WrestleStat Rankings - Week 5 2018-2019

[andegre]

As expected, there's quite a bit of movement in this weeks rankings after the CKLV tournament last weekend.

Looking through these, there are a couple confusing movements, but keep in mind that the low number of match results for some wrestlers will skew their ranking. Once a wrestler gets to the "equilibrium" match count [14], their ranking should become more accurate.

I'll provide an example where it's "tough" for the algorithm....Gable Steveson has propelled to #1 at 285 after an impressive start to his college career. He is currently 13 - 0 (12 - 0 vs D1) so he is ALMOST at his equilibrium, and rightfully-so, his ranking is pretty close to reality.

On-the-other-hand, Anthony Cassar from Penn State is also undefeated this year, where he just majored #6 Jordan Wood, while Cassar is still at #8. The reason for this is because he's well past his equilibrium match count, so those big upsets will not move the wrestler as much.

Here are the rankings for the week.

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

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

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

Statistical Rankings: https://www.wrestlestat.com/rankings/statistical

Don't forget you can project out the NCAA Championships for all top 120 placements to see who will win this year.

https://www.wrestlestat.com/tourney/projection

Lastly, the Pick'Em leaderboard has been updated for this week, and the race is getting a little tighter!

https://www.wrestlestat.com/pickem/leaderboard

 

[PSU_Pride_93]

can you elaborate on the "equilibrium" a little more? because im looking at these rankings go wtf. you talk about equilibrium for heavyweights and it seems like your equilibrium in order to be ranked appropriately deals with a minimum number of matches. I am just not sure how what the algorithm is to have micah jordan still above ashnault after we witnessed this matchup this weekend, same with eierman and mckenna. thanks

 

[Hartzie]

As expected, there's quite a bit of movement in this weeks rankings after the CKLV tournament last weekend.

Looking through these, there are a couple confusing movements, but keep in mind that the low number of match results for some wrestlers will skew their ranking. Once a wrestler gets to the "equilibrium" match count [14], their ranking should become more accurate.

I'll provide an example where it's "tough" for the algorithm....Gable Steveson has propelled to #1 at 285 after an impressive start to his college career. He is currently 13 - 0 (12 - 0 vs D1) so he is ALMOST at his equilibrium, and rightfully-so, his ranking is pretty close to reality.

On-the-other-hand, Anthony Cassar from Penn State is also undefeated this year, where he just majored #6 Jordan Wood, while Cassar is still at #8. The reason for this is because he's well past his equilibrium match count, so those big upsets will not move the wrestler as much.

Here are the rankings for the week.

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

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

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

Statistical Rankings: https://www.wrestlestat.com/rankings/statistical

Don't forget you can project out the NCAA Championships for all top 120 placements to see who will win this year.

https://www.wrestlestat.com/tourney/projection

Lastly, the Pick'Em leaderboard has been updated for this week, and the race is getting a little tighter!

https://www.wrestlestat.com/pickem/leaderboard

How is 6-3 freshman Brian Courtney ranked ahead of 6-0 freshman RBY?

 

[andegre]

The algorithm was analyzed to come up with an "equilibrium" match count for wrestlers that would indicate how many matches it would take to get a wrestler as close as possible to what their ranking should be. This was mainly put in place to accelerate the movement of a YOUNG wrestlers ranking as quickly as possible to "get them where they should be". This was mainly for the super freshman, ie Fix, RBY, Lee, Steveson, etc. There were hundreds of rounds of ranking simulations done in order to get to this number which resulted in the current algorithm gaining the #1 spot in the "Rank the Rankers" results as put out by someone on "The Mat" forums. The algorithm has some side-effects, which I'm sure you've seen already, but it's proven, so far, to be quite accurate.

 

[KCLion]

can you elaborate on the "equilibrium" a little more? because im looking at these rankings go wtf. you talk about equilibrium for heavyweights and it seems like your equilibrium in order to be ranked appropriately deals with a minimum number of matches. I am just not sure how what the algorithm is to have micah jordan still above ashnault after we witnessed this matchup this weekend, same with eierman and mckenna. thanks

Wood ahead of Cassar? Not being critical, but that raised my eyebrow as well.

