NCAA Tourney Projections Based on Rankings - Mind The Gap Edition

[Wrestleknownothing]

For the final edition of this post, prior to the final edition of this post, I thought I would take a look at the full season.

Using Intermat's rankings from today to determine the current top 10 teams I traced their expected points back through time to the pre-season rankings. 

• The gap at the top has grown significantly as the PSU freshmen first entered the rankings and then began to climb the rankings, growing from an expected lead over Iowa of 16.8 points to the current gap of 30.9 points.
• Meanwhile, the expected gap between #2 and #3 has shrunk from 14.8 to 10.2
• While #1 and #2 have remained fairly constant, #3 and #4 have not. Ohio State, Cornell, Michigan, and Iowa State have each had dalliances with the podium positions, but with Nebraska's meteoric rise and Missouri's relative stability these two now occupy the expected podium positions.
• But that occupancy is not a foregone conclusion. The race for the podium has tightened up significantly. At the beginning of the season there was a 98% difference between #4 (52.1 expected points) and #10 (26.3 expected points). That is down to a 25% difference (48.8 vs 39.1).

Per usual, I am only looking at placement and advancement points based on the historical results from 2010 - 2022.

Edited February 21, 2023 by Wrestleknownothing

[Pinnacle]

WKN, is there any sense of how the top teams might fare with bonus? I know for Iowa it is not just Spencer Lee because Real Woods tilt game has been pretty good. 

[Wrestleknownothing]

29 minutes ago, Pinnacle said:

WKN, is there any sense of how the top teams might fare with bonus? I know for Iowa it is not just Spencer Lee because Real Woods tilt game has been pretty good. 

I only have bonus point data for 2013 thru 2022. During those 9 tournaments the first place team typically also leads in bonus points (8 out of 9 times).

Average bonus points by place looks like this:

 

Zooming in on the top 4 teams:

The exception year was 2018 when a third place Iowa outscored everyone else when it came to bonus points:

So the question is what kind of tournament do you expect? If it is the average tournament then PSU should make the gap bigger when factoring in bonus points. However, PSU has been below average of late, only averaging 16.25 bonus points the last two tourneys.

With PSU losing Nick Lee, but potentially gaining bonus points at 157 is that a push?

With Iowa seeing the return of an all-time tournament bonus point machine and Woods' upgraded offense, is that a net gain?

With Lee back I could see a path for Iowa to finish second, but have the most bonus points.

[BigRedFan]

42 minutes ago, Wrestleknownothing said:

I only have bonus point data for 2013 thru 2022. During those 9 tournaments the first place team typically also leads in bonus points (8 out of 9 times).

Average bonus points by place looks like this:

 

Zooming in on the top 4 teams:

The exception year was 2018 when a third place Iowa outscored everyone else when it came to bonus points:

So the question is what kind of tournament do you expect? If it is the average tournament then PSU should make the gap bigger when factoring in bonus points. However, PSU has been below average of late, only averaging 16.25 bonus points the last two tourneys.

With PSU losing Nick Lee, but potentially gaining bonus points at 157 is that a push?

With Iowa seeing the return of an all-time tournament bonus point machine and Woods' upgraded offense, is that a net gain?

With Lee back I could see a path for Iowa to finish second, but have the most bonus points.

Wouldn't it make more sense to determine historical bonus points by wrestler for a given seed, rather than by team placement?  #1 seeds historically get Z bonus points, #2 seeds get Y < Z points, etc.

[Wrestleknownothing]

1 minute ago, BigRedFan said:

Wouldn't it make more sense to determine historical bonus points by wrestler for a given seed, rather than by team placement?  #1 seeds historically get Z bonus points, #2 seeds get Y < Z points, etc.

Yes, but it is a lot more work.

[BigRedFan]

1 minute ago, Wrestleknownothing said:

Yes, but it is a lot more work.

You have time on your (fewer fingered) hands, don't you?

[Wrestleknownothing]

5 minutes ago, BigRedFan said:

You have time on your (fewer fingered) hands, don't you?

To quote the better half, "You have soooo many projects." For the less careful reader, that is not meant as a compliment.

[Wrestleknownothing]

55 minutes ago, BigRedFan said:

Wouldn't it make more sense to determine historical bonus points by wrestler for a given seed, rather than by team placement?  #1 seeds historically get Z bonus points, #2 seeds get Y < Z points, etc.

I did this once.

I wanted to see if more bonus points are scored on the championship side of the bracket or the consolation side. It breaks down to about 52% / 48%, so there is probably a modest correlation to seed. This is for 2010 - 2022, so a different time period than above.

[Pinnacle]

I was not even thinking historically, but rather which schools had known bonus scorers this year like Spencer Lee and Real Woods. PSU has a few guys but quantifying or projecting anything seems like guesswork. 

[Wrestleknownothing]

41 minutes ago, Pinnacle said:

I was not even thinking historically, but rather which schools had known bonus scorers this year like Spencer Lee and Real Woods. PSU has a few guys but quantifying or projecting anything seems like guesswork. 

Here is my pure guesswork for PSU and Iowa.

PSU had 15.5 bonus points last year. I think they lose a net of 3.5 at 141 (swapping Beau Bartlett for Nick Lee) and gain 2.5 among 149, 157 and 165,  with Levi Haines the most likely to score bonus. Everything else is a push. That gives them 14.5 bonus points.

Iowa had 9 bonus points last year. I think they gain 5.5 at 125, loses 3.5 (swapping Brody Teske for Austin DeSanto) at 133, gains 3.5 at 141, and loses 2 at 174 (swapping Nelson Brands for Michael Kemerer). Everything else is a push. That gives them 12.5 bonus points.

