definitely are underperforming with pfp recruits. Trying to understand the disconnect in the data. Any ideas??

I just don’t believe they have been coached up to the degree the stats say. Curious.

 

[west2east]

I saw a stat the other day. In the past 10 years, PSU has hauled in 18 top ten pfp recruits. Ohio State has had 14, the next highest by far. So they may be over performing against seed but they definitely are underperforming with pfp recruits. Trying to understand the disconnect in the data. Any ideas??

Them underperforming all year so they get bad seeds and then they do a little better at NCAA but nowhere near where they should

 

[wrestleknownothing]

I saw a stat the other day. In the past 10 years, PSU has hauled in 18 top ten pfp recruits. Ohio State has had 14, the next highest by far. So they may be over performing against seed but they definitely are underperforming with pfp recruits. Trying to understand the disconnect in the data. Any ideas??

I just don’t believe they have been coached up to the degree the stats say. Curious.

Here is my post that I think you are referring to.

As for Ohio State they have been the third best team over that time period, and one of three to win a title, so they are not exactly slacking.

Yes, their more recent monster classes have not yet paid big dividends, but they also still have a lot of eligibility left.

Posted November 3, 2023
Now that we have established that the Big Board matters, lets take a look at who is optimizing (haha, we know who it is) the Big Board talent.

• From 2015 to 2022 (2023 class has not competed yet) the race for the most Big Board Top 20 wrestlers is pretty tight.
• PSU has the edge over Ohio State 18 to 15.
• But PSU has gotten 27 AA finishes from those 18 wrestlers (1.5 per) while Ohio State has gotten "only" 18 from 15 (1.2).
• But where things get really crazy is that 23 of PSU's 27 AA finishes were first, second, or third (85%), while Ohio State had 7 top threes among their 18 AA finishes (39%).
• Iowa gets trashed a lot for under-performing, but the real issue is they under-recruit. They got commitments from only 9 Big Board Top 20 wrestlers. So, even though those wrestlers averaged 2 AA finishes per, there just were not enough of them.
• One caveat. The school where the wrestler committed gets credit for all the wrestlers AA finishes regardless of where they were achieved. This hurts Iowa with a lot of recent transfer AAs (DeSanto, Eierman, Woods) and PSU (Dean, Kerkvliet). My logic is this is a measure of high school recruiting rather than retention. For a fuller picture, you would want to look at retention too (independent study opportunity?).

It could be that PSU is just getting the best of the best leading to those best in class results.

• But, that is not completely the case. Ohio State has brought in slightly higher ranked wrestlers on average.
• ASU has also brought in higher ranked wrestlers on average, though significantly fewer of them.

 

[Pennstate1985]

Now those are some chewy stats for recruitment.

I wonder what the average ncaa placement is per average top 10 pfp by team. You’ve done a fabulous job on the recruiting side but I’m trying to get a sense for how well the coaches are getting them to progress.

 

[tom_v31]

I saw a stat the other day. In the past 10 years, PSU has hauled in 18 top ten pfp recruits. Ohio State has had 14, the next highest by far. So they may be over performing against seed but they definitely are underperforming with pfp recruits. Trying to understand the disconnect in the data. Any ideas??

I just don’t believe they have been coached up to the degree the stats say. Curious.

Well, performing below expectations during the season, getting low seeds and then performing closer to expectations in postseason, is the first thing that comes to my mind. I always wonder how Ohio state could be in such a hotbed of talent, yet seem to perform, what I would consider, marginally well, I never understand. They seem to have the universities support, they have the recruits, I imagine they have the funding. 🤷🏼‍♂️

 

[wrestleknownothing]

Now those are some chewy stats for recruitment.

I wonder what the average ncaa placement is per average top 10 pfp by team. You’ve done a fabulous job on the recruiting side but I’m trying to get a sense for how well the coaches are getting them to progress.

I think the first table answers your question. PSU got 15 titles from their 18 top 20 recruits. Cornell got 5 from 6. Same ratio, but divided by three. No one else is anywhere near those results.

 

[west2east]

Here is my post that I think you are referring to.

As for Ohio State they have been the third best team over that time period, and one of three to win a title, so they are not exactly slacking.

Yes, their more recent monster classes have not yet paid big dividends, but they also still have a lot of eligibility left.

Posted November 3, 2023
Now that we have established that the Big Board matters, lets take a look at who is optimizing (haha, we know who it is) the Big Board talent.

• From 2015 to 2022 (2023 class has not competed yet) the race for the most Big Board Top 20 wrestlers is pretty tight.
• PSU has the edge over Ohio State 18 to 15.
• But PSU has gotten 27 AA finishes from those 18 wrestlers (1.5 per) while Ohio State has gotten "only" 18 from 15 (1.2).
• But where things get really crazy is that 23 of PSU's 27 AA finishes were first, second, or third (85%), while Ohio State had 7 top threes among their 18 AA finishes (39%).
• Iowa gets trashed a lot for under-performing, but the real issue is they under-recruit. They got commitments from only 9 Big Board Top 20 wrestlers. So, even though those wrestlers averaged 2 AA finishes per, there just were not enough of them.
• One caveat. The school where the wrestler committed gets credit for all the wrestlers AA finishes regardless of where they were achieved. This hurts Iowa with a lot of recent transfer AAs (DeSanto, Eierman, Woods) and PSU (Dean, Kerkvliet). My logic is this is a measure of high school recruiting rather than retention. For a fuller picture, you would want to look at retention too (independent study opportunity?).

It could be that PSU is just getting the best of the best leading to those best in class results.

• But, that is not completely the case. Ohio State has brought in slightly higher ranked wrestlers on average.
• ASU has also brought in higher ranked wrestlers on average, though significantly fewer of them.

Funny because people say Iowa can't coach yet that's not what those #s say. In that time period they've been better than Ohio St and Oklahoma st. People don't like how Tom & Terry act at times and they dont like them so they let that turn into they can't coach.