Ratings - please read before commenting - Archived

[JCS]

But why are even mention weight jumping? For all practical purposes, here we have career WWs and BWs. I see Lujan hasn't fought above WW in 4 years. Mike Jones is a WW prospect cum contender, with no noteworthy fights above. Moreno, Mares, Agbeko are all established champs at BW, have not moved.

So when you say that points are division adjusted, is this only with respect to moving up/down classes? Or will a cruiser end up with way more points than a flyweight for "comparable" achievement, provided neither has moved up/down?

Welterweight is a historically stronger division. More fighters, more points. More good fighters, more points.

Ok, sorry to keep harping on this. It might be good to factor in the current relative strength of the division, else guys like Sturm and Geale would figure in p4p. And the super 6 guys would be unfairly penalized for the historical weakness of their div. I think bantam has been better, or at least deeper than welter for the past few years.

I think the "pound for pound" rating on the main page is a little bit misleading.. as it's not really "pound for pound" in the minds of the public.

[marcianofan]

But why are even mention weight jumping? For all practical purposes, here we have career WWs and BWs. I see Lujan hasn't fought above WW in 4 years. Mike Jones is a WW prospect cum contender, with no noteworthy fights above. Moreno, Mares, Agbeko are all established champs at BW, have not moved.

So when you say that points are division adjusted, is this only with respect to moving up/down classes? Or will a cruiser end up with way more points than a flyweight for "comparable" achievement, provided neither has moved up/down?

Welterweight is a historically stronger division. More fighters, more points. More good fighters, more points.

Ok, sorry to keep harping on this. It might be good to factor in the current relative strength of the division, else guys like Sturm and Geale would figure in p4p. And the super 6 guys would be unfairly penalized for the historical weakness of their div. I think bantam has been better, or at least deeper than welter for the past few years.

Welter features more than 50% more active fighters than does Bantam, though. More fighters = more fights = more points in the system = more points won from fights high on the food chain. Just look at heavyweight- its contenders have probably the most points of any other division, and yet all we hear is how awful the division is. It's not that Heavyweights are better in a pound for pound or any other sense, but there are simply more heavyweights fighting and thus more points available. Think of it like an ecosystem with a food chain. If there is more grass for the rabbits and more rabbits for the birds of prey, the birds of prey are going to tend to be better fed. In comparing bantam to welter, there is more welterweight grass (ie bums and guys who never beat anybody significant) and more welterweight rabbits (mid-range fighters/journeymen) which leads to a larger and better point-fed group of welterweight birds of prey (contenders/fringe contenders).

[Purse Bid Shakedown]

Welterweight is a historically stronger division. More fighters, more points. More good fighters, more points.

Ok, sorry to keep harping on this. It might be good to factor in the current relative strength of the division, else guys like Sturm and Geale would figure in p4p. And the super 6 guys would be unfairly penalized for the historical weakness of their div. I think bantam has been better, or at least deeper than welter for the past few years.

Welter features more than 50% more active fighters than does Bantam, though. More fighters = more fights = more points in the system = more points won from fights high on the food chain. Just look at heavyweight- its contenders have probably the most points of any other division, and yet all we hear is how awful the division is. It's not that Heavyweights are better in a pound for pound or any other sense, but there are simply more heavyweights fighting and thus more points available. Think of it like an ecosystem with a food chain. If there is more grass for the rabbits and more rabbits for the birds of prey, the birds of prey are going to tend to be better fed. In comparing bantam to welter, there is more welterweight grass (ie bums and guys who never beat anybody significant) and more welterweight rabbits (mid-range fighters/journeymen) which leads to a larger and better point-fed group of welterweight birds of prey (contenders/fringe contenders).

So it's not a weight bias, but quantity. And seems to boil down to the same thing. Either way, if star ratings are to be useful, this would merit a tweak. I mean, 2 top bantam fights, excellent division, vs Jones - Lujan? Or does welter go to 11?

[conan_the_cribber]

Ok, sorry to keep harping on this. It might be good to factor in the current relative strength of the division, else guys like Sturm and Geale would figure in p4p. And the super 6 guys would be unfairly penalized for the historical weakness of their div. I think bantam has been better, or at least deeper than welter for the past few years.

