Wrestle Stat gonna wrestle stat

nerfstate

Well-Known Member
Oct 10, 2017
8,018
19,392
1
Why do you even look at that site? Clearly the algorithms are out of whack. Don’t waste your time
Is there a better place to go to look at matchups, results and scoring? Sure, their rankings and predictions don’t always pass the sniff test, but it’s still a fantastic resource for fans.

hey, at least Nick has the highest P4P ELO, that’s pretty cool.
 

Sportfan2017

Well-Known Member
Gold Member
Jun 28, 2017
2,744
5,423
1
Why wouldn't Fix be ahead of RBY? Fix has only 2 losses in the last 2 full years while RBY has just been undefeated. Fix beat Desanto the last 3 times they have wrestled why RBY has defeated Desanto only 6 consecutive times. Finally in Head to Head matches Fix has only lost twice while RBY only has 2 wins and 2 NCAA Championships. You just don't understand.......it is based on simple math that is just too complicated for us to understand. :).
 

GogglesPaizano

Well-Known Member
Feb 6, 2018
3,169
4,827
1
I both critique and defend wrestlestat, but if you have multiple major ranking flaws at the end of a seasion, it raises more than an eyebrow. These are embarrassing.

I think they should shit can ELO period.

I Googled it long ago and there are better algorithms designed for individual head to head events like professional tennis.
 

johnstownsteel

Well-Known Member
Aug 2, 2013
6,318
13,838
1
Johnstown Pa
Why do you even look at that site? Clearly the algorithms are out of whack. Don’t waste your time
Do you stay away because of the rankings? I understand if you do but think you should check it out again sometime… browse around the site. I’d bet my life you’d love the rest of the site.
 

Sportfan2017

Well-Known Member
Gold Member
Jun 28, 2017
2,744
5,423
1
The algorithm needs to be adjusted to be less of a lifetime ELO and give more recent results more weight. I think the concept is simple, but the implementation is hard.
Yes......give more weight to recent matches and more weight to head to head matches. For instance...give the winner of a head to head matchup a multiplier such as a number between 1.5 & 2.0 or such. So if the winner of a head to head match is supposed to get 18 points......he now gets 27 or more ( or less ) depending on the relative ranking of the guy he beat head to head. Fix can NOT be ahead of RBY in any remotely sane rubric. It is NOT mathematically hard to shift the allotted percentages. Someone wrestling 2 years....Count the years 42%-58% or such to more heavily weigh the 2nd year. After 3 years....it could be weighted 25%-32%-43% or such. The other problem is starting every "New" wrestler so low that they have to win for YEARS to come even close to those with far more losses because the longer you are "In" their rubric.....the higher you go even if your record isn't as good as those chasing you.....even if they are winning more and defeating the higher ranked wrestler......even multiple times. Comments like " You have to give the wrestlers time.....if they keep winning they will eventually catch up" is just an admission that their Rubric is faulty.
That said......if anyone on FLO has a question about who lost and to whom....they are on Wrestlestat in 2 seconds to get all the info they need. For info and records.....they are the Gold Standard. They just need to bite the bullet and accept that their chess algorithm for lifetime achievement doesn't work for wrestlers limited to 4 years of countable matches. In chess.....the ratings are "Provisional" until you hit 20 matches. In wrestling during the Covid era......that could be 2 full seasons....or more.......especially if you are an ASU heavyweight :).
 
Last edited:

billrag

Well-Known Member
Jan 25, 2011
524
899
1
I also defend and criticize wrestlestat, but I'm smart enough and pragmatic enough to know what it is, what it's good for and what it's not good for. I do find it interesting that at each one of those weights, there's a considerable drop off between the top 2 (or the top 1 at HWT) and the rest of the field.
 

NothingMeaningful

Well-Known Member
Dec 21, 2018
464
511
1
I both critique and defend wrestlestat, but if you have multiple major ranking flaws at the end of a seasion, it raises more than an eyebrow. These are embarrassing.

I think they should shit can ELO period.

I Googled it long ago and there are better algorithms designed for individual head to head events like professional tennis.
Doesn't wrestlestats produce ratings, not rankings? There's a difference.
 
  • Like
Reactions: Cali_Nittany

Fink

Well-Known Member
Gold Member
Jul 30, 2001
188
206
1
Yes. I also thought I saw that they are going to re-examine them this offseason to adjust for recency and hth.
 

82bordeaux

Well-Known Member
Nov 19, 2019
3,881
12,110
1
58
eriespecialty.com
Doesn't wrestlestats produce ratings, not rankings? There's a difference.
Now I understand. So she


giphy.gif


could be RANKED higher but RATED lower than

giphy.gif


Why didn't you just say so?
 

pawrestlersintn

Well-Known Member
Jan 26, 2013
15,671
23,109
1
Yes. I also thought I saw that they are going to re-examine them this offseason to adjust for recency and hth.
But, you see, nationals ended 10 days ago. And, all of the people on here that say making changes to the algorithm is so easy, yet still somehow never do their own, hold Wrestlestat to a different standard.
 

jack66

Well-Known Member
May 29, 2001
7,378
10,856
1
That's the 1st I've seen that AM is staying with the Hawkeye Wrestling Club as a coach.

He will help assure the 'Iowa Style' is preserved well into the future.
 
  • Like
Reactions: slushhead

nitlion6

Well-Known Member
Mar 26, 2012
11,913
13,347
1
But, you see, nationals ended 10 days ago. And, all of the people on here that say making changes to the algorithm is so easy, yet still somehow never do their own, hold Wrestlestat to a different standard.
Just because someone has no interest in designing their own doesn't mean they aren't capable of noticing mistakes / or issues with one in existence. Pointing that nobody seems interested in designing their own doesn't dismiss the reality of the existing quirks.
 

