ADVERTISEMENT

WrestleStat Rankings | Week 4 2022

Assuming Michigan will move up once Suriano is plugged into the Tournament weight.
 
  • Like
Reactions: andegre
Here's WrestleStat's latest rankings. Not much movement today, but Penn State is back at #2 after moving up 3 spots.

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

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

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

Stat Leaders | https://www.wrestlestat.com/rankings/statistical
Individual rankings are ridiculous: RBY #3, Nick Lee #2, C Star #4, Brooks #2, and Kerk #9. Dean at #2 seems fair, though he is probably better than Ferrari.
 
  • Haha
Reactions: cheddarwurst69
Rankings are what they are, conversation starters.
174 is particularly ridiculous though.
Kemerer, last year's runner-up ranked ahead of 2 previous champions and last year's champion ranked behind 2 kids never good enough to win a championship plus another who has won a championship, 3 years ago.
 
133 is nearly as ridiculous as 174. Last year's champ ranked behind one kid who the champ beat so bad the kid actually left the mat crying after he was mentally beaten so bad he quit, and the other kid couldn't score a single point in the championship match.
 
  • Like
Reactions: danoftw
Guys, Wrestlestat is an algorithm. It's not a ranking in the traditional sense. I feel like this has to be explained every year. As I understand it, this early in the season the algorithm doesn't have enough data to reconcile the previous year's results. Though, @andegre can surely explain it much more clearly than I can.

Wrestlestat is a great tool, but don't mistake it for a human ranking system.
 
Guys, Wrestlestat is an algorithm. It's not a ranking in the traditional sense. I feel like this has to be explained every year. As I understand it, this early in the season the algorithm doesn't have enough data to reconcile the previous year's results. Though, @andegre can surely explain it much more clearly than I can.

Wrestlestat is a great tool, but don't mistake it for a human ranking system.
It is a ranking in the normal sense. The algorithm uses the designer's measuring tools. What the algorithm does is make certain the rankings are determined in a constant manner.

I know this is an annual conversation on here, I also know the early season rankings have nothing to go on from this season, so what has happened in seasons past are the determining factors. Based upon 133 and 174 it is obvious whatever is used as early season measure parameters has problems and needs some tweaking.
 
  • Haha
Reactions: cheddarwurst69
Rankings are what they are, conversation starters.
174 is particularly ridiculous though.
Kemerer, last year's runner-up ranked ahead of 2 previous champions and last year's champion ranked behind 2 kids never good enough to win a championship plus another who has won a championship, 3 years ago.
The algorithm apparently doesn't put much weight on head-to-head meetings.

In the dual comparison against Lehigh, they have Wood beating Kerk, this in spite of last year's match where Kerk won 11-2.

I'm not complaining, love the site ... just an observation.
 
  • Like
Reactions: nitlion6
It is a ranking in the normal sense. The algorithm uses the designer's measuring tools. What the algorithm does is make certain the rankings are determined in a constant manner.

I know this is an annual conversation on here, I also know the early season rankings have nothing to go on from this season, so what has happened in seasons past are the determining factors. Based upon 133 and 174 it is obvious whatever is used as early season measure parameters has problems and needs some tweaking.
The algorithm apparently doesn't account for how mentally beat Austin DeSanto was against RBY.
 
  • Like
Reactions: nitlion6
Guys, Wrestlestat is an algorithm. It's not a ranking in the traditional sense. I feel like this has to be explained every year. As I understand it, this early in the season the algorithm doesn't have enough data to reconcile the previous year's results. Though, @andegre can surely explain it much more clearly than I can.

Wrestlestat is a great tool, but don't mistake it for a human ranking system.
how can they use this so called tool to rank someone higher than the champ when they havent wrestled in 2 yrs???
 
  • Haha
Reactions: cheddarwurst69
I've been asking for years for any of the bitchers/moaners/whiners/complainers to come up with their own rankings. And, yet, strangely, I'm still waiting for the first one of them to come back with some.

