ADVERTISEMENT

WrestleStat Rankings | Week 9 2022

andegre

Well-Known Member
May 18, 2004
4,185
906
1
Here's our latest rankings...we have a new #1 in the tournament rankings this week with the Spencer Lee (Iowa) news, as well as still not having Michael Kemmerer in the lineup yet.

Check out the rest of the rankings below.

Wrestler Rankings | https://www.wrestlestat.com/rankings/starters

Dual Rankings | https://www.wrestlestat.com/rankings/dual

Tournament Rankings | https://www.wrestlestat.com/rankings/tournament

RPI Rankings | https://www.wrestlestat.com/rankings/rpi

Stat Leaders | https://www.wrestlestat.com/rankings/statistical
 
Here's our latest rankings...we have a new #1 in the tournament rankings this week with the Spencer Lee (Iowa) news, as well as still not having Michael Kemmerer in the lineup yet.

Check out the rest of the rankings below.

Wrestler Rankings | https://www.wrestlestat.com/rankings/starters

Dual Rankings | https://www.wrestlestat.com/rankings/dual

Tournament Rankings | https://www.wrestlestat.com/rankings/tournament

RPI Rankings | https://www.wrestlestat.com/rankings/rpi

Stat Leaders | https://www.wrestlestat.com/rankings/statistical
I don’t want to give you too hard a time, and there are several I question, but how in the world does your algorithm have Jordan Wood ahead of Kerk?

Edit: and 4-3 Matt Stencil ahead of both of them? Stencil has no top 20 wins and has lost twice to Wood. Yet ranks above Wood (and Kerk) at #5?
 
I don’t want to give you too hard a time, and there are several I question, but how in the world does your algorithm have Jordan Wood ahead of Kerk?

Edit: and 4-3 Matt Stencil ahead of both of them? Stencil has no top 20 wins and has lost twice to Wood. Yet ranks above Wood (and Kerk) at #5?

ELO chess needs to be retired in preference for an algorithm that adds weight more recent matches. Hildy (vs Ayala), RBY, and Dean are all projected to lose vs Iowa. Nick Lee just recently was corrected. As far as I am concerned matches more than 18 months old should be excluded entirely. Maybe then we will see a higher fidelity model. That said I still love wrestlestat, and the service it provides.
 
I don’t want to give you too hard a time, and there are several I question, but how in the world does your algorithm have Jordan Wood ahead of Kerk?

Edit: and 4-3 Matt Stencil ahead of both of them? Stencil has no top 20 wins and has lost twice to Wood. Yet ranks above Wood (and Kerk) at #5?
How about the algorithm rocketing Ayala up to #7.
 
Edit: and 4-3 Matt Stencil ahead of both of them? Stencil has no top 20 wins and has lost twice to Wood. Yet ranks above Wood (and Kerk) at #5?
Hehe, head to head prediction has Wood beating Stencel 5-4
 
  • Like
Reactions: ccdiver
Here's our latest rankings...we have a new #1 in the tournament rankings this week with the Spencer Lee (Iowa) news, as well as still not having Michael Kemmerer in the lineup yet.

Check out the rest of the rankings below.

Wrestler Rankings | https://www.wrestlestat.com/rankings/starters

Dual Rankings | https://www.wrestlestat.com/rankings/dual

Tournament Rankings | https://www.wrestlestat.com/rankings/tournament

RPI Rankings | https://www.wrestlestat.com/rankings/rpi

Stat Leaders | https://www.wrestlestat.com/rankings/statistical
Btw, thank you for a wonderful wonderful site. I’m in it daily. One of my go to sites. You guys did a great job putting it together.
 
I don’t want to give you too hard a time, and there are several I question, but how in the world does your algorithm have Jordan Wood ahead of Kerk?

Edit: and 4-3 Matt Stencil ahead of both of them? Stencil has no top 20 wins and has lost twice to Wood. Yet ranks above Wood (and Kerk) at #5?
Bear in mind that the ELO algorithm, in its raw form is much more of a "career ranking," than an in-season, week to week ranking system. Stencil has 150 career matches, Wood has 120 and Kerk has 25. With those number of matches for the other two guys, their ELO score doesn't change a lot. For instance, Stencil's 4-3 record this year has moved his score from 1502 to 1494. Kerk, however has moved up 15 points with his 6 wins.

