ADVERTISEMENT

157 lbs- Is #81 better than #31?

Wrestlestat is a "go to site" for most wrestling fans but their rankings haven't been the best. I haven't watched much of the two so I'll let your fellow Hawk fans be the judge on who's the better option. Hope local kid Murin stays healthy this year and gets on the podium. Great kid!
 
First, I don't recall Murin having much history of injury, so that's not a concern at all.

Second, if I understand this correctly, WrestleStat rankings are based on an algorithm (tweeked over time), whereas rankings on other sites are opinion-based. Algorithms aren't perfect, but they try to keep out bias/opinions and instead base rankings on statistical criteria. Therefore certain individual rankings may seem strange, but will be adjusted appropriately with sufficient data (match results).
 
First, I don't recall Murin having much history of injury, so that's not a concern at all.

Second, if I understand this correctly, WrestleStat rankings are based on an algorithm (tweeked over time), whereas rankings on other sites are opinion-based. Algorithms aren't perfect, but they try to keep out bias/opinions and instead base rankings on statistical criteria. Therefore certain individual rankings may seem strange, but will be adjusted appropriately with sufficient data (match results).
The results of the algorithm are only as good as the algorithm itself. and despite being an amazing resource for results, wrestlestat has never come close to producing a set of competent rankings. use it for results only. its results and projections are not sound.

best resource in wrestling hands down. but know what it's good for and use it correctly.
 
best resource in wrestling hands down. but know what it's good for and use it correctly.

2f9db62d-767a-45bc-81eb-c4cd29866f89_text.gif
 
The results of the algorithm are only as good as the algorithm itself. and despite being an amazing resource for results, wrestlestat has never come close to producing a set of competent rankings. use it for results only. its results and projections are not sound.

best resource in wrestling hands down. but know what it's good for and use it correctly.
Agreed.

It's quite difficult to create algorithms, there will always be certain flaws, but I think Greg has done an excellent job and has continued to make improvements. If anyone has suggestions for how to do it better, you can contact him.
 
Agreed.

It's quite difficult to create algorithms, there will always be certain flaws, but I think Greg has done an excellent job and has continued to make improvements. If anyone has suggestions for how to do it better, you can contact him.
The information on it is mind blowing. I originally had problems with the rankings but after being brow beat and schooled by the pool in what too expect I stfu. Lol
I appreciate the effort!
 
First, I don't recall Murin having much history of injury, so that's not a concern at all.

Second, if I understand this correctly, WrestleStat rankings are based on an algorithm (tweeked over time), whereas rankings on other sites are opinion-based. Algorithms aren't perfect, but they try to keep out bias/opinions and instead base rankings on statistical criteria. Therefore certain individual rankings may seem strange, but will be adjusted appropriately with sufficient data (match results).



you dont remeber the shoulder brace?

 
  • Like
Reactions: el dub
Second, if I understand this correctly, WrestleStat rankings are based on an algorithm (tweeked over time), whereas rankings on other sites are opinion-based. Algorithms aren't perfect, but they try to keep out bias/opinions and instead base rankings on statistical criteria.

GIGO (?).
 
I think WrestleStat rankings are pretty good once you get a big enough sample size.
If the recipe (algorithm) for baking the cake is flawed, the cake (rankings) is flawed.
'Garbage In - Garbage Out'. GIGO. Precision counts.
 
  • Like
Reactions: el dub
I like the site for the details and the comparisons, but their ranking system is flawed.

Somehow nowadays people can make the argument that 10x10 may or may not be 100, they refer to it as the new normal.

Is zero a number, well you cannot multiple or divide by it so maybe its not. Crazy stuff!
 
Wrestlestat uses a modified Elo ranking, which was initially built for chess, but is now probably the most common ranking system in the world for zero sum games.

I don't know the specific adjustments Greg (or Andre, as I heard he likes to be called :)) made to tailor it to wrestling, but its about as objective as you can get.

If any of you follow Iowa HS football, there is an ISU economist named Brent Moore who has produced a regression analysis ratings system for years. Its pretty interesting and very accurate as well. (Accurate meaning fitting a line or curve) Individual performances and outings have variability, but the predictive power is very strong with the amount of data that goes into that system. it also has College football, and Boys and Girls High School basketball. http://ia.bcmoorerankings.com/fb/2022/latest/
 
If the recipe (algorithm) for baking the cake is flawed, the cake (rankings) is flawed.
'Garbage In - Garbage Out'. GIGO. Precision counts.
It's really Irritating that Greg (who happens to be a Hawkeye wrestling fan) would be criticised for creating WrestleStat rankings using an algorithm. Anyone doing so is ignorant about data-driven calculations. I steer clear of those who criticise and complain - they bring nothing of value to the table.
 
you dont remeber the shoulder brace?
You're right, I definitely remember the shoulder brace. I should have worded my comment better. We all know that wrestlers are always banged up, especially later in the season. What I meant to point out is that Max, as best as I remember, hasn't lost lost any significant amount of time to injury and is not injury-prone. Therefore I'm not concerned about needing a solid back-up.
 
  • Like
Reactions: armbartilt
There is a reason why ELO rating system does not work well for physical sports. Chess doesn't have variables like wrestling. Injuries probably would be the most significant variable and rankings can't really build that in without human input. Guess the machines can't take over just yet!
 