The math for the algorithm is readily available. It's called the Elo rating system. Look it up, tweak it to your heart's content, put in all of the matches that Wrestlestats does, and bring us your rankings. Easy, and I'll bet they'll be the best rankings you've ever seen.

Problem is, the energy required to bitch is about 0.00001% of the energy required to even just start a website, let alone put all of the effort into the algorithm and data entry.

And, before anyone comes up with the annual, "Oh, I forgot we aren't allowed to criticize Wrestlestat rankings," I'll say this: Oh, I forgot I'm not allowed to criticize the people who are criticizing Wrestlestat rankings.
 
Like any algorithm, it has biases programmed into it by the creator. All ELO-based algorithms reward prolonged performance. Not recent results.

It also doesn't take the relative importance of a match into account; ELO doesn't differentiate between a dual against Oregon State, the Southwest Missouri State Open Invite and the NCAA championship.

So, when you are Carter Starocci, with a lifetime record of 36-2 (with a , your ELO isn't as high as a guy with a 110-10 record like Kemerer, or a 114-11 record like Hidlay.
 
Guys, Wrestlestat is an algorithm. It's not a ranking in the traditional sense. I feel like this has to be explained every year. As I understand it, this early in the season the algorithm doesn't have enough data to reconcile the previous year's results. Though, @andegre can surely explain it much more clearly than I can.

Wrestlestat is a great tool, but don't mistake it for a human ranking system.
Then the algorithm needs fixed because it's flawed end of story. If the algorithm is placing champions behind kids who never won or who lost to the champion head to head then something needs tweaked. It's not the bitching that should be focused on here. Enough of hiding behind algorithms to justify the rankings if you can't get it right in anyway stop ranking wrestlers your not good at it moral of the story their algorithm sucks out loud I think we can all agree on that. Now they need to stop ranking kids because they suck at it
 
I've been asking for years for any of the bitchers/moaners/whiners/complainers to come up with their own rankings. And, yet, strangely, I'm still waiting for the first one of them to come back with some.

The math for the algorithm is readily available. It's called the Elo rating system. Look it up, tweak it to your heart's content, put in all of the matches that Wrestlestats does, and bring us your rankings. Easy, and I'll bet they'll be the best rankings you've ever seen.

Problem is, the energy required to bitch is about 0.00001% of the energy required to even just start a website, let alone put all of the effort into the algorithm and data entry.

And, before anyone comes up with the annual, "Oh, I forgot we aren't allowed to criticize Wrestlestat rankings," I'll say this: Oh, I forgot I'm not allowed to criticize the people who are criticizing Wrestlestat rankings.
Don't get me wrong - I think the site is great for what it is. I was just annoyed that none of our 4 champs is #1 right now. I've been doing web/software development a long time, so I could do my own, but you've got it covered pretty well.
 
Then the algorithm needs fixed because it's flawed end of story. If the algorithm is placing champions behind kids who never won or who lost to the champion head to head then something needs tweaked. It's not the bitching that should be focused on here. Enough of hiding behind algorithms to justify the rankings if you can't get it right in anyway stop ranking wrestlers your not good at it moral of the story their algorithm sucks out loud I think we can all agree on that. Now they need to stop ranking kids because they suck at it
Then, by all means, fix it.
 
  • Like
Reactions: cheddarwurst69
Like any algorithm, it has biases programmed into it by the creator. All ELO-based algorithms reward prolonged performance. Not recent results.

It also doesn't take the relative importance of a match into account; ELO doesn't differentiate between a dual against Oregon State, the Southwest Missouri State Open Invite and the NCAA championship.

So, when you are Carter Starocci, with a lifetime record of 36-2 (with a , your ELO isn't as high as a guy with a 110-10 record like Kemerer, or a 114-11 record like Hidlay.
I don’t care about any rankings to be completely honest. Rarely do I even look at them unless it’s Highschool rankings… and that’s only to get a sense of who the top kids are throughout the country and find their matches to watch. Only opinions I care about are those who are seeding the Big 10 and NCAA’s. Then it maters.