SpikingWhamos

Well-Known Member
Aug 26, 2015
349
534
1
Felt like digging in. Ignore any rounding issues in the math, and also any "me being an idiot" issues in mapping rankings histories to specific weeks, I think I have it directionally correct

In general, very similar climbs and results between RBY and Fix. Both have taken relatively elite losses that wouldn't great impact their score. The difference is some bad losses for RBY

In 2019, Fix's first real ranking was a 1468, which counts some scoring from his redshirt year. By the end of that year, he had lost to Cole Matthews and Suriano, neither of which brought his score down much. Cole Matthews week dropped it by ~11, and he went up overall with his NCAA wins more than offsetting the Suriano loss.

I'll use 1404 as RBY's starting point, with no redshirt season. He took some Fix-equivalent "good" ELO losses to Desanto / Erneste / Lizak as well. Those losses didn't hurt his score much, but also didn't give him the equivalent gains that Fix got by beating Erneste (by major) and Desanto

RBY on the other hand had a bunch of bad ELO losses that year. The losses to Gomez and Thornton were both significant ELO setbacks. If I'm lining weeks up right, the scuffle including the Gomez loss cost him 9 points, and the Thornton loss cost him about 16.

Total 2019 gain for RBY = 94
Total 2019 gain for Fix = 100

So end of 2019:
Fix = 1567
RBY = 1497 (started 64 points down, lost 6 in results)

Seems valid given the results at the time. ELO doesn't really have a "head to head" or "common opponent" adjustment, it's just giving points based on the equivalent rankings at the time.

2020: Fix doesn't wrestle, stays at 1567
2020: RBY gains ~49 points

2021: In 2021, RBY made up about 10 points overall. That all feels generally reasonable given the two seasons. Both undefeated, Fix with only one "good loss" to #2, RBY gets a reasonable gain from beating #1.

2022: RBY makes up about 7 points overall. Very similar to the year before!

So we end up with RBY behind by about 5 points overall in the rankings. The difference there is really the 2019 season where RBY had his overall rankings growth hampered by two losses.

But if you trend it, you clearly see RBY making up 66 out of the 70 points starting in 2020.

Doubt this will change anyone's thinking either way!
 
  • Like
Reactions: Reslo

STAND with PRIDE

Well-Known Member
Sep 9, 2018
4,538
6,592
1
Felt like digging in. Ignore any rounding issues in the math, and also any "me being an idiot" issues in mapping rankings histories to specific weeks, I think I have it directionally correct

In general, very similar climbs and results between RBY and Fix. Both have taken relatively elite losses that wouldn't great impact their score. The difference is some bad losses for RBY

In 2019, Fix's first real ranking was a 1468, which counts some scoring from his redshirt year. By the end of that year, he had lost to Cole Matthews and Suriano, neither of which brought his score down much. Cole Matthews week dropped it by ~11, and he went up overall with his NCAA wins more than offsetting the Suriano loss.

I'll use 1404 as RBY's starting point, with no redshirt season. He took some Fix-equivalent "good" ELO losses to Desanto / Erneste / Lizak as well. Those losses didn't hurt his score much, but also didn't give him the equivalent gains that Fix got by beating Erneste (by major) and Desanto

RBY on the other hand had a bunch of bad ELO losses that year. The losses to Gomez and Thornton were both significant ELO setbacks. If I'm lining weeks up right, the scuffle including the Gomez loss cost him 9 points, and the Thornton loss cost him about 16.

Total 2019 gain for RBY = 94
Total 2019 gain for Fix = 100

So end of 2019:
Fix = 1567
RBY = 1497 (started 64 points down, lost 6 in results)

Seems valid given the results at the time. ELO doesn't really have a "head to head" or "common opponent" adjustment, it's just giving points based on the equivalent rankings at the time.

2020: Fix doesn't wrestle, stays at 1567
2020: RBY gains ~49 points

2021: In 2021, RBY made up about 10 points overall. That all feels generally reasonable given the two seasons. Both undefeated, Fix with only one "good loss" to #2, RBY gets a reasonable gain from beating #1.

2022: RBY makes up about 7 points overall. Very similar to the year before!

So we end up with RBY behind by about 5 points overall in the rankings. The difference there is really the 2019 season where RBY had his overall rankings growth hampered by two losses.

But if you trend it, you clearly see RBY making up 66 out of the 70 points starting in 2020.

Doubt this will change anyone's thinking either way!
The only numbers that should matter.
RBY-2
Fix-0
This isnt hard.
 

pawrestlersintn

Well-Known Member
Jan 26, 2013
15,671
23,109
1
Just because someone has no interest in designing their own doesn't mean they aren't capable of noticing mistakes / or issues with one in existence. Pointing that nobody seems interested in designing their own doesn't dismiss the reality of the existing quirks.
It's been pointed, and whined about, to death. Everyone, and I mean everyone, knows the issues. But, here we go again.
 

PSU Mike

Well-Known Member
Gold Member
Jul 29, 2001
14,704
7,353
1
Glencoe, IL
Why wouldn't Fix be ahead of RBY? Fix has only 2 losses in the last 2 full years while RBY has just been undefeated. Fix beat Desanto the last 3 times they have wrestled why RBY has defeated Desanto only 6 consecutive times. Finally in Head to Head matches Fix has only lost twice while RBY only has 2 wins and 2 NCAA Championships. You just don't understand.......it is based on simple math that is just too complicated for us to understand. :).
In the last two 133 finals Fix finished 2nd and RBY finished next to last. Makes total sense.
 
  • Like
Reactions: PASLP2