Iowa Separating From The Pack (Intermat 1/14/25 rankings)

[wrestleknownothing]

• Iowa just keeps grinding higher. Credit Kyle Parco's move from #4 to #3 this week and Joey Cruz's return to the rankings. Next stop Penn State (not shown but expected points = 139.6)
• Oklahoma State took a breather this week. While Reece Witcraft, Carter Young, and Caleb Fish saw incremental rankings gains, Troy Spratley dropped from #2 to #4, more than offsetting the gains at 133, 149, and 157.
• The race for the final podium position remains tight. Minnesota, Nebraska, Ohio State, Virginia Tech, and Northern Iowa are within 14.8 points of each other. Even Cornell and Illinois have a punchers chance.

 

[Iowawrestling03]

Great stuff! Tableau graph? I haven't followed the rankings but it looked like we are performing better than usual as a team. Could you share a graph with PSU so we can observe how they are trending? Would be cool to see other rankers like Flo, wrestlestat, etc and include actual end of season result for the last data point.

BTW - how many points is Cruz expected to get?

 

[MSU158]

Great stuff! Tableau graph? I haven't followed the rankings but it looked like we are performing better than usual as a team. Could you share a graph with PSU so we can observe how they are trending? Would be cool to see other rankers like Flo, wrestlestat, etc and include actual end of season result for the last data point.

BTW - how many points is Cruz expected to get?

With a ranking in the high 20's he shouldn't be scoring any....

 

[wrestleknownothing]

Great stuff! Tableau graph? I haven't followed the rankings but it looked like we are performing better than usual as a team. Could you share a graph with PSU so we can observe how they are trending? Would be cool to see other rankers like Flo, wrestlestat, etc and include actual end of season result for the last data point.

BTW - how many points is Cruz expected to get?

Yes, Tableau.

For trends I tend to stick with one service as the graph gets too noisy with multiple dates and services for a single team.

Here is a snapshot across services for the most recent set of rankings.



And here is the expected scoring breakdown by wrestler, by service, along with the rank and the probability of achieving AA (all based on current rank).



Summing across all ten wrestlers gets me an expected 6.1 AAs based on Intermat rankings for Iowa. The distribution of outcomes looks like this:

 

[wrestleknownothing]

With a ranking in the high 20's he shouldn't be scoring any....

There are still advancement points to be had for high 20's wrestlers.

 

[MSU158]

There are still advancement points to be had for high 20's wrestlers.

How? In a standard 32 man bracket, 25-32 should lose to 1-8, respectively. Then, in the consolation, 25-32 should lose to 17-24, respectively. The points may be "there" to be had, but they shouldn't be assumed in a tournament ranking format...

 

[wrestleknownothing]

How? In a standard 32 man bracket, 25-32 should lose to 1-8, respectively. Then, in the consolation, 25-32 should lose to 17-24, respectively. The points may be "there" to be had, but they shouldn't be assumed in a tournament ranking format...

Au contraire. That is very binary thinking and assumes things always go chalk. But things never go chalk.

The 28 seed has gone 0-2 only 64% of the time in the 33 seed era. The other 36% of the time they score advancement and even bonus points. Heck there has even been a 28 seed finish 8th.

It is also a mistake to assume the #1 seed will score 20+ points when the #1 seed only wins 48% of the time.

Giving too much credit to top 8 seeds and not enough credit to 9-33 seeds is the mistake Intermat, Wrestlestat, Flo all make when assembling their tournament rankings. If you look at the way WrestleStat does it they even go so far as to assume 0 points for the 32 and 33 wrestlers when they will definitionally score a minimum of 1 point between them.

 

[johnyt9]

does the 32 seed get advancement for beating the 33?

 

[wrestleknownothing]

does the 32 seed get advancement for beating the 33?

Yes. Whoever wins the pigtail on the championship side gets 1 advancement point. Winner of the consolation pigtail gets 0.5 advancement points. They also have bonus opportunities in those matches.

And the 33 has done slightly better on average than the 32.

 

[MSU158]

Au contraire. That is very binary thinking and assumes things always go chalk. But things never go chalk.

The 28 seed has gone 0-2 only 64% of the time in the 33 seed era. The other 36% of the time they score advancement and even bonus points. Heck there has even been a 28 seed finish 8th.

It is also a mistake to assume the #1 seed will score 20+ points when the #1 seed only wins 48% of the time.