As for Wrestlestat and Elo, it looks like they need to weigh most recent matches more heavily and tail off to practically nothing when it comes to results 4 years ago. The ranking is ranking wrestlers where they stand now, not what they did 4 years ago so most recent results mean a lot more.
 
Then, by all means, fix it.
Or be more like me and not give a s”t about where Wrestlestat ranks guys. Do people even go there to look at rankings anyway? All the great features of that site…viewing their rankings never really even crosses my mind. To be transparent though, even if they were considered the best out there as far as their rankings go, I still wouldn’t care to look at them much.
 
  • Like
Reactions: Gebmo
Like any algorithm, it has biases programmed into it by the creator. All ELO-based algorithms reward prolonged performance. Not recent results.

It also doesn't take the relative importance of a match into account; ELO doesn't differentiate between a dual against Oregon State, the Southwest Missouri State Open Invite and the NCAA championship.

So, when you are Carter Starocci, with a lifetime record of 36-2 (with a , your ELO isn't as high as a guy with a 110-10 record like Kemerer, or a 114-11 record like Hidlay.
As I said, algorithm has the bias built in of what the designer sets as weighted importance. Having no weighted difference between NCAA championships and a 2 year old open tournament result is kind of an interesting perspective.
Just my opinion, but last year's champion should carry some extra weight, as should more recent head to heads.
 
Or be more like me and not give a s”t about where Wrestlestat ranks guys. Do people even go there to look at rankings anyway? All the great features of that site…viewing their rankings never really even crosses my mind. To be transparent though, even if they were considered the best out there as far as their rankings go, I still wouldn’t care to look at them much.
Apparently, some of these guys are betting, and losing, their houses based on WS rankings. And, they're doing it year after year. 🤣

I'm going to go out on another limb here. I'm willing to bet that not a single one of the complainers has ever contributed a dime to WS. I'll bet I can dance a jig on this limb, and it won't ever fall down.
 
  • Like
Reactions: Reslo
I'm going to go out on another limb here. I'm willing to bet that not a single one of the complainers has ever contributed a dime to WS. I'll bet I can dance a jig on this limb, and it won't ever fall down.
OMG, please wear a harness this time!
 
  • Like
Reactions: nitlion6
BWI Algorithm Rankings

125 Shunk #1
133 RBY #1
141 Lee #1
149 Bartlett #1
157 Barraclough #1
165 Facundo #1
174 Starocci #1
184 Brooks #1
197 Dean #1
285 Kerkvliet #1

 
  • Like
Reactions: pawrestlersintn
Then the algorithm needs fixed because it's flawed end of story. If the algorithm is placing champions behind kids who never won or who lost to the champion head to head then something needs tweaked. It's not the bitching that should be focused on here. Enough of hiding behind algorithms to justify the rankings if you can't get it right in anyway stop ranking wrestlers your not good at it moral of the story their algorithm sucks out loud I think we can all agree on that. Now they need to stop ranking kids because they suck at it
 
So, I defended Elo before, but this makes me reconsider. Just ran a dual meet comparison vs Nebraska.

Aaron Brooks, 20-1 vs common opponents, 3-1 Venz, is predicted to lose to Venz, who is 15-8 vs common opponents and has lost the last three vs Brooks.

If it were a straight Elo comparison it seems impossible that it could favor Venz.
 
So, I defended Elo before, but this makes me reconsider. Just ran a dual meet comparison vs Nebraska.

Aaron Brooks, 20-1 vs common opponents, 3-1 Venz, is predicted to lose to Venz, who is 15-8 vs common opponents and has lost the last three vs Brooks.

If it were a straight Elo comparison it seems impossible that it could favor Venz.
That one seems like a math error. Aaron's Elo ranking is 1563, Venz is 1487. No way that should lead to a loss. @andegre?
 
ADVERTISEMENT

Latest posts

ADVERTISEMENT