Ratings - please read before commenting - Archived

[Jens S]

I'll make it easy for you:

1) How should anybody be able to see, how the system works?
2) Where can anyone see, that all the ratings of a boxer is changed after his fights?

...

My last try with you, but I fear, you don't read.

I said, you cannot see the fight impact NOW. But I will ADD a value to each bout, where will be able to see the influece of each bout to the rating.

Oh, yes I do, but you don't answer my questions. Some of them are implicit answered, but clearly not all. You reply, that we can't see it, but you don't let us know how you will make it visible.

Take Callum Smith:
https://boxrec.com/en/proboxer/631178

7) How many points extra did he get after his 23rd fight after his 27th fight?

It is a simple question you must know? Or?

Look at Drake Thadzi:
https://boxrec.com/en/proboxer/907

9) An extra question: how come Drake Thadzis career peak after a win against nobody, but deteriorate after the by far biggest win in his career?

10) Am I to understand it like this: Boxer A knockouts boxer B - KO1. Boxer A gets an injury that seriously diminish his boxing abilities. Fx a torn rotator cuff, gets in to a car accident or gets a drug addiction. He fights on despite being in poor shape. He loses to fighter C half a year after his win against boxer B. Does it affect his score after the win against boxer B?

11) Why?

That you add a value for clarity doesn't change the system and and doesn't reply many of my questions.

[computerrank]

My last try with you, but I fear, you don't read.

I said, you cannot see the fight impact NOW. But I will ADD a value to each bout, where will be able to see the influece of each bout to the rating.

Oh, yes I do, but you don't answer my questions. Some of them are implicit answered, but clearly not all. You reply, that we can't see it, but you don't let us know how you will make it visible.

Take Callum Smith:
https://boxrec.com/en/proboxer/631178

7) How many points extra did he get after his 23rd fight after his 27th fight?

It is a simple question you must know? Or?

Look at Drake Thadzi:
https://boxrec.com/en/proboxer/907

9) An extra question: how come Drake Thadzis career peak after a win against nobody, but deteriorate after the by far biggest win in his career?

10) Am I to understand it like this: Boxer A knockouts boxer B - KO1. Boxer A gets an injury that seriously diminish his boxing abilities. Fx a torn rotator cuff, gets in to a car accident or gets a drug addiction. He fights on despite being in poor shape. He loses to fighter C half a year after his win against boxer B. Does it affect his score after the win against boxer B?

11) Why?

That you add a value for clarity doesn't change the system and and doesn't reply many of my questions.

You moan: "You reply, that we can't see it, but you don't let us know how you will make it visible." -> I posted an example yesterday (you don't follow the thread ):
viewtopic.php?p=5371470#p5371470

ad 7:
ad 9:
I answered that question before in the thread - the ratings follow the big line - not the detailed bout.
ad 10: Yes - because the ratings only follow the official results.

[felinoboxing]

Greetings to all.
I was making P4P lists per year based on boxrec ratings. I had made 11 lists when the formula to calculate the points was modified.

Here are two P4P lists from late 1968. One old system and one new system.
I remind you that in those days Ali was suspended, for that reason he is not on the list.
Compare, and help me choose which one to stay with.

Old system

Curtis Cokes 5170
Carlos Ortiz 3648
Carlos Teo Cruz 3261
Bob Foster 3101
Sandro Mazzinghi 3027
Nino Benvenuti 3017
Nicolino Locche 2996
Jose Napoles 2423
Luis Rodriguez 2396
Ismael Laguna 2199

New system

Joe Frazier 283.3
Jimmy Ellis 97.04
Jose Napoles 96.53
Bob Foster 95.56
Oscar Bonavena 89.87
Nicolino Locche 74.77
Floyd Patterson 69.53
Emile Griffith 45.10
Curtis Cokes 43.55
Luis Rodriguez 41.50

Your list NeW System is not correct. I just looked at those and found ????
Curtis Cokes 189.5
Carlos Ortiz 160.8
Sandro Mazzinghi 171.4

Where are those numbers that you show?

I use the points that are awarded to each boxer in "after fight".
Always from the last fight they made in 1968.

The points of the 3 boxers you saw, I take your points from
these fights:

Curtis Cokes-Ramón La Cruz, in "after fight" records 43.55 for Cokes
Carlos Ortiz-Carlos Teo Cruz, 28.61 for Ortiz
Sandro Mazzinghi-Freddie Little, 000 for Mazzinghi

I'm doing it wrong. I take the points from the wrong place?
Correct me please

[computerrank]

Your list NeW System is not correct. I just looked at those and found ????
Curtis Cokes 189.5
Carlos Ortiz 160.8
Sandro Mazzinghi 171.4

Where are those numbers that you show?

I use the points that are awarded to each boxer in "after fight".
Always from the last fight they made in 1968.

The points of the 3 boxers you saw, I take your points from
these fights:

Curtis Cokes-Ramón La Cruz, in "after fight" records 43.55 for Cokes
Carlos Ortiz-Carlos Teo Cruz, 28.61 for Ortiz
Sandro Mazzinghi-Freddie Little, 000 for Mazzinghi

I'm doing it wrong. I take the points from the wrong place?
Correct me please

For Curtis Cokes-Ramón La Cruz, in "after fight" records 43.55 for Cokes.

