WrestleStat Week 3 Rankings

[andegre]

Here's the newest installment of the rankings. Just want to point out that a bug was found in the rankings process copying the EloRank of a wrestler from one week to the next. I've fixed this and re-processed the rankings accordingly.

Here we go...

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

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

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

 

[pish69]

Appreciate your Work Andegre.

I understand younger wrestlers have to get matches in to get a good feel for where they are. But how is Dean Heil, a senior two time champ who hasn’t lost in awhile, ranked third?

Also, how is Rasheed over Cutch when Cutch has been starter for 3 yrs and Rasheed hasn’t?

Look forward to your answer

 

[dicemen99]

Just started browsing but noticed something you may want to address.

Lane Peters. Has lost big to everyone in the top 30 he has faced. Plus a loss to #88. Highest ranked win is over #56. Yet this week he zooms up over 100 spots to crack the top 20. Something is off there. How did that happen when his week consisted of a win over 56, 107, 161 and a blowout loss to 12?

 

[andegre]

Just started browsing but noticed something you may want to address.

Lane Peters. Has lost big to everyone in the top 30 he has faced. Plus a loss to #88. Highest ranked win is over #56. Yet this week he zooms up over 100 spots to crack the top 20. Something is off there. How did that happen when his week consisted of a win over 56, 107, 161 and a blowout loss to 12?

Definitely sounds fishy, I'll take a look, thanks for pointing it out!

 

[andegre]

Appreciate your Work Andegre.

I understand younger wrestlers have to get matches in to get a good feel for where they are. But how is Dean Heil, a senior two time champ who hasn’t lost in awhile, ranked third?

Also, how is Rasheed over Cutch when Cutch has been starter for 3 yrs and Rasheed hasn’t?

Look forward to your answer

I'll have to research that and get back to you...

 

[lobo lion]

Interesting and fun stuff. Here's one to look at. A. Valencia loses 2 of his last 3, both by MD (to the #2 and #10 guys), and leapfrogs Tyler Chandler, who won 2 of his last 3, and whose only loss was a 1-point decision to #7 and whose latest match was a decision over the #19.

 

[pawrestlersintn]

Interesting and fun stuff. Here's one to look at. A. Valencia loses 2 of his last 3, both by MD (to the #2 and #10 guys), and leapfrogs Tyler Chandler, who won 2 of his last 3, and whose only loss was a 1-point decision to #7 and whose latest match was a decision over the #19.

Keep in mind, the math doesn't care about #1, #2, etc. It cares about the points that it calculates (right-hand column). Then, it cares (horrible math term) about "should have" or "shouldn't have", rewarding guys for beating guys they should have or punishing them for losing to someone they shouldn't have. For beating or losing to someone close to their rank value, the math says, "Meh."

With that in mind, go back and look at the ratings history of the guys you're talking about, and I'll bet they make more sense.

 

[lobo lion]

Keep in mind, the math doesn't care about #1, #2, etc. It cares about the points that it calculates (right-hand column). Then, it cares (horrible math term) about "should have" or "shouldn't have", rewarding guys for beating guys they should have or punishing them for losing to someone they shouldn't have. For beating or losing to someone close to their rank value, the math says, "Meh."

With that in mind, go back and look at the ratings history of the guys you're talking about, and I'll bet they make more sense.

Thanks.

 

[andegre]

Keep in mind, the math doesn't care about #1, #2, etc. It cares about the points that it calculates (right-hand column). Then, it cares (horrible math term) about "should have" or "shouldn't have", rewarding guys for beating guys they should have or punishing them for losing to someone they shouldn't have. For beating or losing to someone close to their rank value, the math says, "Meh."

With that in mind, go back and look at the ratings history of the guys you're talking about, and I'll bet they make more sense.

You're giving much better answers than I could, you're hired!

 

[csiebott]

Definitely sounds fishy, I'll take a look, thanks for pointing it out!

Mark Hall has two matches in the Binghamton dual. That can't be right.

 

[lions#1]

How are the points calculated? Jason Nolf has pinned all 7 of his opponents while Zain has pinned 5 of his 7 opponents. Yet, Zain has more points......

 

[RoarLions1]

How are the points calculated? Jason Nolf has pinned all 7 of his opponents while Zain has pinned 5 of his 7 opponents. Yet, Zain has more points......

It's not based on just this year.

A basic understanding of Elo can be found here, though tweaked for college wrestling...
https://en.wikipedia.org/wiki/Elo_rating_system

 

[pawrestlersintn]

You're giving much better answers than I could, you're hired!

I've been your online bodyguard for a few years now. Good news for you, I work cheap.

Two questions.

As it relates to another question on here already, have you adjusted the math for various types of wins or losses? Or do 15 Falls get you the same rating as 15 decisions?

Secondly, what are the point differentials that you use for a dual comparison for various types of winds? In other words what does the rating point differential have to be for say a pin versus a Tech fall?

 

[KCLion]

Here's the newest installment of the rankings. Just want to point out that a bug was found in the rankings process copying the EloRank of a wrestler from one week to the next. I've fixed this and re-processed the rankings accordingly.

Here we go...

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

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

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

Might want to adjust for Michigan vs Lehigh

 

[andegre]

I've been your online bodyguard for a few years now. Good news for you, I work cheap.

Two questions.

As it relates to another question on here already, have you adjusted the math for various types of wins or losses? Or do 15 Falls get you the same rating as 15 decisions?

Secondly, what are the point differentials that you use for a dual comparison for various types of winds? In other words what does the rating point differential have to be for say a pin versus a Tech fall?

Good/valid questions.

The answer is yes, win types are accounted for.

In Elo, there's a "k" factor that's a part of the algorithm, that value is modified per calculation based on the score difference of the match. If it's a 3-2 decision, then the K factor goes up by something proprotionate to 1 (don't remember the exact calculation). A 14 - 3 MD would move k by something proportionate to 11 (14 - 3). TFs are strictly by their point differential as well. Falls, on the other hand, are factored at 20 - 0 in the calculations, so it's always 20 points (proportionate of) for moving the K value.