  • Like
Reactions: armbartilt
It's really Irritating that Greg (who happens to be a Hawkeye wrestling fan) would be criticised for creating WrestleStat rankings using an algorithm. Anyone doing so is ignorant about data-driven calculations. I steer clear of those who criticise and complain - they bring nothing of value to the table.
When legitimate research is published on a public forum (Hawkeye Report) the publisher absolutely must expect that work to be Peer Reviewed, and that work to be Assessed Critically. Peer Review is not a personal attack on the researcher, rather it is an effort to move information in the direction of greater accuracy. And Critical Analysis simple provides a different frame-of-reference for viewing that data.

Figuratively, perhaps literally, the algorithm is the Black Box of data analytics. The consumer/reader has prompt access to what goes into the Black Box (data). And the consumer/reader may readily observe what exits the Black Box (work product). What transpires inside the Black Box, more often than not, is totally unseen by the reader/end user. My critique.

Hopefully this posting brings something of value to the table.

The Best to the University of Iowa Hawkeyes against the Black Knights of the Hudson.
 
When legitimate research is published on a public forum (Hawkeye Report) the publisher absolutely must expect that work to be Peer Reviewed, and that work to be Assessed Critically. Peer Review is not a personal attack on the researcher, rather it is an effort to move information in the direction of greater accuracy. And Critical Analysis simple provides a different frame-of-reference for viewing that data.

Figuratively, perhaps literally, the algorithm is the Black Box of data analytics. The consumer/reader has prompt access to what goes into the Black Box (data). And the consumer/reader may readily observe what exits the Black Box (work product). What transpires inside the Black Box, more often than not, is totally unseen by the reader/end user. My critique.

Hopefully this posting brings something of value to the table.

The Best to the University of Iowa Hawkeyes against the Black Knights of the Hudson.
I concur with the above however the using the description of GIGO to describe the rankings sounded insulting to Greg/wrestlestat when it was actually a valid insult to the data --specifically the lack of it. All algorithms have increased statistical outliers with fewer data points. From your jargon I suspect you know that. The increased statistical probabilities of outliers increases with smaller sample size. IMO it is not the algorithm it is the lack of sample size. Hence I disagree with your GIGO because there is insufficient data to make that analysis and interpretation of the algorithm. Precision does matter but you can be precisely wrong. Precise but not accurate is how an incorrectly sighted rifle shoots. Precise tight groupings that miss the target.

My IT manager would probably call that an Id 10t error. ;)
 
I concur with the above however the using the description of GIGO to describe the rankings sounded insulting to Greg/wrestlestat when it was actually a valid insult to the data --specifically the lack of it. All algorithms have increased statistical outliers with fewer data points. From your jargon I suspect you know that. The increased statistical probabilities of outliers increases with smaller sample size. IMO it is not the algorithm it is the lack of sample size. Hence I disagree with your GIGO because there is insufficient data to make that analysis and interpretation of the algorithm. Precision does matter but you can be precisely wrong. Precise but not accurate is how an incorrectly sighted rifle shoots. Precise tight groupings that miss the target.

My IT manager would probably call that an Id 10t error. ;)
And another problem is that the model is presumably deterministic, no randomness which clearly is relevant. But we don't like confidence intervals on rankings, so here we are. And Lord help us if the equation is nonlinear, which would invoke chaos theory. All bets are off, model might even predict ISU is #1 and @MitchLL is their biggest fan.
 
And another problem is that the model is presumably deterministic, no randomness which clearly is relevant. But we don't like confidence intervals on rankings, so here we are. And Lord help us if the equation is nonlinear, which would invoke chaos theory. All bets are off, model might even predict ISU is #1 and @MitchLL is their biggest fan.
Quantum Wrestling Stats…

Do results change due to someone observing the results?
 
There is a reason why ELO rating system does not work well for physical sports. Chess doesn't have variables like wrestling. Injuries probably would be the most significant variable and rankings can't really build that in without human input. Guess the machines can't take over just yet!
I dunno, I get a headache (injury) when I play chess.
 
And another problem is that the model is presumably deterministic, no randomness which clearly is relevant. But we don't like confidence intervals on rankings, so here we are. And Lord help us if the equation is nonlinear, which would invoke chaos theory. All bets are off, model might even predict ISU is #1 and @MitchLL is their biggest fan.
I did sit in the ISU section at Nattys for about 8-10 years back in the 90s.
 
Perhaps we should moderate our level of criticism over the WrestleStat rankings using an algorithm, i.e. GIGO. Statistics present an expected outcome with all things considered equal. The problem is that wrestling doesn't work that way. An individual may have strengths, or weaknesses against a certain style of wrestler that may not present itself in the statistics. The results of many matches can come down to a particular attack and reaction in a fraction of a second, especially if a match is close in the closing seconds.
 
Perhaps we should moderate our level of criticism over the WrestleStat rankings using an algorithm, i.e. GIGO. Statistics present an expected outcome with all things considered equal. The problem is that wrestling doesn't work that way. An individual may have strengths, or weaknesses against a certain style of wrestler that may not present itself in the statistics. The results of many matches can come down to a particular attack and reaction in a fraction of a second, especially if a match is close in the closing seconds.
I love wrestle stat just can’t take rankings early too seriously… if you are new to it ask questions, the nerds will explain it. My brolo is working on his own algorithm similar and I think anything wrestling is cool.
 
And Lord help us if the equation is nonlinear, which would invoke chaos theory. All bets are off!!!
Trapper85; Terrific post. Although it should be self-evident to all, your invocation of 'Chaos Therory' on HR wrestling was nothing short of literary genius. Implying Chaos Therory on HR, how dare you. You certainly made my day, still smiling. Great post.

Tonight; Hawkeyes against the Black Knights of the Hudson. May test some wrestlers. Go Hawks.
 
  • Like
Reactions: Trapper85
ADVERTISEMENT
ADVERTISEMENT