[flyingcement]

@Wrestleknownothingdid you ever live in the UK?  I've seen you drop all sorts of hints.  I'm pretty sure you mentioned Ladbrokes a few weeks back :)

[Wrestleknownothing]

16 minutes ago, flyingcement said:

@Wrestleknownothingdid you ever live in the UK?  I've seen you drop all sorts of hints.  I'm pretty sure you mentioned Ladbrokes a few weeks back

Just vacations and work trips.

But I am a big Liverpool fan. So if you hear me saying things like he gave him a Scouse kiss, it's twitchy bum time, he ate all the pies, he hit a purple patch, or they are at sixes and sevens, you will know where it came from.

[D3 for LU]

Mind the Gap, ol' chap!

D3

Edited February 22, 2023 by D3 for LU

[flyingcement]

2 minutes ago, Wrestleknownothing said:

Just vacations and work trips.

But I am a big Liverpool fan. So if you hear me saying things like he gave him a Scouse kiss, it's twitchy bum time, he ate all the pies, he hit a purple patch, or they are at sixes and sevens, you will know where it came from.

Lived there from age 24 through age 30 on an extended work assignment.  I had some flatmates in Kings Cross who were from Liverpool and introduced me to the scouse accent as well as the dish (a bit of an everything stew).  One thing UK is completely ignorant to in my experience is wrestling.  I joined an MMA gym while very out of shape and managed to do very well against those slobs.  

[Wrestleknownothing]

24 minutes ago, flyingcement said:

Lived there from age 24 through age 30 on an extended work assignment.  I had some flatmates in Kings Cross who were from Liverpool and introduced me to the scouse accent as well as the dish (a bit of an everything stew).  One thing UK is completely ignorant to in my experience is wrestling.  I joined an MMA gym while very out of shape and managed to do very well against those slobs.  

I am jealous.

[flyingcement]

1 minute ago, Wrestleknownothing said:

I am jealous.

If I told you a little more about my flatmates you wouldn't be 

[82bordeaux]

11 hours ago, Wrestleknownothing said:

Here is my pure guesswork for PSU and Iowa.

PSU had 15.5 bonus points last year. I think they lose a net of 3.5 at 141 (swapping Beau Bartlett for Nick Lee) and gain 2.5 among 149, 157 and 165,  with Levi Haines the most likely to score bonus. Everything else is a push. That gives them 14.5 bonus points.

Iowa had 9 bonus points last year. I think they gain 5.5 at 125, loses 3.5 (swapping Brody Teske for Austin DeSanto) at 133, gains 3.5 at 141, and loses 2 at 174 (swapping Nelson Brands for Michael Kemerer). Everything else is a push. That gives them 12.5 bonus points.

I would expect Kerkvliet to out bonus last year by a fair margin. He scored 2.5 last year.

[82bordeaux]

10 hours ago, flyingcement said:

If I told you a little more about my flatmates you wouldn't be 

[MPhillips]

12 hours ago, Wrestleknownothing said:

But I am a big Liverpool fan.

Same.

Nunez and Salah both score and Liverpool lose by three at Anfield. What a difference nine months make...

[Coach Boom]

WWK, I love this kind of analysis.  Thanks.  Can you tell us, where do you get all of you stats from?  I'd like to determine what is the expected score per wrestler by seed, e.g., 1st gets 16, plus 4 adv, plus bonus pts.  But I don't know where to find a listing of bonus points averages.     

[BerniePragle]

54 minutes ago, Coach Boom said:

WWK, I love this kind of analysis.  Thanks.  Can you tell us, where do you get all of you stats from?  I'd like to determine what is the expected score per wrestler by seed, e.g., 1st gets 16, plus 4 adv, plus bonus pts.  But I don't know where to find a listing of bonus points averages.     

 

[BerniePragle]

Just kidding.   WKN is amazing.  He not only is great at digging out and analyzing the data, he is the best I've ever seen at actually understanding data, drawing meaningful conclusions, and explaining it all to everyone. We're VERY  lucky to have him "onboard".

[Wrestleknownothing]

56 minutes ago, Coach Boom said:

WWK, I love this kind of analysis.  Thanks.  Can you tell us, where do you get all of you stats from?  I'd like to determine what is the expected score per wrestler by seed, e.g., 1st gets 16, plus 4 adv, plus bonus pts.  But I don't know where to find a listing of bonus points averages.     

Last year @BerniePragle and I spent time inputting the results for 2010 - 2022 by year, weight, seed, and school. Unfortunately I did not include bonus points. Doing bonus points by seed is a LOT more work than just doing it the way we did.

What we came up with was an expected advancement and placement points per seed rather than a score that reflects what they would get if they wrestled exactly to their seed. This recognizes the fact that seeds and results can be very different things. For example, from 2010 - 2022 a #1 seed's max points would be 20 plus bonus, but the actual average is 16.81 points. This reflects that the average finish for a #1 seed is between second and third.

It looks like this:

Now, if you are willing to do the same with bonus points by year, weight, seed, and school and share it would be greatly appreciated.

[Wrestleknownothing]

1 hour ago, MPhillips said:

Same.

Nunez and Salah both score and Liverpool lose by three at Anfield. What a difference nine months make...

It was hard to watch. Up 2-0 and lose 2-5. Oof.

[ionel]

10 minutes ago, BerniePragle said:

Just kidding.   WKN is amazing.  He not only is great at digging out and analyzing the data, he is the best I've ever seen at actually understanding data, drawing meaningful conclusions, and explaining it all to everyone. We're VERY  lucky to have him "onboard".

All true, well said!

But he does believe in the "pinfall" which is not a data supported term and he has a limited punctuation vocabulary.