BoxRec

Posted: **17 Aug 2006, 06:11**

You didn't answer the question.

The winner of Briggs vs Minto becomes number 1?

If Briggs vs Wlad or Byrd or Valuev is 50/50 according to your predictive ratings, then great, he can be the number one.

But Briggs beats "46yo Tony Tubbs victim" Minto and he's number one?

You think that's OK? You as a boxing fan, think that's OK? What's your answer?

conan

Posted: **17 Aug 2006, 06:29**

Sorry Conan i did not realise it was you, and apologies to JCS.

I dont thnk im going to have enough time to do a complete 8 week study, as far as I can see. the upsets have been 2 to boxrec 1 to IBO lately excluding common upsets.

Manny Siaca was rated higher in IBO than Branco but vice versa in boxrec, we know Branco won

Roy Jones was higher than Ajamu in IBO, etc and Ouma was ahead of powell. I cant see any other differences in fights although there have been a lot of upsets recently;
Romero bt Viloria nashiro bt Castillo and Koshimoto got beat, prediction rate probably under 50% for IBO & boxrec at the moment for top 30

Posted: **17 Aug 2006, 06:38**

computerrank wrote:conan,

It may be painful to hear it over and over again, but hey, answer the question above as a boxing fan, and you'll see why the questions will be repeated.

I very carefully follow everthing regarding boxing - and sometimes follow the top fights, if available on television. But I don't find it necessary to discuss in the current scene forum in threads nearly all concerning "I hate a boxer", "I love a boxer", "I am absolutley sure, ...", "you are an idiot ..." - I simply find it quite boring to spend my time with this.

By this, it is also not difficult to see, what you saying with the Brocks, Briggs, Dawsons etc.

I cant understand your post
I even see the structure behind:

- they all are rated quite high above the the ratings of the boxers, they defeated at all ( this regarding no loss boxers) or in a recent time period ( this regarding Briggs)

- it so is absolutely unapproved, that they are able to defeat these days opponents at their rating level, yes

- yes, I am sure, that not everyone of them will do so - and everyone is suspect in some way to me

- but in spite of my feelings, I can imagine, that 50% of these boxers, will do so - and I found it in the statistics

- this is, what the ratings say, not more - and then, the ratings are correct from a statistical perspective

- I think, the problem behind - for everybody - including me is:

We all will more accept the situation, where a low rated boxer defeats a higher rated one, who already defeated opponents at his level in a recent time period. Altough we all know, this happens quite often.

But we all have problem to accept, where a high rated boxer, who didn't yet defeat an opponent at his level, loses to a lower rated one. Altough this also happens quite often.

We know this exspection issue from psychology - people more accept to not win, even when there is a good chance - but there is also a considerable risk to lose.

And the observer's losing feeling here is, that he hates to have been wrong, and in addition now cannot defend his exspectation afterwards by simple facts e.g. earlier wins against proven opponents.

A way to go would be, to weight the outcomes by this exspectation:

- a loss of a boxer against an opponent rated higher than the boxers best defeated last opponents (in a recent time period) should be regarded as a multiple upset - by multiple weight.

This exactly is, what people makes pain, I think.

Best regards
Martin

Posted: **17 Aug 2006, 06:38**

conan_the_cribber wrote:You didn't answer the question.

The winner of Briggs vs Minto becomes number 1?

If Briggs vs Wlad or Byrd or Valuev is 50/50 according to your predictive ratings, then great, he can be the number one.

But Briggs beats "46yo Tony Tubbs victim" Minto and he's number one?

You think that's OK? You as a boxing fan, think that's OK? What's your answer?

conan

... you didn't spend the time, to read, I think ...

- I do not like it, and find it questionable, in case of Briggs, as in all such cases

- but you are using this arguement very unprofessional and misleading against the rating's substance
I didn't repeat in my last email, but often before - but also everybody interrprets more into this ratings than they can give

- you absolutely mis-interprete this ratings and their substance, if you state a sharp ranking
- such as just done: Briggs can never be #1, cause implictely everbody then thinks, Briggs must kill everybody else
- the rating values for him may be quite in order, even after this win
- the ratings have a variance of about 100 points
- all interpretations within this margin are quite unprofessional
- you have to interprete as: all boxers within a margin of 100 points are quite one cluster

Posted: **17 Aug 2006, 06:52**

I look at someone like Malik Scott who you have rated at 22. I am confused why he is 22, he is so unproven. As a matchmaker the only worry I have (and Ive not seen a tape of him) is that he has good management, I dont think Shelly would waste his time if this chap could not fight. Now im sure that is not factored into your ratings at all, so why is he number 22. He is obviously a boxer not a puncher and seems to have reached a level where he wins widely on points against non top 100 fighters. That means 'so what'. He is still unknown and unproven, what are you seeing in this fighter that I cant. Your ratings say he will beat John Ruiz. I have seen Gonzalo Basile who you also rate above Ruiz. I can pretty much say barring an amazing collapse in Ruiz's form Basile does not win this fight.

