boxrec has tony bellew as 4 at heavy ...

[man]

... and this is just weird. i am losing some
respect for the system.

[ValMar]

Statistics are like bikinis.............

[Nightmare Roy]

... and this is just weird. i am losing some
respect for the system.

I noticed that the other day, pretty stupid.

[squiggy]

Virtually no one at heavyweight has any points, which is so different from how that division's rankings usually seem to be that I suspect it's the result of some recent change to the system.

[Lackeos]

There's been like 3 threads about this. Move on.

[jujigatame]

You can't have guys starting from 0 in a new weight class, it would do way more harm to the rankings than good. It would be particularly nonsensical at the lower weights where there are only 3 or 4 pounds between classes. Think of how stupid it would have been for Inoue or Chocolatito to have been ranked like a rookie every time they moved up a few pounds.

[Enlightened-One]

Why don't people look at the BoxRec points rating calculation system and identify the flaws in the maths used by their equations, because that's what this discussion ultimately boils down to?

If the logic used for the calculation is flawed, then people should help BoxRec address these issues by listed them and explaining how they can be improved, so that their calculated ratings more accurately reflect reality.

Those whinging about Tony Bellew’s rating without even bothering to review how his ratings points have been calculated aren’t really doing anything constructive.

[joshj909]

After beating Flores Bellew had 316 points from Cruiserweight+LHW.

Pre-Haye, his points were devalued to 184, and after both Haye fights he has 246. On the current rankings, this would be roughly the same as Miller/Ortiz beating Chisora/Kownacki twice and moving up to Bellew's spot at 4th.

I support devaluing the points by say 25%? when changing weightclass, however, in this instance the issue doesn't seem to be with the change in weightclass as 184 points seems reasonable for a cruiserweight champion with 316 points to receive when changing class. Cruiserweight also had a wider spread of points compared to heavyweight as Klitschko/AJ/Fury had stacked most of them for themselves. This could've been part of the issue.

Another source of the issue is the points Haye had, or where they came from. Haye built up 418 points before fighting Klitschko which most people would agree is fair, then went down to 310 after the loss. After this he fought Chisora and got 341 points, again, no issue there. After the 4 year layoff he was relegated to 70 points (again, fair) and after two wins against two padded records he ended up with 113 points, which after the fights seems like a joke but these fighters were 30-1 and 29-0. The issue is the points of Gjergjaj if anything, who had 80 by fighting nobodies.

Therefore in my opinion the only real issue is allowing boxers to gain points for fighting loads of others who have a fraction of their points.

[JCS]

Bellew conjured up the perfect storm of how to beat a rating system like this.

- He moved up a division.. the "biggest" move in the sport, but to one that's pretty thin.
- He beat a guy who was fairly inactive and on the decline, but still had a decent amount of points.
- He beat the same guy again.

Should Haye have still had 83 points going into the 2nd fight? OK, so pretend he was dropped to 50... That still puts Bellew at #4 HW. Maybe the drop from CW to HW should be more severe.. this just means Bellew would've started lower against Haye in the 1st fight, which means his gain would've been elevated for beating Haye., again, putting him at #4.

Combating this type of shit isn't easy. To do it in a mathematically pure sense, would take a great deal of time to research and an advanced mathematics degree. Understand, any fundamental change you make, affects the entire system. If you don't want to do that, then you have to be very selective... and identify the scenario that is indeed the problem.

[man]

Bellew conjured up the perfect storm of how to beat a rating system like this.

that is what i hope. overall i like the
ratings and i like that they are objectively
derived. just as of recent some looked
weird and tony bellew really .stands out.

[man]

Why don't people look at the BoxRec points rating calculation system and identify the flaws in the maths

because i do not want to invest the
time this would take.

[man]

There's been like 3 threads about this. Move on.

didn't see them and don't care.

[jamamb]

a lot of ppl actually have bellew in there top 10 even though haye hasnt scored a meanginful hw win in years

[Thomastearns]

Does it really matter whether Bellew is 4/5/6 etc?

He beat David Haye twice against the odds and like Whyte his star is on the up. Unlike Whyte he freely admits that he's not a contender for the top spot. So he won't be going any higher.

Joshua and Wilder are the clear 1 and 2 and the rest are fighting for the remaining places.
The positions are unlikely to change much before Joshua and Wilder get it on unless they start fighting each other.

[jamamb]

haye hadnt scored a relevant win in years, he shouldnt be launching anyrone up that high

[jujigatame]

The problem is not with Bellew's point total, it's with the HW division at large. It's a total wasteland where top guys are almost never fighting each other. And there are a bunch of situations that "extracted" points from the division as well:

- Wlad retiring

- Fury disappearing for 3 years

- Pulev inactive for 16 months and fighting only cans for a full year before that

- Wilder going over 2 years between remotely decent opponents

Also, considering Huck fought Povetkin effectively to a draw, I don't have any reason to believe Bellew wouldn't be competitive at that level either.

[Redback Rasta]

Statistics are like bikinis.............

Well you got my attention. How so?

[ValMar]

Statistics are like bikinis.............

Well you got my attention. How so?

What they (statistics) reveal is suggestive, but what they conceal is vital (Aaron Levenstain).

[squiggy]

Good line. All I had come up with was "they cover your ass."

[Redback Rasta]

Statistics are like bikinis.............

Well you got my attention. How so?

What they (statistics) reveal is suggestive, but what they conceal is vital (Aaron Levenstain).

I like your work.

[ValMar]

Well you got my attention. How so?

What they (statistics) reveal is suggestive, but what they conceal is vital (Aaron Levenstain).

I like your work.

[candyslim]

The problem is the programme identifies the fact that Bellew has just scored two victories over a former two weight world champion and scores him accordingly.

It is beyond the the system's capability to recognize that David Haye was basically fukced at that point, even before the first fight.

It's not an easy fix without introducing an element of human interference and bias, the relative absence of which, is one of best features of a computer based system.

[jamamb]

ya but haye hadnt scored a decent win in years , that should be able to be accounted

in the last 5-6 years hed only beaten de mori and gjerjag and hed had long inactivity

[candyslim]

Well you got my attention. How so?

What they (statistics) reveal is suggestive, but what they conceal is vital (Aaron Levenstain).

I like your work.

Me too. I particularly like the way the bait was left instead of putting the punchline in the same post

[Lennox]

Those reasons are why www.premierboxingorganisation.com/heavyweight-200 are vastly better.

We account for lack of opposition quality in a strict way.

Bellew is 10. I think but even that is probably 10 places too high.