Giving too much credit to top 8 seeds and not enough credit to 9-33 seeds is the mistake Intermat, Wrestlestat, Flo all make when assembling their tournament rankings. If you look at the way WrestleStat does it they even go so far as to assume 0 points for the 32 and 33 wrestlers when they will definitionally score a minimum of 1 point between them.

So, you AUTOMATICALLY assume a 1 point minimum for EVERY wrestler that qualified in the top 32?

Look, 64% going 0-2 is still A LOT. I get that you want to use statistics to manipulate a standard system, but to me, that adjusted system convolutes everything. It isn't perfect to assume all goes to chalk, but it takes the "calculator" out of the mix and goes 100% by the rankings, which is what a ranking system is supposed to be...

 

[12tdsaw]

god you guys suck the life out of wrestling

 

[DropitlikeitsHawk]

Penn State winning it though right? Let’s go get some individuals.

 

[Coach Mac B]

Au contraire. That is very binary thinking and assumes things always go chalk. But things never go chalk.

The 28 seed has gone 0-2 only 64% of the time in the 33 seed era. The other 36% of the time they score advancement and even bonus points. Heck there has even been a 28 seed finish 8th.

Technically, yes you are correct. However, I think most people understand the assumption that the team scores listed are assuming the individuals perform to "chalk". If the individual rankings play out, the team result is this. Anything else gets difficult - like explaining how a team's only wrestler wins it all, but only scores the equivalent of a 3rd place participant.

If you want to assign 0.34 points to the 28 seed because 36% of the time they won't go 0-2, then I guess you could do it. But you're splitting hairs between a #24 ranked wrestler and a #28 one. Pretty small margin of error for those than rank that far down.

If you want to run an "expected value" ranking, I don't think the end result would be significantly more valuable and comparable to the end results.

Don't make the estimated team score results more technical than the process behind the individual rankings themselves.

 

[Chickenman Testa]

 

[wrestleknownothing]

So, you AUTOMATICALLY assume a 1 point minimum for EVERY wrestler that qualified in the top 32?

Look, 64% going 0-2 is still A LOT. I get that you want to use statistics to manipulate a standard system, but to me, that adjusted system convolutes everything. It isn't perfect to assume all goes to chalk, but it takes the "calculator" out of the mix and goes 100% by the rankings, which is what a ranking system is supposed to be...

You have jumped to a conclusion that is not supported by anything I said. I do not automatically do anything. All of these data points are based on empirical results.

I am not using statistics to manipulate anything. I am using statistics to explain how the real world works. Going 100% by the rankings (or seeds - I am using those terms interchangeable until seeds are available) is to assume away reality. I choose not to do that. If you want to assume away reality, that is certainly your prerogative.

What the ranking/seeding system is supposed to be and what it is are two different things. This is why the tournament gets wrestled to conclusion.

 

[wrestleknownothing]

Technically, yes you are correct. However, I think most people understand the assumption that the team scores listed are assuming the individuals perform to "chalk". If the individual rankings play out, the team result is this. Anything else gets difficult - like explaining how a team's only wrestler wins it all, but only scores the equivalent of a 3rd place participant.

If you want to assign 0.34 points to the 28 seed because 36% of the time they won't go 0-2, then I guess you could do it. But you're splitting hairs between a #24 ranked wrestler and a #28 one. Pretty small margin of error for those than rank that far down.

If you want to run an "expected value" ranking, I don't think the end result would be significantly more valuable and comparable to the end results.

Don't make the estimated team score results more technical than the process behind the individual rankings themselves.

Yes, it is difficult, but I have made it easy for you by doing all the heavy lifting.

I am making the estimated team score reflect reality. What the ranking services do with their individual rankings to come up with tournament team scores is what is unrealistic.

 

[wrestleknownothing]

Amen

 

[wrestleknownothing]

god you guys suck the life out of wrestling

You will probably want to avoid 100% of all threads I start

 

[MSU158]

You have jumped to a conclusion that is not supported by anything I said. I do not automatically do anything. All of these data points are based on empirical results.

I am not using statistics to manipulate anything. I am using statistics to explain how the real world works. Going 100% by the rankings (or seeds - I am using those terms interchangeable until seeds are available) is to assume away reality. I choose not to do that. If you want to assume away reality, that is certainly your prerogative.

What the ranking/seeding system is supposed to be and what it is are two different things. This is why the tournament gets wrestled to conclusion.

Not what I meant. I get what your system does. I just fundamentally disagree with it. To me, assigning zero expected points to positions seeded to lose is a more accurate "ranking system" than the one you are using. That doesn't mean I don't appreciate all the hard work you put into it, or even the data it provides.