Posted: **17 Aug 2006, 07:07**

computerrank wrote:
conan_the_cribber wrote:You didn't answer the question.

The winner of Briggs vs Minto becomes number 1?

If Briggs vs Wlad or Byrd or Valuev is 50/50 according to your predictive ratings, then great, he can be the number one.

But Briggs beats "46yo Tony Tubbs victim" Minto and he's number one?

You think that's OK? You as a boxing fan, think that's OK? What's your answer?

conan
... you didn't spend the time, to read, I think ...

- I do not like it, and find it questionable, in case of Briggs, as in all such cases

- but you are using this arguement very unprofessional and misleading against the rating's substance
I didn't repeat in my last email, but often before - but also everybody interrprets more into this ratings than they can give

- you absolutely mis-interprete this ratings and their substance, if you state a sharp ranking
- such as just done: Briggs can never be #1, cause implictely everbody then thinks, Briggs must kill everybody else
- the rating values for him may be quite in order, even after this win
- the ratings have a variance of about 100 points
- all interpretations within this margin are quite unprofessional
- you have to interprete as: all boxers within a margin of 100 points are quite one cluster

Martin,

I'd like to take some of the tension out of this, if I may try. I understand fully your points. I understand what you see the ratings are, effectively a future performance indicator. I see what you see in the ratings number, it's just a number, that can be used to interpret future success, allowing of course for the home town factor. The predictive quality drops for fighters within a 100 points of each other, as they are even fights. I believe you have spent a lot of time optimizing this and checking the predictive nature of the results. As I said before, I am a statistician, I have a statistics degree, really, I know where you are coming from.

However, the unfortunate thing is, the rating number can be sorted. It is sorted and presented to the public as rankings on this boxing site. Although you may lump Shannon Briggs in with a bunch of other fighters within a 100 pts after a theoretical win against Brian Minto, he will appear at the top of the list at number 1. This is clearly not in anyone's interest, except Shannon Briggs manager.

I don't consider this to be unprofessional to point this out to you as a nasty side effect of the ratings algorithm, especially as you also find the idea ugly. It is also not just I, although I am perhaps the most vocal. Many other people capable of presenting logical arguments are also against this. The people posting on this thread are not of the nature you mentioned before i.s. "I'm a Toney fan, go fornicate yourself", "fornicate you too". The people posting here present meaningful arguments as to why the current use of the predictive ratings as rankings is wrong.

The point about Briggs killing everyone is not true. I expect the number one position to be earned and in boxing, everyone is just one punch away from losing. I dont not expect the number one to kill everyone. However, even if he beat Minto, there's no way Briggs would start favourite against a host of other boxers.

Although bouyed by the support of others here, I am running out of enthusiasm in this discussion. In the end it boils down to the same old yes or no question.

Are you prepared to change the algorithm so that the outliers are less visible with the possible consequence that the prediction quality drops by 1 or 2% ?

conan

Posted: **17 Aug 2006, 07:59**

conan_the_cribber wrote:
computerrank wrote:
conan_the_cribber wrote:You didn't answer the question.

Are you prepared to change the algorithm so that the outliers are less visible with the possible consequence that the prediction quality drops by 1 or 2% ?

conan

I bet I know the answer to that ;)

Posted: **17 Aug 2006, 08:06**

JCS83MD wrote:
conan_the_cribber wrote:
computerrank wrote: I bet I know the answer to that ;)

... so it is ...

But the Boxrec Traditional Rankings will be published soon ... for all those, who need the old flavour.

Thanks and best regards
Martin

Posted: **17 Aug 2006, 08:21**

Conan,

I always appreciate the discussions on this forum - in my last message I only responded to your question concerning current scene forum. Sorry, there was absolutely no offence intended.