Welter features more than 50% more active fighters than does Bantam, though. More fighters = more fights = more points in the system = more points won from fights high on the food chain. Just look at heavyweight- its contenders have probably the most points of any other division, and yet all we hear is how awful the division is. It's not that Heavyweights are better in a pound for pound or any other sense, but there are simply more heavyweights fighting and thus more points available. Think of it like an ecosystem with a food chain. If there is more grass for the rabbits and more rabbits for the birds of prey, the birds of prey are going to tend to be better fed. In comparing bantam to welter, there is more welterweight grass (ie bums and guys who never beat anybody significant) and more welterweight rabbits (mid-range fighters/journeymen) which leads to a larger and better point-fed group of welterweight birds of prey (contenders/fringe contenders).

So it's not a weight bias, but quantity. And seems to boil down to the same thing. Either way, if star ratings are to be useful, this would merit a tweak. I mean, 2 top bantam fights, excellent division, vs Jones - Lujan? Or does welter go to 11?

As I've written before, this is the major flaw in the p4p rankings. There is point inflation in a division because when fighter A fights fighter B, ptsA+PtsB is bigger after the fight than before. This point inflation is only countered by inactivity and lack of good opponents. As such, divisions with the most fighters e.g. welter and heavy produce fighters at the top of the p4p rankings.

The only solutions are to introduce either
a) a zero-sum game for fights. This is difficult when an upset occurs
b) for p4p comparison, use a factor to adjust the individual weight class ratings based upon the number of fights in the db in a particular weight class.

Your complaint is legitimate.

conan

[computerrank]

As I've written before, this is the major flaw in the p4p rankings. There is point inflation in a division because when fighter A fights fighter B, ptsA+PtsB is bigger after the fight than before. This point inflation is only countered by inactivity and lack of good opponents. As such, divisions with the most fighters e.g. welter and heavy produce fighters at the top of the p4p rankings.

The only solutions are to introduce either
a) a zero-sum game for fights. This is difficult when an upset occurs
b) for p4p comparison, use a factor to adjust the individual weight class ratings based upon the number of fights in the db in a particular weight class.

Your complaint is legitimate.

conan

The division top ratings (= top ratings for p4p) do not correlate with the number of active boxers in the divsions:

Code: Select all

+---------------------+-------+-------+
| division            | top_r | boxers|
+---------------------+-------+-------+
| Heavyweight         |  1331 | 1101  |
| Cruiserweight       |   571 |  940  |
| Light Heavyweight   |  1020 |  840  |
| Super Middleweight  |   903 |  877  |
| Middleweight        |   973 | 1293  |
| Light Middleweight  |   760 | 1252  |
| Welterweight        |  1385 | 1545  |
| Light Welterweight  |   862 | 1485  |
| Lightweight         |   963 | 1581  |
| Super Featherweight |   444 | 1173  |
| Featherweight       |   582 | 1303  |
| Super Bantamweight  |   556 | 1041  |
| Bantamweight        |   493 |  950  |
| Super Flyweight     |   322 |  668  |
| Flyweight           |   353 |  728  |
| Light Flyweight     |   236 |  510  |
| Minimumweight       |   258 |  304  |
+---------------------+-------+-------+

A zero-sum game for fights would not work - as the sum of all points would steadily shrink
- by the loss of retiring boxers, carrying their points out
- while new boxers join the business with 0 points

[jujigatame]

Is something wrong with the pre/post ratings displays on the right side of each fighter page? Right now just looking at Yhonny Perez's page, his first Agbeko fight has "315 470 367 239" as the pre/post ratings but right below it, the Silence Mabuza fight has "0 0 0 0". I am seeing a few cases of this on almost every fighter page now.

[computerrank]

Is something wrong with the pre/post ratings displays on the right side of each fighter page? Right now just looking at Yhonny Perez's page, his first Agbeko fight has "315 470 367 239" as the pre/post ratings but right below it, the Silence Mabuza fight has "0 0 0 0". I am seeing a few cases of this on almost every fighter page now.

Sorry, Silence Mabuza fight has "154 315 276 170". ???

[m1kee50]

What was the first fight to be held in the super-middle division, i.e 168lbs?
Not neccessarily a title fight... Think I might have to post this where martin can see it too.

[computerrank]

What was the first fight to be held in the super-middle division, i.e 168lbs?
Not neccessarily a title fight... Think I might have to post this where martin can see it too.