If you want to classify it as a predictor based on historical data, I am all for it. However, calling it a tournament ranking system, just isn't accurate in my opinion...

 

[so cal hawkfan]

You will probably want to avoid 100% of all threads I start

Im going to assume he didn’t mean you. Likely referring to another particularly long winded poster……….

 

[wrestleknownothing]

Not what I meant. I get what your system does. I just fundamentally disagree with it. To me, assigning zero expected points to positions seeded to lose is a more accurate "ranking system" than the one you are using. That doesn't mean I don't appreciate all the hard work you put into it, or even the data it provides.

If you want to classify it as a predictor based on historical data, I am all for it. However, calling it a tournament ranking system, just isn't accurate in my opinion...

I did not call it a tournament ranking system. I refer to it very plainly as expected points. I even provide distributions of possible outcomes for AA's, for example, without getting married to one.

Now if you can explain to me the difference between a ranking system for the tournament and a predictor given that the final rank is based on the team scores, I am all ears.

 

[HawkFan1986]

Not what I meant. I get what your system does. I just fundamentally disagree with it. To me, assigning zero expected points to positions seeded to lose is a more accurate "ranking system" than the one you are using. That doesn't mean I don't appreciate all the hard work you put into it, or even the data it provides.

If you want to classify it as a predictor based on historical data, I am all for it. However, calling it a tournament ranking system, just isn't accurate in my opinion...

You’re insufferable with your consistent smartest guy in the room act and needing the last word as though the world is waiting with bated breath to hear your coy remark.

I bet when your wife asks you for a bj you critize the way she asks.

The guy you keep nitpicking is providing a FREE service using analytics and data. The proper thing to do is look at the data, say thanks, and move on. Knowing that nearly all projected statistics have a margin of error. But nope, you’re nitpicking one damn point that Joey Cruz may or may not score. Give it a rest man.

 

[MSU158]

You’re insufferable with your consistent smartest guy in the room act and needing the last word as though the world is waiting with bated breath to hear your coy remark.

I bet when your wife asks you for a bj you critize the way she asks.

The guy you keep nitpicking is providing a FREE service using analytics and data. The proper thing to do is look at the data, say thanks, and move on. Knowing that nearly all projected statistics have a margin of error. But nope, you’re nitpicking one damn point that Joey Cruz may or may not score. Give it a rest man.

Thank you for your stellar contribution. Now, go in peace, my son....

 

[acuhawk]

You’re insufferable with your consistent smartest guy in the room act and needing the last word as though the world is waiting with bated breath to hear your coy remark.

I bet when your wife asks you for a bj you critize the way she asks.

The guy you keep nitpicking is providing a FREE service using analytics and data. The proper thing to do is look at the data, say thanks, and move on. Knowing that nearly all projected statistics have a margin of error. But nope, you’re nitpicking one damn point that Joey Cruz may or may not score. Give it a rest man.

It's not hard to ignore him he obviously likes to debate people so either debate him or ignore him. I don't get the anguish. It's s not like he forces anyone to read his posts. Just skip over the arguments or participate if you have a slow day. I know some guys post on here and have fun stirring crap up when they have a slow day

 

[vhsalum]

It's not hard to ignore him he obviously likes to debate people so either debate him or ignore him. I don't get the anguish. It's s not like he forces anyone to read his posts. Just skip over the arguments or participate if you have a slow day. I know some guys post on here and have fun stirring crap up when they have a slow day

I like him (and any of us) getting called out. Sure, you can ignore him. But he's DEFINITELY being a prick arguing semantics over a graph he could never and will never provide/create himself.

It's just weird and loser-y. Like roasting a guy who brings in donuts for the team, but there's only one glaze in the box.

 

[MSU158]

I did not call it a tournament ranking system. I refer to it very plainly as expected points. I even provide distributions of possible outcomes for AA's, for example, without getting married to one.

Now if you can explain to me the difference between a ranking system for the tournament and a predictor given that the final rank is based on the team scores, I am all ears.

I am sorry if I misunderstood. Again, I DO appreciate the work you put into these and the data it provides. I just was admittedly very surprised that ANY system expects/assigns points to seed slotted not to win a match.

I didn't mean to "poop" on your system. I just think "historical" data doesn't really apply when rankings assume the people are all correctly slotted and therefore all seeds are correctly slotted and represent their respective, expected outcomes until the matches are actually wrestled.