Also, the scores between the three guys are very close. Stencil 1494, Kerk 1479 and Wood 1491. Since Wood and Stencil's scores don't change much, win or lose, Kerk will overtake them soon with a few more wins.
 
How about the algorithm rocketing Ayala up to #7.
Ayala has 12 matches, and is 10-2, with some wins over higher ranked (high ELO score) guys. Those good wins mean a lot to an ELO score that is only built on a limited number of matches. Also, a low ELO guy beating a guy he is not "supposed" to beat (a high ELO guy), causes a big jump in ELO score. Finally, his losses to McKee don't hurt him much, because he is "supposed" to lose to him.

One of the tweaks that Andegre did work on was to tweak the early career calcs, so that a new wrestler's scores would jump quicker than it would in the chess world, due to the limited number of matches a wrestler gets, versus the number of matches a chess player gets.
 
Bear in mind that the ELO algorithm, in its raw form is much more of a "career ranking," than an in-season, week to week ranking system. Stencil has 150 career matches, Wood has 120 and Kerk has 25. With those number of matches for the other two guys, their ELO score doesn't change a lot. For instance, Stencil's 4-3 record this year has moved his score from 1502 to 1494. Kerk, however has moved up 15 points with his 6 wins.

Also, the scores between the three guys are very close. Stencil 1494, Kerk 1479 and Wood 1491. Since Wood and Stencil's scores don't change much, win or lose, Kerk will overtake them soon with a few more wins.
Thanks; I did not know or take the time to understand that.
 
  • Like
Reactions: hlstone
Thanks; I did not know or take the time to understand that.
No problem. You and hundreds of others. LOL.

An interesting exercise is to pick a particular wrestler, then go to their rank history. And, you can look at their ELO score overtime. Couple that with their results during their freshmen year, and you can see the effect that a good win or a bad loss had on that ELO number. Just bear in mind that they only run the calculations once per week, not after each match.
 
Bear in mind that the ELO algorithm, in its raw form is much more of a "career ranking," than an in-season, week to week ranking system. Stencil has 150 career matches, Wood has 120 and Kerk has 25. With those number of matches for the other two guys, their ELO score doesn't change a lot. For instance, Stencil's 4-3 record this year has moved his score from 1502 to 1494. Kerk, however has moved up 15 points with his 6 wins.

Also, the scores between the three guys are very close. Stencil 1494, Kerk 1479 and Wood 1491. Since Wood and Stencil's scores don't change much, win or lose, Kerk will overtake them soon with a few more wins.
I would like to know how career matches applies. So if he lost all 150 what then. I'm not saying it shouldn't apply I just want to know how it applies to this f d up algorithm they use. Once again if your algorithm is broke. Fix it
 
  • Like
Reactions: ccdiver
I would like to know how career matches applies. So if he lost all 150 what then. I'm not saying it shouldn't apply I just want to know how it applies to this f d up algorithm they use. Once again if your algorithm is broke. Fix it
Or, alternatively, if you don't think the algorithm/results are worthwhile, you could simply not look at it.
 
I'm wondering how a change in weight class is factored into the algorithm. Lewis and Hidlay are both moving up and are already ranked above Carter. Their resume was built at 165 and 157 respectively.

Also, Max Dean has ascended the rankings pretty quickly even though he has faced only one top-10 wrestler at 197.

I'm not complaining ... love the site ... just curious.
 
Or, alternatively, if you don't think the algorithm/results are worthwhile, you could simply not look at it.
True but I would rather bitch about it lmfao and was only adding to another response. I look at all the rankings and initially think what BS then I realize there's some algorithm that doesn't take into account anything important and this is for fun and then I let it go.
 
  • Like
Reactions: ccdiver
Or, alternatively, if you don't think the algorithm/results are worthwhile, you could simply not look at it.
Just out of curiosity and I mean that. I'm not trying to be offensive. But what exactly about the algorithm is worthwhile. It seems everyone excepts the algorithm is flawed why shouldn't we say something so whoever came up with it makes changes. I get the argument that not everyone does rankings so having rankings is nice and it gives us something to talk about. My argument relies on the notion that if something is broken you fix it otherwise it never gets better or there is never any improvement in any event I'm sure it's complicated as most algorithms are so I just except they are what they are and I'll keep bitching about it lmfao.
 
