Boxrec ratings are f*cked up sometimes

[boxing_rocks]

Golovkin only got 2 points rating increase for beating the 2nd best middleweight, while he got 136 points for beating welterweight Brook.

Canelo got 411 (!!!) points for beating Cotto.

That is just ridiculous. The boxrec rating system works fine for normal scenarios, but when people jump weights, it is broken.

[crusader]

Yes, sometimes stuff like that happens.

Another example is Richard Commey being rated 4 spots ahead of Denis Shafikov, even though Shafikov fairly beat him in December! Apparently all Commey had to do was outpoint Hedi Slimani, taking him from 173 to 204, whereas Shafikov went from merely 156 to 181 for beating Commey.

[JCS]

Golovkin only got 2 points rating increase for beating the 2nd best middleweight, while he got 136 points for beating welterweight Brook.

Canelo got 411 (!!!) points for beating Cotto.

That is just ridiculous. The boxrec rating system works fine for normal scenarios, but when people jump weights, it is broken.

Golovkin barely beat Jacobs.... the margin of victory matters.

[SaadOffTheDeck]

The ratings have always been terrible. If you don't like them, don't take them seriously.

[boxing_rocks]

Golovkin only got 2 points rating increase for beating the 2nd best middleweight, while he got 136 points for beating welterweight Brook.

Canelo got 411 (!!!) points for beating Cotto.

That is just ridiculous. The boxrec rating system works fine for normal scenarios, but when people jump weights, it is broken.

Golovkin barely beat Jacobs.... the margin of victory matters.

Did Canelo destroy Cotto ? He got 200x points compared to Golovkin.

[boxing_rocks]

The ratings have always been terrible. If you don't like them, don't take them seriously.

I like them in general. They were better though before the last major revamp a few months ago.

[BitPlayer]

Golovkin only got 2 points rating increase for beating the 2nd best middleweight, while he got 136 points for beating welterweight Brook.

Canelo got 411 (!!!) points for beating Cotto.

That is just ridiculous. The boxrec rating system works fine for normal scenarios, but when people jump weights, it is broken.

Golovkin barely beat Jacobs.... the margin of victory matters.

Did Canelo destroy Cotto ? He got 200x points compared to Golovkin.

He did on the cards.

[crusader]

How many points would GGG be at if he outpointed Jacobs by the same scores that Canelo beat Cotto?

Cotto started with about 500 more points than Jacobs, so I imagine that explains much of the disparity. GGG and Canelo won UDs, and it would be strange if the latter got 200x more points simply due to the greater margin of victory.

[BitPlayer]

Just for the stats.

Cotto Canelo
Judge: Burt A. Clements 118-110
Judge: Dave Moretti 119-109
Judge: John McKaie 117-111

750 points before fight 804 points
1161 points after fight 530 points

Golovkin Jacobs
Judge: Max DeLuca 114-113
Judge: Don Trella 115-112
Judge: Steve Weisfeld 115-112

736 points before fight 375 points
738 points after fight 393 points

[greg]

..I can relate to "margin of victory" factor, however the 2-point increase beating the second best MW unanimously on all three cards seems to be disproportionately small IMO...

[computerrank]

..I can relate to "margin of victory" factor, however the 2-point increase beating the second best MW unanimously on all three cards seems to be disproportionately small IMO...

The result was close - only 2.3 points difference per judge. A clear win on points would have required at least 6 points difference per judge in a 12 rounds bout.

And Golovkin with 736 points was rated much higher than Jacobs with 375 points.

So actually regarding regular points Golovkin even lost 18 points and Jacobs got 18 points.

Only due to Jacobs ranked at #2 before the bout, Golovkin got additional 20 points and ended with 2 point higher than before.

[boxing_rocks]

..I can relate to "margin of victory" factor, however the 2-point increase beating the second best MW unanimously on all three cards seems to be disproportionately small IMO...

The result was close - only 2.3 points difference per judge. A clear win on points would have required at least 6 points difference per judge in a 12 rounds bout.

And Golovkin with 736 points was rated much higher than Jacobs with 375 points.