Look here:
https://boxrec.com/en/proboxer/12555

I see:



189.5 for Curtis after the bout.

[felinoboxing]

I already saw it. I was looking in the wrong place.

Last thing, boxrec has the option of being able to see the ratings on a specific date?

I would like to see the P4P ratins after Marciano retirement, After Ali suspended, P4P rating of the day I was born, and many other moments.
Checking one x one is heavy work

[computerrank]

I already saw it. I was looking in the wrong place.

Last thing, boxrec has the option of being able to see the ratings on a specific date?

I would like to see the P4P ratins after Marciano retirement, After Ali suspended, P4P rating of the day I was born, and many other moments.
Checking one x one is heavy work

Boxrec would have to go over all bouts within the year before of about 20 to 40 thousand active boxers for each request. That would cost too much machine power ... I will update the Boxrec Annual Ratings next ...

[felinoboxing]

Hi. I see that the P4P rating is made with the values ​​of each fighter after his last fight. Except Canelo Alvarez. After beating Kovalev he is 1730, and at the P4P rating 2263. Why this?

[computerrank]

Hi. I see that the P4P rating is made with the values ​​of each fighter after his last fight. Except Canelo Alvarez. After beating Kovalev he is 1730, and at the P4P rating 2263. Why this?

That fight was at Light Heavyweight and Alvarez is rated at Middleweight.

[felinoboxing]

I am redoing the P4P list with the new formula. I make the cut at the end of each year. And I saw these 2 things. To the 1950 list, Harry Matthews jumped from position 45 in the old formula to position 4 in the new. Mathews is on top of big names like Marciano, Pep, Saddler, etc.
In 1967, Jose Napoles in his penultimate fight was 265.5, ranked # 3 on the P4P list. He made his last fight of that year against Charlie Watson. Napoles ended the year with only 65 points. In just 2 months I went from # 3 to not being in P4P.

What happened in those 2 cases?

[computerrank]

I am redoing the P4P list with the new formula. I make the cut at the end of each year. And I saw these 2 things. To the 1950 list, Harry Matthews jumped from position 45 in the old formula to position 4 in the new. Mathews is on top of big names like Marciano, Pep, Saddler, etc.
In 1967, Jose Napoles in his penultimate fight was 265.5, ranked # 3 on the P4P list. He made his last fight of that year against Charlie Watson. Napoles ended the year with only 65 points. In just 2 months I went from # 3 to not being in P4P.

What happened in those 2 cases?

Matthews was in the up to the win against Murphy at end of 1950.
Napoles last bout in 1967 was withoutt weight assignment. So the points are low as at Heavyweight.

[DazBoxingFan]

But still it begs the question why the formula ranking Past fighters is the same as ranking current ones. Makes no sense. Like someone mentioned before people who’s offering two variation forms, one for prediction purposes like the current one and one for actual ranking systems like the past ones.

[computerrank]

But still it begs the question why the formula ranking Past fighters is the same as ranking current ones. Makes no sense. Like someone mentioned before people who’s offering two variation forms, one for prediction purposes like the current one and one for actual ranking systems like the past ones.

The current Boxrec ratings only can use the past results as there are no future results. So Boxrec doesn't predict. The current formula has had best matching, when comparing the ratings before the bout (without knowledge of the later results) with the real results, It simply transforms all knowledge of the past bouts into best available ratings.

[JCS]

Added WHR on Fight Matrix. We now display 4 purely mathematical systems... two slightly different applications of Elo, Glicko-1 and now WHR. Can pick which flavor you like. In my honest opinion, Glicko-1 is probably the best of the 4. They all have a prediction rate somewhere in the mid 60s... MMA is a much harder sport to build a prediction system around.

In addition to the new system, I'm now displaying icons next to fighters on the standard rankings if there is a certain level of disagreement between our standard rankings and the other variants.

One thing I noticed in my testing is that the w2 value that WHR depends on, shows highest prediction rate in high-profile fights at a much lower value (~10) than the commonly used 60. A guy who was analyzing UFC results on Reddit landed in a similar area... around 12 I think.

Furthermore, WHR seems much more susceptible to rating inflation by way of can crushing... most notably, the "smart" kind that you see in eastern Europe and Russia.... where matchmakers often pay for fighters that are building glossy records elsewhere, to get beat by their fighters that are doing the same. Due to its nature of iterating and analyzing the opponent chains before and after bouts, it seems that winning records are much, much more important than opponent quality.

My take on WHR so far is that it seems to be an optimum system to use to determine when a fighter is entering or exiting their peak years. Outside of those moments, I much prefer an Elo/Glicko.

[SportsRatings]

Added WHR on Fight Matrix. We now display 4 purely mathematical systems... two slightly different applications of Elo, Glicko-1 and now WHR.

I see that WHR still has Cormier as the #1 HW

So what is the "standard" ranking on Fight Matrix?