Just out of curiosity and I mean that. I'm not trying to be offensive. But what exactly about the algorithm is worthwhile. It seems everyone excepts the algorithm is flawed why shouldn't we say something so whoever came up with it makes changes. I get the argument that not everyone does rankings so having rankings is nice and it gives us something to talk about. My argument relies on the notion that if something is broken you fix it otherwise it never gets better or there is never any improvement in any event I'm sure it's complicated as most algorithms are so I just except they are what they are and I'll keep bitching about it lmfao.
I'm not sure what you mean by "fix it". There is no singular solution that will "fix" wrestlestat's algorithm. As far as I can tell, Andre seems to be fairly responsive when issues are brought up and he's trying to incrementally make it better.

I find wrestlestat to be a great database for results first and foremost. The head to head predictions I take with a gigantic grain of salt because I know there are weaknesses/shortcomings with the approach. But, when someone is trying to use mathematics or some kind of objective analysis to predict outcomes of specific wrestling matches, I'm pretty lenient when it comes to the accuracy. Because a) that's really hard to do, and b) almost no one even tries to do it.
 
I would like to know how career matches applies. So if he lost all 150 what then. I'm not saying it shouldn't apply I just want to know how it applies to this f d up algorithm they use. Once again if your algorithm is broke. Fix it
Tell everyone that you have no idea what "fixing it" would entail, without telling everyone that you have no idea what "fixing it" would entail.

Each tweak that they make would require re-running the calculations, followed by quality control to see if the calculations did what you think they should do. It would take hours just to make one tweak and QC that tweak.

Let's say you think there is a recency problem and a problem with how slowly freshman move up the ranks.
125# wrestlers A,B and C are ranked BCA, and you don't like it because A recently beat C. Meanwhile, at 133, wrestlers D, E and F are ranked FED, but you think they should be FDE, because D is an obvious freshman stud, and should simply be ranked higher.

So, WS starts working on the recency issue. They spend days tweaking the "recency weighting," until finally BAC, your perfect 125 ranking. But, 133 FED. So, now they go to work on the freshman thing. They spend days tweaking and tweaking, until finally they get FDE, announce the new "District four approved WS rankings."

In great anticipation, you start clicking, knowing that these rankings are going to look just like every other ranking service's rankings (won't that be great). You get to the WS home page, click on 125, the page loads slower than Flo during the match of the century. Finally, the 125 link opens, and what do you see? BCA.

Why? Because you can't adjust the weighting of any variable without it affecting every last wrestlers ELO score, even the ones whose rankings were "right." And, to get every last ranking "right" would take a Herculean effort.
 
Just out of curiosity and I mean that. I'm not trying to be offensive. But what exactly about the algorithm is worthwhile. It seems everyone excepts the algorithm is flawed why shouldn't we say something so whoever came up with it makes changes. I get the argument that not everyone does rankings so having rankings is nice and it gives us something to talk about. My argument relies on the notion that if something is broken you fix it otherwise it never gets better or there is never any improvement in any event I'm sure it's complicated as most algorithms are so I just except they are what they are and I'll keep bitching about it lmfao.
I hope to God that the whole goal of these rankings is just to torque a couple of people off every time they are released. If that's the case, then that's what makes them worthwhile to me. 😂
 
I'm not sure what you mean by "fix it". There is no singular solution that will "fix" wrestlestat's algorithm. As far as I can tell, Andre seems to be fairly responsive when issues are brought up and he's trying to incrementally make it better.