I also appreciate a clear language, you know, every perspective allowed.

And so to your last question, the BoxRec Current Ratings will strictly follow the predicition rate measure as the single evaluation criteria - and withstand the critics.

This is not in contradiction to my very clear understanding of your valid considerations regarding possible interpretations.

But we will have the Traditional Rankgs now too - and also for this, thanks for your contributions and help.

Have a nice day and stay alive here
Martin

Posted: **17 Aug 2006, 08:37**

Lennox wrote:I look at someone like Malik Scott who you have rated at 22. I am confused why he is 22, he is so unproven. As a matchmaker the only worry I have (and Ive not seen a tape of him) is that he has good management, I dont think Shelly would waste his time if this chap could not fight. Now im sure that is not factored into your ratings at all, so why is he number 22. He is obviously a boxer not a puncher and seems to have reached a level where he wins widely on points against non top 100 fighters. That means 'so what'. He is still unknown and unproven, what are you seeing in this fighter that I cant. Your ratings say he will beat John Ruiz. I have seen Gonzalo Basile who you also rate above Ruiz. I can pretty much say barring an amazing collapse in Ruiz's form Basile does not win this fight.

Lennox,

I think you read all the messages - so you will understand, that Malik or Basile simply collected the points for this rating - the program simply uses the facts from the records - no insight from exsperts.

Regarding their position to Ruiz, these guys around him in the ratings are regarded to have a chance at minimum 50% against him - not that they will win definitely, because they are 2 ranks in lead. See, the difference is only some points and the uncertainty is about 100 points.

Best regards
Martin

P.S.: I sent a request to IBO, whether they would give me their ratings formula for a comparision of the prediciton rate. Answer: They will not publish their methods - whatever they do really.

So for me, it would mean too much effort to get a valid result by hand comparision. I will not follow-up, sorry.

Posted: **17 Aug 2006, 17:54**

Hi, now that we've cleared up the fact that it's pointless to try and change the predictive rankings, regardless of the outliers, it's on to the next problem child, the 'traditional rankings' or as I now prefer to call them 'who beat who' rankings.

Trouble ahead....why? Johnny bloody Nelson. Well let's say, the Johnny Nelsons of the world.

Let's say, Johnny Nelson makes it to number five and his and then via attrition make it to number one in the Cruiser division. All the other top Cruisers, Jirov, Toney, JC Gomez all moved out. He now goes on to fight a bunch of fighters around the bottom of the top 45. He's got the number one spot by default.

Meanwhile, Mormeck, Braithwaite, Jones, Cunningham and O'Neil Bell all show stellar form and have a nice old bash up to decide who really is the champ. Bell prevails, has unified the three major belts and by all accounts is the man to beat in the division. Which rank does he get? You guessed it, the best he can do despite kicking butt at the highest level is number 2. The current number one, got there earlier, and not by beating the best, rather the fifth best. Yet the number one cannot be dethroned by Bell, because........

..... the quality of opponents is only indirectly rewarded. Johnny Nelson could be Gary St Clair or some other people.

Although I view the 'who beat who' rankings as a fast improvement on the predictive rankings, they are fundamentally flawed. What I'm really looking for is the IBO rankings. Call them what you will. The best fighters, with the best form and the biggest wins against the top contendors should rise to the top of the ratings. That's what i'm looking for. That's what I interpret as 'traditional'.

conan

Posted: **17 Aug 2006, 18:09**

Its possible to change the predictive ratings. I just put a conditional calculation in today that better handles fighters who are coming off a long layoff. In testing now, will probably be in a future update. Martin appears to have a couple changes in as well.

:)

Posted: **18 Aug 2006, 06:06**

JCS83MD wrote:Its possible to change the predictive ratings. I just put a conditional calculation in today that better handles fighters who are coming off a long layoff. In testing now, will probably be in a future update. Martin appears to have a couple changes in as well.

:)

Well you know what I meant. Change it, in the sense to eliminate the very visible outliers.

conan

Posted: **18 Aug 2006, 08:17**

conan_the_cribber wrote:
JCS83MD wrote:Its possible to change the predictive ratings. I just put a conditional calculation in today that better handles fighters who are coming off a long layoff. In testing now, will probably be in a future update. Martin appears to have a couple changes in as well.

:)
Well you know what I meant. Change it, in the sense to eliminate the very visible outliers.

conan

It has to be months ago now when you said you weren't gonna spend much time on this whole fiasco and not make anymore than a few posts.