Also I notice the ELOs and Glicko have Fedor at #7 or #8, is that because he amassed such a high rating at one stage that he still hasnt been whittled down to size yet?

[JCS]

Added WHR on Fight Matrix. We now display 4 purely mathematical systems... two slightly different applications of Elo, Glicko-1 and now WHR.

I see that WHR still has Cormier as the #1 HW

So what is the "standard" ranking on Fight Matrix?

Also I notice the ELOs and Glicko have Fedor at #7 or #8, is that because he amassed such a high rating at one stage that he still hasnt been whittled down to size yet?

Standard is homegrown... It's something like a hybrid between the old and new BoxRec systems... probably a little more practical than analytical... it is an Elo-like core, with a bunch of shit wrapped around it to handle various scenarios.

You are mostly correct on Fedor.. but also that his losses have all been to high-quality opponents... stretching that definition for Mitrione, but he was a fringe Top 10 guy as well...

[SportsRatings]

You are mostly correct on Fedor.. but also that his losses have all been to high-quality opponents... stretching that definition for Mitrione, but he was a fringe Top 10 guy as well...

I keep forgetting that technically he beat Maldonado, too, so he's won 8 of his last 10 fights.

[JCS]

Here's a fairly ridiculous one.

https://boxrec.com/en/proboxer/813592

The guy had a higher rating after his pro debut than he has after his last fight... and he is 9-0, 9KO??

[computerrank]

Here's a fairly ridiculous one.

https://boxrec.com/en/proboxer/813592

The guy had a higher rating after his pro debut than he has after his last fight... and he is 9-0, 9KO??

No, he hadn't.

[computerrank]

...
Furthermore, WHR seems much more susceptible to rating inflation by way of can crushing... most notably, the "smart" kind that you see in eastern Europe and Russia.... where matchmakers often pay for fighters that are building glossy records elsewhere, to get beat by their fighters that are doing the same. Due to its nature of iterating and analyzing the opponent chains before and after bouts, it seems that winning records are much, much more important than opponent quality.
...

Boxrec WHR has an extension to use a 'connected' parameter for the boxers. Only bouts with both opponents connected to the community have full value. Other fights only have value of 10 percent.
The second extension in Boxrec WHR is, boxers are seeded at different initial ratings for different initial weight divisions. Heavyweights are seeded higher, Flyweights are seeded lower.
So in Boxrec WHR you will not find tomato can crushers up in the ratings.

[JCS]

...
Furthermore, WHR seems much more susceptible to rating inflation by way of can crushing... most notably, the "smart" kind that you see in eastern Europe and Russia.... where matchmakers often pay for fighters that are building glossy records elsewhere, to get beat by their fighters that are doing the same. Due to its nature of iterating and analyzing the opponent chains before and after bouts, it seems that winning records are much, much more important than opponent quality.
...

Boxrec WHR has an extension to use a 'connected' parameter for the boxers. Only bouts with both opponents connected to the community have full value. Other fights only have value of 10 percent.
The second extension in Boxrec WHR is, boxers are seeded at different initial ratings for different initial weight divisions. Heavyweights are seeded higher, Flyweights are seeded lower.
So in Boxrec WHR you will not find tomato can crushers up in the ratings.

Understood, but from a pure mathematical sense, WHR is much more susceptible to this problem.

[computerrank]

Boxrec WHR has an extension to use a 'connected' parameter for the boxers. Only bouts with both opponents connected to the community have full value. Other fights only have value of 10 percent.
The second extension in Boxrec WHR is, boxers are seeded at different initial ratings for different initial weight divisions. Heavyweights are seeded higher, Flyweights are seeded lower.
So in Boxrec WHR you will not find tomato can crushers up in the ratings.

Understood, but from a pure mathematical sense, WHR is much more susceptible to this problem.

Elo, Glicko, WHR with no specific corrections, all show the same problems.

[felinoboxing]

Based on its formula, throughout history, which and how many boxers have been P4P # 1 ?.

[felinoboxing]

I'm still with P4P # 1. At the end of 1956, # 1 was Spider Webb with 285.9, second Floyd Patterson 260.3, 3 is Basil 243.4 .... Am I wrong?

[gilgamesh]

I'm still with P4P # 1. At the end of 1956, # 1 was Spider Webb with 285.9, second Floyd Patterson 260.3, 3 is Basil 243.4 .... Am I wrong?

Ring Magazine I would imagine would be your source for this sorta thing. I'm not sure which years had a comprehensive Pound for Pound list, but I'll see if I can dig some up.

[gilgamesh]

I'm still with P4P # 1. At the end of 1956, # 1 was Spider Webb with 285.9, second Floyd Patterson 260.3, 3 is Basil 243.4 .... Am I wrong?

This page shows you annual Ratings by year.

https://boxrec.com/media/index.php/The_ ... al_Ratings

They didn't start doing an Annual P4P ratings thing until 1989. So your best bet would be just to study who the Champions and Top contenders are of each Calendar year, do your research into their records, and what footage you can find of them, and draw your own conclusions.