I find wrestlestat to be a great database for results first and foremost. The head to head predictions I take with a gigantic grain of salt because I know there are weaknesses/shortcomings with the approach. But, when someone is trying to use mathematics or some kind of objective analysis to predict outcomes of specific wrestling matches, I'm pretty lenient when it comes to the accuracy. Because a) that's really hard to do, and b) almost no one even tries to do it

I hope to God that the whole goal of these rankings is just to torque a couple of people off every time they are released. If that's the case, then that's what makes them worthwhile to me. 😂
Lol. I agree with post above there is a lot to like about wrestle stat as a hole and I'm not that invested in rankings I would rather have what they offer as opposed to not
 
Wrestlestat hiccup - Joey O of Rutgers over RBY, hmmmm!
It is a computer algorithm. Should the punishment be to hit it with 130volts and fry it's core? Or just accept it is a computer mathematical analysis and stop trying to punish an inanimate entity. Your call :)
 
  • Haha
Reactions: NoVa Lion
It is a computer algorithm. Should the punishment be to hit it with 130volts and fry it's core? Or just accept it is a computer mathematical analysis and stop trying to punish an inanimate entity. Your call :)

I think the point is, the input data or the algorithm clearly has flaws and could be adjusted.

These kind of errors are head scratchers, and intuitively it's hard to contemplate how the algorythm comes up with such a result.

I still think keeping a record of common opponents past 2 years is informative but results older than 2 years should not be pushed through the algorythm. I don't think that is what happened with this prediction, but clearly something is amiss and with a little investigation the flaw in the data or the algorythm can be identified and corrected.

In the end as you get into January, if you see obvious anomalies like this, then the overall tool, is less useful than it could or should be.
 
Last edited:
  • Like
Reactions: NittanyLion84
It is a computer algorithm. Should the punishment be to hit it with 130volts and fry it's core? Or just accept it is a computer mathematical analysis and stop trying to punish an inanimate entity. Your call :)
I totally understand, I’m just highlighting in case andegre would like to be made aware of these anomalies with the algorithm.
 
I totally understand, I’m just highlighting in case andegre would like to be made aware of these anomalies with the algorithm.
It won't matter it is what it is it's been like this since I can remember and it isn't going to change.
 
It won't matter it is what it is it's been like this since I can remember and it isn't going to change.
It has changed, though, even though it doesn't meet your idea of "fixed." And, "as long as you can remember" goes back to February of 2018, when you first showed up here, right? WS goes back to at least 2014, I believe.

It used to take several more matches for an up and coming wrestler to zoom up the charts. WS fixed that, and that is why you see a guy like Olivieri moved up so quickly. You used to have to wait 16 or 17 matches to see where new wrestler was going to begin to settle out. But, with Olivieri, he jumped from 1200 to 1404 ELO points in two weeks, thanks to beating a few kids he was "supposed" to beat, and a few kids he wasn't "supposed" to beat. (The "supposed to" scale based on ELO score at the time of the match.) So, that's one example of WS making an improvement, despite your claim.

I will also say that I just did a quick search for "Wrestlestat" in this forum, and some of the same people have been bitching about these rankings going back to at least December 2018. I can't vouch for their sanity, but bitching about something for four years, yet still going back week after week and beating your head against that wall, meets one definition for insanity.
 
I think the point is, the input data or the algorithm clearly has flaws and could be adjusted.

These kind of errors are head scratchers, and intuitively it's hard to contemplate how the algorythm comes up with such a result.

I still think keeping a record of common opponents past 2 years is informative but results older than 2 years should not be pushed through the algorythm. I don't think that is what happened with this prediction, but clearly something is amiss and with a little investigation the flaw in the data or the algorythm can be identified and corrected.

In the end as you get into January, if you see obvious anomalies like this, then the overall tool, is less useful than it could or should be.
What do you think its usefulness is? Serious question. What is the usefulness of, say, Intermat's rankings? Betting? Playing KYPSW? As a conversation piece? As a means of thoroughly ticking off aggrieved wrestling fans because their guy was snubbed by a computer?
 
  • Like
Reactions: 86PSUPaul
It has changed, though, even though it doesn't meet your idea of "fixed." And, "as long as you can remember" goes back to February of 2018, when you first showed up here, right? WS goes back to at least 2014, I believe.

It used to take several more matches for an up and coming wrestler to zoom up the charts. WS fixed that, and that is why you see a guy like Olivieri moved up so quickly. You used to have to wait 16 or 17 matches to see where new wrestler was going to begin to settle out. But, with Olivieri, he jumped from 1200 to 1404 ELO points in two weeks, thanks to beating a few kids he was "supposed" to beat, and a few kids he wasn't "supposed" to beat. (The "supposed to" scale based on ELO score at the time of the match.) So, that's one example of WS making an improvement, despite your claim.