I guess, the Boxrec records look a bit strange for bouts assigned to super middleweight:

Code: Select all

+------------+---------------------+---------------------+
| date       | boxer_a             | boxer_b             |
+------------+---------------------+---------------------+
| 1867-03-05 | Joe Goss            | Tom Allen           |
| 1872-02-01 | Charley Davis       | Abe Daultry         |
| 1872-02-01 | Charley Davis       | Bat Mullins         |
| 1872-02-01 | Bat Mullins         | Jim Gollagher       |
| 1876-03-20 | James Clocker Keogh | Jack Boylan         |
| 1878-01-28 | Jack Boylan         | James Clocker Keogh |
| 1878-12-02 | Larry Foley         | Peter Newton        |
| 1878-12-27 | Jack Boylan         | ? Merrin            |
| 1879-05-26 | Denny Harrington    | Florrie Barnett     |
| 1879-11-27 | Alf Greenfield      | Denny Harrington    |
+------------+---------------------+---------------------+

From my perspective, super middleweight assignment should be used not earlier than from the 1970s.

[squiggy]

I haven't noticed a fight actually receiving the Super Middleweight designation on Boxrec earlier than Rocchigiani Malinga in 89.

[SportsRatings]

What was the first fight to be held in the super-middle division, i.e 168lbs?
Not neccessarily a title fight... Think I might have to post this where martin can see it too.

I guess, the Boxrec records look a bit strange for bouts assigned to super middleweight:

Code: Select all

+------------+---------------------+---------------------+
| date       | boxer_a             | boxer_b             |
+------------+---------------------+---------------------+
| 1867-03-05 | Joe Goss            | Tom Allen           |
| 1872-02-01 | Charley Davis       | Abe Daultry         |
| 1872-02-01 | Charley Davis       | Bat Mullins         |
| 1872-02-01 | Bat Mullins         | Jim Gollagher       |
| 1876-03-20 | James Clocker Keogh | Jack Boylan         |
| 1878-01-28 | Jack Boylan         | James Clocker Keogh |
| 1878-12-02 | Larry Foley         | Peter Newton        |
| 1878-12-27 | Jack Boylan         | ? Merrin            |
| 1879-05-26 | Denny Harrington    | Florrie Barnett     |
| 1879-11-27 | Alf Greenfield      | Denny Harrington    |
+------------+---------------------+---------------------+

From my perspective, super middleweight assignment should be used not earlier than from the 1970s.

Interesting ! Are fights all designated with a weight class now? Where does one see this on the fight page ?

[squiggy]

A fighter's results page shows a "division" column between each fight's results and the Boxrec rankings. (Recently I looked at the site on my parents' computer, and that column wouldn't display... otherwise I always see it.) I'm sure it isn't meant to be gospel truth in every case, but I think it's interesting.

[squiggy]

So you can spot little strange things like how Duilio Loi's last two fights were apparently welterweight fights for a light welterweight title belt.

[m1kee50]

Any chance that you could tell me the biggest bouts to have taken part in the location United Kingdom >> Merseyside?

[computerrank]

Any chance that you could tell me the biggest bouts to have taken part in the location United Kingdom >> Merseyside?

Code: Select all

+------------+--------------+-------------+---------------+-------------------------+--------------------+----------------+---------------------+------+
| date       | country_code | region_name | towncity_name | venue_name              | division           | boxer_a        | boxer_b             | r    |
+------------+--------------+-------------+---------------+-------------------------+--------------------+----------------+---------------------+------+
| 1940-09-21 | UK           | MSY         | Liverpool     | Anfield Football Ground | Welterweight       | Ernie Roderick | Eric Boon           | 1102 |
| 1958-03-27 | UK           | MSY         | Liverpool     | The Stadium             | Middleweight       | Dick Tiger     | Pat McAteer         | 1010 |
| 1935-06-27 | UK           | MSY         | Liverpool     | Anfield Football Ground |                    | Freddie Miller | Seaman Tommy Watson |  914 |
| 1954-04-29 | UK           | MSY         | Liverpool     | The Stadium             |                    | Jean Sneyers   | Hogan Kid Bassey    |  862 |
| 1940-11-02 | UK           | MSY         | Liverpool     | The Stadium             | Featherweight      | Nel Tarleton   | Tom Smith           |  845 |
| 1932-11-10 | UK           | MSY         | Liverpool     | The Stadium             | Featherweight      | Nel Tarleton   | Seaman Tommy Watson |  807 |
| 1951-04-19 | UK           | MSY         | Liverpool     | The Stadium             | Welterweight       | Alf Danahar    | Wally Thom          |  785 |
| 1955-06-16 | UK           | MSY         | Liverpool     | The Stadium             | Middleweight       | Pat McAteer    | Johnny Sullivan     |  784 |
| 1934-09-20 | UK           | MSY         | Liverpool     | Anfield Football Ground | Featherweight      | Nel Tarleton   | Freddie Miller      |  777 |
| 1957-07-25 | UK           | MSY         | Liverpool     | The Stadium             | Super Middleweight | Leen Jansen    | Pat McAteer         |  766 |
+------------+--------------+-------------+---------------+-------------------------+--------------------+----------------+---------------------+------+