Now, you've only spent hours posting about the system and now i have to ask... if you are a statistician, why not analyze the calculations we're doing and submit some REAL suggestions in terms of altering the system we use?

Posted: **18 Aug 2006, 12:12**

Those look pretty good. The 50% inactivity hit seems kinda steep... how would 33.3% look?

Posted: **18 Aug 2006, 12:18**

JCS83MD wrote:Those look pretty good. The 50% inactivity hit seems kinda steep... how would 33.3% look?

I worked around with such, but you must keep in mind - a boxer can more than double his rating in 1 bout ...

you cannot compare this with current ratings - they are based on point diferrences - this achievment rating is based on point ratios - you see the wide ranges and big steps ...

Posted: **19 Aug 2006, 10:16**

... but this is only valid for the top ranks ...

below 250 points, we have a wild mix of good and bad boxers ....

So for the traditionals the only solution seems to be the Traditional Ranking - the prediction rate is quite low with 62% - but the ranking is quite ok from the perspective of separation of good and bad boxers ...

Posted: **19 Aug 2006, 11:12**

Cobwebcat wrote:Winner Always above loser???? Nooooooo!

Absolutely essential I'd say. Ratings look better Martn.

Posted: **19 Aug 2006, 12:51**

76.9% for a somewhat traditional ranking? Hey thats pretty good.. may definitely be able to push that up with some tweaking.

Posted: **20 Aug 2006, 15:07**

Cobwebcat wrote:
Lennox wrote:
Cobwebcat wrote:Winner Always above loser???? Nooooooo!
Absolutely essential I'd say. Ratings look better Martn.
Absolutely NOT

Tell me any other respected sports ranking that has this!!?

Oh for friggen fvcks sake. Wake up cobwebcat. We're talking about boxing and not friggen Manchester City. For ever and a day in boxing, you beat a man, you're above him. Always was always will be. Hell, it's even part of the Muhammed Ali Act. It's part of boxing, if you dont know that part, then you may as well question why the boxers are wearing gloves and cant hit low.

conan

Posted: **20 Aug 2006, 15:20**

Hi Martin,

Those performance rankings are so friggen hot. They are simple, easy to understand and fair. A brilliant, simple, algorithm, much better than the complex stuff I was too intimidated to implement. All credit to you.

The results are excellent. Hell, you even get a decent p4p out of it, for nothing. Every division looks good.

Great work. The lower rankings might be very dynamic, but that's ok too, in boxing you are only as good as your last win.

The prediction rates are incredible. It is conceivable, that you could get within one or two percentage points of the 'prediction rankings'. Maybe, even overtake them. Now that would be a win-win for everyone.

One small suggestion, it goes back to the weight change stuff. I still believe the transition from Cruiser to Heavyweight to be the hardest. Maybe you could implement a special transition for this, so that the performance ranks of the cruisers coming into heavy is more accurate and the predictability increases.

cheers

conan

Posted: **20 Aug 2006, 16:02**

conan_the_cribber wrote:Hi Martin,

Those performance rankings are so friggen hot. They are simple, easy to understand and fair. A brilliant, simple, algorithm, much better than the complex stuff I was too intimidated to implement. All credit to you.

The results are excellent. Hell, you even get a decent p4p out of it, for nothing. Every division looks good.

Great work. The lower rankings might be very dynamic, but that's ok too, in boxing you are only as good as your last win.

The prediction rates are incredible. It is conceivable, that you could get within one or two percentage points of the 'prediction rankings'. Maybe, even overtake them. Now that would be a win-win for everyone.

One small suggestion, it goes back to the weight change stuff. I still believe the transition from Cruiser to Heavyweight to be the hardest. Maybe you could implement a special transition for this, so that the performance ranks of the cruisers coming into heavy is more accurate and the predictability increases.

cheers

conan

I believe we found through statistics that the move from Cruiser to Heavy was not a very significant change. Light Heavy to Cruiser constitutes the largest change based on the maximum weights of the division and these are what we used to calculate rating transitions.

