Ratings - please read before commenting - Archived

[computerrank]

So the greatest heavyweight of all time was KO'd by Purrity, Brewster and Sanders. Ookaay.

Why not, and he learned from that - and he peaked 10 years later - at a level 7 times higher.

[pugilisticspecialist]

So the greatest heavyweight of all time was KO'd by Purrity, Brewster and Sanders. Ookaay.

Who do you consider the greatest heavyweight of all time?

[JohnL]

Ali

[pugilisticspecialist]

Ali

Wlad fought basically everyone and went almost a full decade without being in a close fight. He had a close one with Peter in 2005 then his next close one was Jennings in 2015.

Ali...
Lost to 205½ lb Frazier in 1971.
Lost to 210 lb Norton in 1973.
Taken to a split decision by a 205 lb Norton in 1973.

[JohnL]

Both men are hall of famers. As are Moore, Patterson, Liston, Foreman, Bob Foster, Holmes. How many of Wlad's opponents are or will be?

[Maleel]

I need an explanation on how the ratings work now and how to read the ratings, because at some point Boxrec seems to have made the ratings make no sense.

Lets look at Jesse Hart's points rating;
https://boxrec.com/en/proboxer/610132

Next to the first Gilberto Ramirez fight it says he had 139.5 points.
Next fight he fought a fighter with .007 points. For this fight it says he had 89.54 points. So he lost points.
Next fight he fought a fighter with 15.94 points. For this fight it says he had... 161.1 points!?!?!

So that gives me A LOT of questions.
Did Jesse Hart enter the Gilberto Ramirez fight with 139.5 points, or did he leave it with 139.5 points? Up until some time ago, when you clicked on the bout icon, you could see that information very easily, Boxrec was far more transparent with its points rankings. Under Jesse Hart it would have said something like 114.8 -> 139.5. Now Boxrec no longer has a points row on the bout page.
Secondly, Jesse Hart did not fight anyone with a points total close to his after losing narrowly to Gilberto Ramirez, so how did he suddenly nearly double his points total? All of his opponents between his fights with Ramirez had awful rankings before fighting him and continued to have awful rankings after fighting him. If he had 161.1 points directly after fighting Ramirez I would be able to rationalize that as him having a close fight with a long reigning champion and gaining points that way, but that's not the case, he doubled his points well inbetween both of his Ramirez fights.

I even went as far as looking into the records of all of his opponents around 2018 to see if maybe Boxrec's algorithm now boosts fighters rankings if a fighter they fought in the past, has a big points surge afterwards, because Boxrec before (when you could actually see a fighter's before and after points under the bout page) definitely didn't care about that in the past.
But none of his opponents, Gilberto Ramirez included, had a decent points surge around the time Jesse Hart randomly doubled his points total. The vast majority continued to always be single digit points fighters.

So I'm out of possible explanations. The only thing I can think of is that Boxrec made its actual points total histories completely unreadable and not at all indicative of how many points they actually had at a certain point in time with its modern algorithm. That, or the algorithm isn't the only thing boosting a fighters points anymore.

I used to be able to defend Boxrec's rankings because it was very easy to see and understand where the points for fighters came from if you followed the paper trail. X fighter beat X fighter with X points and the bout page would show his points before the bout and after the bout.
Now it shows you nothing of the sort, and the way fighters gain and lose points is completely inexplicable.

[computerrank]

@Maleel
When you look into the profile of a boxer, the ratings at bout time mean to be the best estimation of the boxer's fight strength at time of that bout. But the perspective considers all bouts AFTER that bout too. So the profile is a view to the career of a boxer from a present perspective.

The second bout you refer was a Cruiserweight. bout The first an third bout you refer were Super Middleweight bouts. The ratings points always reflect the fighter strength at the bout weight. The fighter's strength at Cruiserweight is lower than at Super Middleweight. So you cannot compare the raw points, they refer to different weight divisions.

[computerrank]

- ratings improved
- winner prediction ratio from 83.18 up to 84.20 percent

[margaret thatcher]

jermall approves of these new ratings

[Manrae]

- ratings improved
- winner prediction ratio from 83.18 up to 84.20 percent

Have the annual rankings been updated as well?

[pugilisticspecialist]

So how many points would they have with the reduction?

There is a weight division factor. For Heavyweight it is 1. For Cruiserweight it is (240/200)**3 = 1.2 * 1.2 * 1.2 = 1.728.
So multiply their current Heavyweight all-time ratings by that factor.

Would look like that:

Thanks. It's amazing how far ahead of the pack Haye and Usyk are.

Can I see where Usyk, Adamek and Huck would rank all-time at cruiserweight with the new ratings, please?

[computerrank]

- ratings improved
- winner prediction ratio from 83.18 up to 84.20 percent

Have the annual rankings been updated as well?

done ...

[computerrank]

Can I see where Usyk, Adamek and Huck would rank all-time at cruiserweight with the new ratings, please?

[jujigatame]

Tank above Teo now? Seems questionable.

[margaret thatcher]

of course it is, no one but tank and his family (and now boxrec) has that

[JCS]

Klitschko brothers at #1 and #2 at HW All-time.... Interesting. As is Laszlo Papp at #6 MW.

[margaret thatcher]

by interesting.......

do you mean WTF?

[margaret thatcher]

the goats

[JCS]

Seems like the latest changes have elevated those with impressive Win percentages...

[Manrae]

- ratings improved
- winner prediction ratio from 83.18 up to 84.20 percent

Have the annual rankings been updated as well?

done ...

Thanks a lot

[JCS]

Martin...

How is this guy ranked so high?

https://boxrec.com/en/proboxer/972272

The displayed rating of his last opponent looks suspect. A bug?

[jujigatame]

Martin...

How is this guy ranked so high?

https://boxrec.com/en/proboxer/972272

The displayed rating of his last opponent looks suspect. A bug?

Gotta be a bug.

[jujigatame]

Tank above Teo, not just by a thin margin but by a solid 45% of his point total, seems like something has gone awry.

[computerrank]

Martin...

How is this guy ranked so high?

https://boxrec.com/en/proboxer/972272

The displayed rating of his last opponent looks suspect. A bug?

a bug

[JCS]

Martin...

How is this guy ranked so high?

https://boxrec.com/en/proboxer/972272

The displayed rating of his last opponent looks suspect. A bug?

a bug

Carlos Canizales still a bit higher than the guy who just KO'd him. Intended?