I will also say that I just did a quick search for "Wrestlestat" in this forum, and some of the same people have been bitching about these rankings going back to at least December 2018. I can't vouch for their sanity, but bitching about something for four years, yet still going back week after week and beating your head against that wall, meets one definition for insanity.
Im not sure why I got singles out I have mentioned several times I have accepted the rankings and I'm glad we have them. Making comments on how they can improve is not necessarily bitching in my opinion. If those who run the rankings want to improve them then they need to listen but it doesn't matter either way to me just saying. I still look at them and see them for what they are which isn't fact but what the computer spits out. Still a good source for conversation
 
  • Like
Reactions: ccdiver and 1032004
pawrestlersintn, I admire your fortitude.

Btw, WrestleStat did start in 2014, as all wrestlers records for that year and since can be retrieved, but nothing before.

Lastly, andegre shows up here and has responded to questions often in the past. Admirable.
I am not doxxing andegre as he has mentioned this on other threads. His first name is Greg.
And if anyone cares, I am Joe
 
pawrestlersintn, I admire your fortitude.

Btw, WrestleStat did start in 2014, as all wrestlers records for that year and since can be retrieved, but nothing before.

Lastly, andegre shows up here and has responded to questions often in the past. Admirable.
Lol. Regarding the sanity statement...I, perhaps, could consider my defense of WS similarly to the guys who keep coming back for more. 🤪

The user experience today, versus those days, should lead people to two conclusions.
1. WS has been listening to user input.
2. Their time, in their estimation, has been better spent working mostly on things other than tweaking the algorithm, although they have done some of that, as well.
 
Wrestlestat hiccup - Joey O of Rutgers over RBY, hmmmm!

Olivieri is ranked #4 with his best win over #25 Kyle Berwick. Losses to #7 Phillippi, #22 Matin, and #86 Maida.

A head-scratcher for sure. I love the site as a resource though, but I've learned to take the rankings with a grain of salt.
 
  • Like
Reactions: NittanyChris
Ayala has 12 matches, and is 10-2, with some wins over higher ranked (high ELO score) guys. Those good wins mean a lot to an ELO score that is only built on a limited number of matches. Also, a low ELO guy beating a guy he is not "supposed" to beat (a high ELO guy), causes a big jump in ELO score. Finally, his losses to McKee don't hurt him much, because he is "supposed" to lose to him.

One of the tweaks that Andegre did work on was to tweak the early career calcs, so that a new wrestler's scores would jump quicker than it would in the chess world, due to the limited number of matches a wrestler gets, versus the number of matches a chess player gets.

Those "some wins" prior to Sunday were 2 wins over the same guy who himself is now only #16. I don't see any way that ranking was possible based on the results. But after Sunday looks like it was a good prediction!

I do agree with some of the comments that the algorithm could likely be improved by not putting as much weight on results from 3+ years ago (for example, still having Desanto ahead of RBY). I also think NCAA tournament matches should have more weight (for example, still having Kemerer ahead of Starocci).

Obviously there is not much "usefulness" to ANY rankings, but in the conversation of rankings as a whole, having a set of rankings available that have literally no subjectivity is pretty cool. So I don't see the issue with making suggestions...obviously they don't have to take them. I would also say having the rankings be "more accurate" would be helpful considering how awesome the rest of the site is...then for example if you are already looking at the common opponents in a dual comparison there may not be as much need to flip back to Intermat or whatever to check out the rankings.
 
Observations RE Wrestlestat:
1. If you want to inject human interference to manually correct what seem like obvious errors......it is no longer an unbiased ranking.
2. If you change the algorithm to "correct" the perceived obvious errors.......it will change all calculations and will almost assuredly just move the "Obvious errors" to different matchups. Basically kicking the error(s) to a different location
3. After reading all these suggestions/complaints.......I am going to Bitch-Slap my toaster* until further notice :)

* YES.....that's what they are calling it now :)
 
  • Like
Reactions: danoftw
ADVERTISEMENT
ADVERTISEMENT