Posted: **20 Aug 2006, 16:18**

Hi,

from Martin's list, here is the p4p.

conan

1 M Middleweight 1929 Ronald Wright
2 M Middleweight 1877 Jermain Taylor
3 M Middleweight 1678 Bernard Hopkins
4 M Super Featherweight 1659 Marco Antonio Barrera
5 M Welterweight 1657 Floyd Mayweather Jr
6 M Welterweight 1479 Ricky Hatton
7 M Super Featherweight 1241 Manny Pacquiao
8 M Welterweight 1226 Carlos Manuel Baldomir
9 M Light Welterweight 1164 Jose Luis Castillo
10 M Super Middleweight 1133 Joe Calzaghe
11 M Light Heavyweight 1132 Antonio Tarver
12 M Light Middleweight 1117 Cory Spinks
13 M Welterweight 1114 Antonio Margarito
14 M Light Middleweight 1114 Shane Mosley
15 M Bantamweight 1096 Rafael Marquez
16 M Heavyweight 1065 Wladimir Klitschko
17 M Light Heavyweight 1043 Glen Johnson
18 M Cruiserweight 1034 O'Neil Bell
19 M Heavyweight 971 Nikolay Valuev
20 M Light Welterweight 958 Isaac Hlatshwayo
21 M Featherweight 951 Chris John

Posted: **20 Aug 2006, 18:08**

conan_the_cribber wrote:Hi,

from Martin's list, here is the p4p.

conan

1 M Middleweight 1929 Ronald Wright
2 M Middleweight 1877 Jermain Taylor
3 M Middleweight 1678 Bernard Hopkins
4 M Super Featherweight 1659 Marco Antonio Barrera
5 M Welterweight 1657 Floyd Mayweather Jr
6 M Welterweight 1479 Ricky Hatton
7 M Super Featherweight 1241 Manny Pacquiao
8 M Welterweight 1226 Carlos Manuel Baldomir
9 M Light Welterweight 1164 Jose Luis Castillo
10 M Super Middleweight 1133 Joe Calzaghe
11 M Light Heavyweight 1132 Antonio Tarver
12 M Light Middleweight 1117 Cory Spinks
13 M Welterweight 1114 Antonio Margarito
14 M Light Middleweight 1114 Shane Mosley
15 M Bantamweight 1096 Rafael Marquez
16 M Heavyweight 1065 Wladimir Klitschko
17 M Light Heavyweight 1043 Glen Johnson
18 M Cruiserweight 1034 O'Neil Bell
19 M Heavyweight 971 Nikolay Valuev
20 M Light Welterweight 958 Isaac Hlatshwayo
21 M Featherweight 951 Chris John

Don't want tot start anything, but didn't you s ay yourself earlier you don't deal with p4p rankings because they're too difficult?

Posted: **20 Aug 2006, 19:15**

[quote="Cobwebcat"]Conan, if you are truly a statitician you should know that you are talking nonsense.

There are many sports that have one on one and none of them use the ridiclous system that if you win one fight/game/match whatever then you are the best...no matter who you beat.

Just because boxing has belts and proclaims someone "The Champ" because he "beat the man" does not mean this has any basis in statistical analysis. James Buster Douglas beat Tyson when Tyson had the clap and hardly trained, but did that make him the best fighter in the world?

I thought you had some interesting points re the predictive ranking but all I see now is you getting Martin to produce a ranking system that is flawed beyond belief and you cant take the same criticism of yours you were so quick to give out to other peoples.

This new "ranking" is based on a ladder system used by club squash players for god sakes.

Any system is open to scrutiny. Yours is no different and is badly flawed. Just because it might seem popular, mainly because people actually want to see a form guide, does not make it accurate.

Martin/ JCS: I'm disappointed in you guys! Didn't getting rid of this A beats B nonsense improve your predictive rankings? Shouldn't it be disregarded in any meaningful ranking traditional or otherwise?

Conan: Congratulations on being very persuasive but you are a hypocrite. If you cant take it, don't dish it out. I'm afraid you can see other people's elephants but are terrible at spotting your own.

Having said all that I hope the three of you hit on something and prove me wrong but just as Conan stood on the outside pointing out possible flaws in the predictive system, some of which were good points, I believe he cant see the wood for the trees either now but hey, he's got his own way so who cares right?

Boxing is very different to golf, tennis or sports that teams play weekly. In boxing they may only have 2 contests per year so they are more important. Boxing will always be A beats B beats C etc. The victor until he loses must always be ahead unless the loser makes a signifcant victory of someone even higher in the rankings.

BoxRec

Ratings - please read before commenting - Archived

Johnny Nelson.