 

[MSU158]

I like him (and any of us) getting called out. Sure, you can ignore him. But he's DEFINITELY being a prick arguing semantics over a graph he could never and will never provide/create himself.

It's just weird and loser-y. Like roasting a guy who brings in donuts for the team, but there's only one glaze in the box.

In this case, I wasn't meaning to be a prick. I just didn't understand any system that assigns points to those most likely to go 0-2.

As far as weird and loser-y goes, which one are you? The guy that wrestled for Iowa or the one that keeps DT's nuts warm?

 

[12tdsaw]

You will probably want to avoid 100% of all threads I start

you do you no nothing just not my bag man. plenty of room here for both of us

 

[wrestleknownothing]

I am sorry if I misunderstood. Again, I DO appreciate the work you put into these and the data it provides. I just was admittedly very surprised that ANY system expects/assigns points to seed slotted not to win a match.

I didn't mean to "poop" on your system. I just think "historical" data doesn't really apply when rankings assume the people are all correctly slotted and therefore all seeds are correctly slotted and represent their respective, expected outcomes until the matches are actually wrestled.

One of the most popular threads every year has to do with how the committee got my guy's seed wrong (too low) or their guy's seed wrong (too high). Look at Carter Starocci last year as a #9 seed (my guy is too low says PSU), or Braeden Davis as a #1 seed (their guy is too high says everyone else).

I do not agree that it is safe to assume the people are all slotted correctly. And history bears that out.

• Since they started seeding 6 or more wrestlers (1963) there have been 614 total brackets and only 16 of those (2.6%) have had all the guys seeded in the AA spots all finish as AAs.
• In the 16 seed era it happened once (2%),
• and in the 33 seed era it happened once (2%).
• And none of those resulted in the podium in the same order as the seeds.
• The closest any weight has ever come was 1,2,3,4, but then it breaks down That has only happened twice.
• It just never happens.

 

[MSU158]

One of the most popular threads every year has to do with how the committee got my guy's seed wrong (too low) or their guy's seed wrong (too high). Look at Carter Starocci last year as a #9 seed (my guy is too low says PSU), or Braeden Davis as a #1 seed (their guy is too high says everyone else).

I do not agree that it is safe to assume the people are all slotted correctly. And history bears that out.

• Since they started seeding 6 or more wrestlers (1963) there have been 614 total brackets and only 16 of those (2.6%) have had all the guys seeded in the AA spots all finish as AAs.
• In the 16 seed era it happened once (2%),
• and in the 33 seed era it happened once (2%).
• And none of those resulted in the podium in the same order as the seeds.
• The closest any weight has ever come was 1,2,3,4, but then it breaks down That has only happened twice.
• It just never happens.

All good, I respect what you do and have found you to be a very level minded poster. I get where you are coming from, I just don't like historical data that only looks at numbers without looking at the guy that is in that slot. Mind you a #1 seed Starocci is much more likely to finish there than a #1 seed Davis.

I would be happy to revisit this after NCAA's this year and see where the 2 match up(obviously not including bonus points). Again, I do want to be clear that I respect and appreciate the work and data your provide, even if there are some areas I disagree with...

 

[wrestleknownothing]

All good, I respect what you do and have found you to be a very level minded poster. I get where you are coming from, I just don't like historical data that only looks at numbers without looking at the guy that is in that slot. Mind you a #1 seed Starocci is much more likely to finish there than a #1 seed Davis.

I would be happy to revisit this after NCAA's this year and see where the 2 match up(obviously not including bonus points). Again, I do want to be clear that I respect and appreciate the work and data your provide, even if there are some areas I disagree with...

I do not mind arguing......especially when I know I am right.

 

[el dub]

Au contraire. That is very binary thinking and assumes things always go chalk. But things never go chalk.

The 28 seed has gone 0-2 only 64% of the time in the 33 seed era. The other 36% of the time they score advancement and even bonus points. Heck there has even been a 28 seed finish 8th.

It is also a mistake to assume the #1 seed will score 20+ points when the #1 seed only wins 48% of the time.

Giving too much credit to top 8 seeds and not enough credit to 9-33 seeds is the mistake Intermat, Wrestlestat, Flo all make when assembling their tournament rankings. If you look at the way WrestleStat does it they even go so far as to assume 0 points for the 32 and 33 wrestlers when they will definitionally score a minimum of 1 point between them.

Only 64% of the time? Vegas likes those odds, I’m guessing.

 

[wrestleknownothing]

Only 64% of the time? Vegas likes those odds, I’m guessing.

I have wondered if there was anywhere you could bet on NCAA outcomes. I would be very curious to see how efficient those markets are.