since 2000

Code: Select all

+------------+--------------+-------------+---------------+----------------------------+--------------------+-------------------+-------------------+------+
| date       | country_code | region_name | towncity_name | venue_name                 | division           | boxer_a           | boxer_b           | r    |
+------------+--------------+-------------+---------------+----------------------------+--------------------+-------------------+-------------------+------+
| 2007-03-10 | UK           | MSY         | Liverpool     | Olympia                    | Light Welterweight | Souleymane M'baye | Andriy Kotelnyk   |  279 |
| 2011-07-16 | UK           | MSY         | Liverpool     | Echo Arena                 | Lightweight        | John Murray       | Kevin Mitchell    |  208 |
| 2011-10-15 | UK           | MSY         | Liverpool     | Echo Arena                 | Light Heavyweight  | Nathan Cleverly   | Tony Bellew       |  203 |
| 2011-10-15 | UK           | MSY         | Liverpool     | Echo Arena                 | Super Middleweight | Piotr Wilczewski  | James DeGale      |  154 |
| 2010-03-12 | UK           | MSY         | Liverpool     | Echo Arena                 | Welterweight       | Kell Brook        | Krzysztof Bienias |  152 |
| 2010-12-11 | UK           | MSY         | Liverpool     | Echo Arena                 | Light Heavyweight  | Nathan Cleverly   | Nadjib Mohammedi  |  143 |
| 2008-07-04 | UK           | MSY         | Liverpool     | Everton Park Sports Centre | Light Welterweight | David Barnes      | Barry Morrison    |  135 |
| 2001-09-25 | UK           | MSY         | Liverpool     | Everton Park Sports Centre | Light Middleweight | Richard Williams  | Andrew Murray     |  135 |
| 2010-12-11 | UK           | MSY         | Liverpool     | Echo Arena                 | Super Middleweight | Paul Smith        | James DeGale      |  119 |
| 2012-01-21 | UK           | MSY         | Liverpool     | Olympia                    | Heavyweight        | David Price       | John McDermott    |  119 |
+------------+--------------+-------------+---------------+----------------------------+--------------------+-------------------+-------------------+------+

[m1kee50]

Whats the chances of the ratings pages having an option to see only a boxers points gained at a certain weight... i.e at the minute Roy Jones Jr does not appear at SMW, but could we have it so there is the option to see where he would rate (say, all-time) if only his (and everyone else for whom it applied) SMW points were given?

[m1kee50]

??

[Fritz Werner]

In the rateing area, I feel that it should be noted somewhere when a Boxer who (Example) is rated as a Bantam Weight steps up and fights as a Super Bantam Weight or higher. In other words the number of times he or she has fought out of his Weight Class. This provides those and others such as myself (Boxing Judge) a good perpective of that individual. Please note that some Judges as myself do detailed background research on each fighter more especially those that are involved in High Profile bouts when they receive judgeing assignments. I also feel that you folks that run/work BOXREC,COM are doing a Noteworthy job, please continue to march as I view and use it/you on a daily basis.

Thankyou & Respectfully, H.J."Fritz" Werner Lt Col Marines Retired CSAC Lic./IBF Boxing Judge

[JCS]

Whats the chances of the ratings pages having an option to see only a boxers points gained at a certain weight... i.e at the minute Roy Jones Jr does not appear at SMW, but could we have it so there is the option to see where he would rate (say, all-time) if only his (and everyone else for whom it applied) SMW points were given?

That would be an epic effort to calculate that.. Just saying.


On another topic, I have an interesting query.

Judges with more than 100 fights judged, who are the opposing judge the most often, in majority or split decisions? Let's limit it to fights 10 or more rounds.

[turbostar]

I've observed something with regards to active heavyweights in South America - they're very over-rated. Is there anything that can be done about this?

