BoxRec

I don't pay any attention to them. I guess there's some kind of algorhythm behind the way it's worked out.

Ezzard wrote:I don't pay any attention to them. I guess there's some kind of algorhythm behind the way it's worked out.

I wonder what algorithm gives you Tommy Burns as the #11 ATG heavyweight. Or what algoritym puts Earnie Schaaf, Primo Carnera and Young Stribling (!) ahead of Max Schmeling.

I had mentioned on another thread shortly after the last change, how rediculous the rankings were...insulting and embarassing, the piont spread between the top men would leave the ill-informed to thinking that every one below the top few fighters wasn't up to much , when in fact it's quite the opposite!!!

BoxRec is an EXCELLENT site, there are a lot of informed people here and their record keeping and information gathered is great for resourcing, but these ratings Have Got to be Corrected - for the integrity of the site, the worth & pride of the fighters and for use of the researchers and users, giving us 'usable' information to the betterment and eventual return to a once TOP Sport.

As pundit says; "don't pay any attention to them!" However we'd like to give them respectful consideration if they only just Reflect the True rankings of the fighters!

Jim.

..the only ratings i pay any attention to are my own. saves time and are more accurate.

pundit wrote:

Ezzard wrote:I don't pay any attention to them. I guess there's some kind of algorhythm behind the way it's worked out.

I wonder what algorithm gives you Tommy Burns as the #11 ATG heavyweight. Or what algoritym puts Earnie Schaaf, Primo Carnera and Young Stribling (!) ahead of Max Schmeling.

...and James J. Jeffries (undisputed champ for 5 years with 7 title defences) is on rank 80 (!)

Friedie wrote:

pundit wrote:

Ezzard wrote:I don't pay any attention to them. I guess there's some kind of algorhythm behind the way it's worked out.

I wonder what algorithm gives you Tommy Burns as the #11 ATG heavyweight. Or what algoritym puts Earnie Schaaf, Primo Carnera and Young Stribling (!) ahead of Max Schmeling.

...and James J. Jeffries (undisputed champ for 5 years with 7 title defences) is on rank 80 (!)

Well Jefferies is going to be down to the fact he only has like 20 fights down on Boxrec with some of the guys records as 1-2-2 or whatever, undisputed champion or not.
Similarly Young Stribling's record on paper will give him a higher than deserved position.

Seriously guys, when you consider the amount of boxers in the database and lack of accurate records for pre 50s boxers it's silly to complain when a computor can't put every heavyweight in history in order anywhere near accurately. I'm amazed how offened people get by it, especially when nobody is trying to say the all time rankings mean anything anyway.

There certainly are some problems with the rating system. It's a good idea to click on the "Technical Description" that explains the system. You will see that a lot of research went into it. It's a also a very complex formula that obviously had a lot of thought put into it.

Incomplete records from way back are one problem. It says that more information is added daily. There have been some improvements; for example about year ago Jimmy bivins was the #1 heavyweight. Now he is something like #28.

Obviously there are some glitches in the formula to rating fighters. For some reason, certain fighters that are nowhere near the best are rated way to high. Conversely, others that should be higher aren't. More often than not though, the system is at least somewhat realistic, which leads me to believe the entire formula isn't all bad. For example Ali is the #1 heavyweight, Louis is #2, and many other greats are atleast close to where you would think they would be.

I tried looking at fighters that are badly underrated and tried to see what they had in common. Some just fought a long time ago (such as Corbett, Fitzsimmons, and Jeffries) and their records and their opponents records are incomplete. Others are sort of a mystery though.

I also looked at fighters that are obviously rated too high and tried to see what they had in common. I found a few things.
1. There seems to be an over emphasis on fighters who simply had a lot of wins. Many guys who had a lot of wins, but didn't have many wins over good or even decent opponents are rated way too. It's seems that in the mathematical formula, way too much weight is given to a fighter that wins a lot of fights over tomato cans.
2. Losses don't seem to be count as badly as they should. Guys with a lot of losses (even in their "prime") are rated much higher than they should be.
3.There may be too much emphasis about being knockouts. For example it has been pointed that Tommy Burns is way too high. That may be in part because according to the data base he was only knocked out once, and that probably doesn't count because it was at the end of his career in the "decline phase" of his career.

As a side note concerning Burns, his loss to Jack Johnson should be counted as a KO loss. The police interrupted the fight. Most record books list this as a TKO. In a similar situation with Fireman Jim Flynn, Johnson is credited with a TKO in the Boxrec database. This may lower Burns rating a little bit as well as raise Johnson's a little.

4. I'm not sure if the "decline phase" of a figher's career is done well. I understand that the intent is to not include losses that a fighter has when he is well past his best. However, it seems like that often too much of a fighter's career isn't counted and I suspect that too many losses are not counted.

One good thing about the rating is that it brings attention to fighters who many seem to have forgotten. Guys like Loi and Locche pop up in the Jr welterweight rankings and Billy Miske appears at #25 in the heavyweight ratings.


Anyway, it would be nice if one of the Boxrec Editors can shed some light on the formula and would consider tweaking it.
Maybe a larger review of the fighters who are obviously rated too and those rated way too low will indicate what these guys have in common and thus show where the flaws are in the formula.

It's just a shame that all of this research is done and that such a detailed formula is used and the results come out this silly.