 

[Trapper85]

I didn't mean to "poop" on your system. I just think "historical" data doesn't really apply when rankings assume the people are all correctly slotted and therefore all seeds are correctly slotted and represent their respective, expected outcomes until the matches are actually wrestled.

All "data" is "historical". This is forecasting of the future. Using 'averages' from the past will....on average... provide more accurate predictors than a single number (i.e. current seed by some ranking system). But the devil is also in the modeling details (e.g. how far back do you consider performance of 28 seeds, should you tweak the historical average by say the record of this year's 28 seed). I think we need to start a Cesspool modeling forum, maybe call it the 83% Forum. More politics-free wrestling talk might occur there, although math nerds can get a little testy at times.

 

[wrestleknownothing]

All "data" is "historical". This is forecasting of the future. Using 'averages' from the past will....on average... provide more accurate predictors than a single number (i.e. current seed by some ranking system). But the devil is also in the modeling details (e.g. how far back do you consider performance of 28 seeds, should you tweak the historical average by say the record of this year's 28 seed). I think we need to start a Cesspool modeling forum, maybe call it the 83% Forum. More politics-free wrestling talk might occur there, although math nerds can get a little testy at times.

This does not answer your question, but is loosely related in that it talks about how much data there is at different times for different seed ranges.

I did this about a month ago, but had forgotten about it until someone reminded me:

Do Seeds Matter?
Don't put too much stock in this as it only covers two 5-year periods (so call it weak evidence), but it occurred to me last night that we have a bit of a natural experiment.

The 33 seed era is now 5 years old and the 16 seed era only lasted 5 years. So I thought I would take a look at what the expansion from 16 seeds to 33 seeds has meant.

In theory, if seeding is accurate, then going from 16 to 33 would have different impacts on different ranges of seeds. Concentrating on the first round:

• In a 33 seed bracket the top 8 seeds face seeds 25 - 32 (or 33). In a 16 seed bracket the top 8 seeds may get anyone in the 17 - 33 range. So seeding all should be beneficial to the top 8.
• In a 33 seed bracket the 9 - 16 seeds face seeds 17 - 24. In a 16 seed bracket the 9 - 16 seeds may get anyone in the 17 - 33 range. So seeding all would be harmful to the 9 - 16 seeds.
• In a 33 seed bracket the bottom 16/17 seeds face the top 16 seeds. In a 16 seed bracket the bottom 17 seeds may get anyone in the 1 - 33 range (other than themselves). So seeding all would be harmful to the bottom 17.

And that is what the results show.

• The top 8 seeds have performed better in the 33 seed era,
• while the next 8 seeds have performed worse,
• and the bottom 17 have performed worse.

So, good job seeding committee.

 

[Sdwrestler_20]

"You can't argue with the numbers"
MSU "Hold my beer..."

 

[wrestleknownothing]

All good, I respect what you do and have found you to be a very level minded poster. I get where you are coming from, I just don't like historical data that only looks at numbers without looking at the guy that is in that slot. Mind you a #1 seed Starocci is much more likely to finish there than a #1 seed Davis.

I would be happy to revisit this after NCAA's this year and see where the 2 match up(obviously not including bonus points). Again, I do want to be clear that I respect and appreciate the work and data your provide, even if there are some areas I disagree with...

Just realized we can review now based on last year.

In general my predictions were pretty good. I was a little low on PSU. They are the only team I made an individual adjustments because of their propensity to nearly match their seed when highly seeded. But they slightly outperformed their seeds last year, so I was a few points light.

My two big whiffs were NC State and Lehigh. Now, NC State may have been the biggest crapping of the bed in tournament history. I do not have every year's relative performance, so hard to be certain. They are lucky they did that the year PSU was so dominant that no one noticed.

This table doesn't include the PSU adjustment, just the raw number based on field averages.

My Ten Worst

 

[MSU158]

"You can't argue with the numbers"
MSU "Hold my beer..."

Except I wasn’t arguing against “the numbers”. I was arguing against a system that assigns advancement points to the group most likely to go 0-2. But, kudos to you for re-running the same joke!

 

[Sharky1203]

I never had one single medal hung around my neck for a ranking. When Nats come, step up and go for it. That's where it means something. Otherwise it's just opinions, and opinions vary.

 

[wrestleknownothing]

I never had one single medal hung around my neck for a ranking. When Nats come, step up and go for it. That's where it means something. Otherwise it's just opinions, and opinions vary.

As does mileage