Fabio Eduardo Moli, for example, has 198 points and is ranked #37 in the HW division - yet he's been stopped by the likes of Skelton and Bydenko, and that was in his prime younger days, he's 43 now.

Likewise, George Arias has lost decisively to every non-South American fighter he's faced, including some average names like Audley Harrison, yet is still ranked at #60.
http://boxrec.com/list_bouts.php?human_ ... &cat=boxer

It's obvious why - the fighters in the region stay active and the likes of Moli are able to assert their dominance over the others, it's just that they rarely fight opponents from outside the region and when they do they are almost always easily beaten, often by fighters well below them in the rankings (take Szpilka KO4 Basile in June 2012, Szpilka totally dominated that fight before the early stoppage but beforehand Basile had 26 points more).

Is there anything that can be done about this, or do we just have to accept that South American fighters (particularly heavyweights) can attain an artificially high ranking by staying active and beating the best fighters in their region?

[JCS]

I've observed something with regards to active heavyweights in South America - they're very over-rated. Is there anything that can be done about this?

Fabio Eduardo Moli, for example, has 198 points and is ranked #37 in the HW division - yet he's been stopped by the likes of Skelton and Bydenko, and that was in his prime younger days, he's 43 now.

Likewise, George Arias has lost decisively to every non-South American fighter he's faced, including some average names like Audley Harrison, yet is still ranked at #60.
http://boxrec.com/list_bouts.php?human_ ... &cat=boxer

It's obvious why - the fighters in the region stay active and the likes of Moli are able to assert their dominance over the others, it's just that they rarely fight opponents from outside the region and when they do they are almost always easily beaten, often by fighters well below them in the rankings (take Szpilka KO4 Basile in June 2012, Szpilka totally dominated that fight before the early stoppage but beforehand Basile had 26 points more).

Is there anything that can be done about this, or do we just have to accept that South American fighters (particularly heavyweights) can attain an artificially high ranking by staying active and beating the best fighters in their region?

It's a tough solve. In my opinion, the only reasonable solution is to do periodic adjustments (rating reduction) for the boxers who never leave, based on an extremely small sample set of the ones that do and how they perform when facing others outside of their regional population.

If the average fighter that leaves the pop has a rating of 60 and loses 90% of the time they are favored... you figure out how much of a rating reduction would have been necessary pre-fight to take that number down to nearly 50. I say nearly due to "home advantage", which is very real in boxing. This definitely isn't perfect and would require some additional logic, but it is better than nothing.

In the case of Arias, he stepped up immensely when he "left", so he's not a good example.

[Bobbyptsd]

I was looking at the ratings, and saw Barrera is ranked #4 all time at 140. I don't understand this, aren't fighters ranked where they spent the bulk of their career, or am I missing something?

[computerrank]

Most titles at super bantamweight ...

[JCS]

Are there any plans to re-work the all-time ratings?

My suggestion would be to create a new database table so that fighters can accumulate points in more than one division, eliminating a large issue with the current ratings... allowing a fighter to be ranked in more than one division for all-time purposes.

Furthermore, I'd like to suggest something similar to a new baseball stat called JAWS which basically uses the total rating combined with a rating of the best 7 years.

Instead of using solely rating points, I would suggest utilizing rank position as something that is equally weighted. While I can appreciate that rating points impose depth consideration, it seems that some appreciation should be given to era-specificity.

I'd also suggest dropping at least two data points per year, if not four. If the all-time rankings are only run once per month, then so be it.. but the outcome will be far superior to what we have now. To make things run more smoothly, only insert a record if a fighter is ranked say.. #25 or better -- why do we care about anything less?

A table like:

BoxerId,Division,Rating,Rank

At this point, the only question mark is how to use rating and rank equally, but it's a good setup to start with.

This is something similar to what I use at my site (http://www.fightmatrix.com/all-time-mma-rankings/)

One more issue will present itself if we allow fighters to be ranked in more than one division.. and that is, using their rating points accumulated in one division in another division, for having fought a soft opponent. There will need to be some sort of flag that will allow count for that division move based on a quality performance of some kind. Example: If Sergio Martinez takes a fight at 168 against ranked #79 and stays there for 10 months using his lofty Middleweight points, he should not get credit for any of that time. Rather, in that case, it would be assigned back at 160.

[computerrank]

...

Where would Mayweather feature in all time P4P if he retired next year?

He would be at about 1400 points, which means about #5 at all time welterweight and about #19 at all time p4p.