 

[pointingdogsrule]

Guys, guys..... the only thing that matters in the end is in March. "World Domination"

Yep, I know it's fun to follow rankings.

 

[Hartzie]

Guys, guys..... the only thing that matters in the end is in March. "World Domination"

Yep, I know it's fun to follow rankings.

No argument there. One of the reasons these ranking are published in season is to stimulate discussion. That is what we are doing... discussing.

 

[STAND with PRIDE]

How is 6-3 freshman Brian Courtney ranked ahead of 6-0 freshman RBY?

This IS a head scratcher, even after reading his seemingly understandable explanation.

 

[psumacw]

for courtney vs RBY, it seems the answer lies with the rankings from last year. courtney came into the year with an elo rank of 1395 points. since then, courtney has only lost to decent guys (ranked 27-31 now) so his elo rank has not dropped much, sitting at 1387 now.

RBY on the other hand started the season with a 1200 elo rank since he's a true freshman. he has moved up after winning each week but his best win was only against a 1366 elo rank (paetzell).

as RBY keeps winning, his elo rank will improve. RBY should jump courtney with a win vs kline on 12/14, even if courtney wins in between, because kline is ranked #20 and courtney's two opponents in between are #60 and #91.

 

[Lyons212]

The algorithm was analyzed to come up with an "equilibrium" match count for wrestlers that would indicate how many matches it would take to get a wrestler as close as possible to what their ranking should be. This was mainly put in place to accelerate the movement of a YOUNG wrestlers ranking as quickly as possible to "get them where they should be". This was mainly for the super freshman, ie Fix, RBY, Lee, Steveson, etc. There were hundreds of rounds of ranking simulations done in order to get to this number which resulted in the current algorithm gaining the #1 spot in the "Rank the Rankers" results as put out by someone on "The Mat" forums. The algorithm has some side-effects, which I'm sure you've seen already, but it's proven, so far, to be quite accurate.

One issue with Cassar and his equilibrium is that almost all of his matches are at 197, there are very few guys who move up to 285. At this point in time, Cassar needs to be treated almost like a freshmen.

Generally I like statistical models for rankings, eliminates many biases. However in this case, Cassar being ranked behind Wood after dismantling him on Sunday does not pass the smell test (challenges of modeling). This is most likely because of Cassar's matches being at 197.

 

[impalanut]

even that doesn't make sense. His only loss was in overtime to a ranked wrestler (Horschlag) and he had wins over ranked wrestlers, including the #1.

 

[psumacw]

even that doesn't make sense. His only loss was in overtime to a ranked wrestler (Horschlag) and he had wins over ranked wrestlers, including the #1.

also lost to garrett hoffman. also only had 1 bonus point victory over an opponent in the top 100.

that said, i think his 2015 results are acting like an anchor on his current ranking. i would have thought his victories over nevills and wood (and others) to be worth more than 20 elo points.

 

[RoarLions1]

Educated guesses, andegre can affirm, or not .

The Courtney / RBY situation is, as someone else said, a result of Courtney's RS year, and having 30 bouts under his belt, well above the "15" threshold for equilibrium. RBY will keep climbing...so hope we have this discussion again after he gets to 15 bouts.

For the 285 situation, Cassar is sitting at 6 wins at 285. It appears the model gives wrestlers that change weights the same points as they had finishing the prior year, which for Anthony as 1356.28. From there, I'm guessing the model reverts to the "15" bout equilibrium to allow the dust to settle. Wood was 44-9 heading into this year (counting RS year), with some pretty good wins. When Anthony gets 9 or so more bouts, he'll be higher ranked in this model.

As an aside, Steveson was 123rd to start the season, then 43rd, then 7th, 5th, 3rd and 1st at the end of weeks 1 through 5. He now has 17 bouts in, counting a few as a HS senior. I suspect Anthony will follow a similar progression if he keeps winning, particularly if they're BP wins.