So actually regarding regular points Golovkin even lost 18 points and Jacobs got 18 points.

Only due to Jacobs ranked at #2 before the bout, Golovkin got additional 20 points and ended with 2 point higher than before.

... and I am fine with Golovkin's addition, but Canelo's one is outrageous.

[computerrank]

... and I am fine with Golovkin's addition, but Canelo's one is outrageous.

The margin was 8 points per judge in favour of Alvarez. 6 points per judge is a clear win.

Alvarez was close behind Cotto before the bout. Alvarez had 750 points and Cotto had 804 points.

So Alvarez won around a third of the points regularly - 274 points and Cotto lost them.

Additionaly Alvarez got 137 extra points for clearly defeating Cotto ranked #1 before.

[boxing_rocks]

... and I am fine with Golovkin's addition, but Canelo's one is outrageous.

The margin was 8 points per judge in favour of Alvarez. 6 points per judge is a clear win.

Alvarez was close behind Cotto before the bout. Alvarez had 750 points and Cotto had 804 points.

So Alvarez won around a third of the points regularly - 274 points and Cotto lost them.

Additionaly Alvarez got 137 extra points for clearly defeating Cotto ranked #1 before.

The algorithm is clearly f*cked up if Canelo's win over Cotto is 200x times more valuable than Golovkin's win over Jacobs and when a victory over a blown up welter Brook is 68 times more valuable than the win over Jacobs.

[Lennox]

Boxrec ratings are awful.

http://www.premierboxingorganisation.com

Produced monthly. You won't find the Shafikov/Commey type situations. These are linear based so YOU rise up in the ratings by beating quality. In Boxrec you can jump up by beating poor opponents. Their formula is dire.

[SaadOffTheDeck]

Boxrec ratings are awful.

http://www.premierboxingorganisation.com

Produced monthly. You won't find the Shafikov/Commey type situations. These are linear based so YOU rise up in the ratings by beating quality. In Boxrec you can jump up by beating poor opponents. Their formula is dire.

Your ratings suck too.

[jezzamundo]

Canelo has an inflated rating on BoxRec because judges seem to love him for some reason, so he gets more credit for his decision wins than he deserves. To my eyes, Canelo beat Cotto no more clearly than GGG beat Jacobs, yet the three blind mice in charge gave him a landslide victory, which gave his rating a huge boost. I think the scoring for GGG-Jacobs was very fair and accurate.

[Lackeos]

Golovkin only got 2 points rating increase for beating the 2nd best middleweight, while he got 136 points for beating welterweight Brook.

a) He barely beat Jacobs. He dominantly beat Brook.
b) The fact that he gained 126 points off of Brook is one of the factors in why he gained 2 points off of Jacobs. If Golovkin had gone into the Jacobs fight with a 613 rating and knocked Jacobs out, he would've gained about 127 points. But since Golovkin went into the Jacobs fight with a 736 rating (thanks to the 126 points he gained off of Brook), combined with the fact that he barely won, he only gained 2 points. You can't just keep linearly gaining 126 points every time you beat another opponent of the same rating, otherwise your rating would just keep linearly growing and become an increasingly inaccurate indication of how good that fighter really is.

[MachoTime]

BoxRec Ratings...I was always under the impression that gaining points had to do with the quality of the fight. Like a 4 star compared to a 3 star fight and so on.

[Covfefe]

Boxrec ratings are awful.

http://www.premierboxingorganisation.com

Produced monthly. You won't find the Shafikov/Commey type situations. These are linear based so YOU rise up in the ratings by beating quality. In Boxrec you can jump up by beating poor opponents. Their formula is dire.

How is Wladimir Klitschko the 'real' heavyweight champion?

[Lennox]

Boxrec ratings are awful.

http://www.premierboxingorganisation.com

Produced monthly. You won't find the Shafikov/Commey type situations. These are linear based so YOU rise up in the ratings by beating quality. In Boxrec you can jump up by beating poor opponents. Their formula is dire.

Your ratings suck too.

It must be your eyes are corrupt. The ratings are based on results of the man that beat the beat man. They are not subjective so you will find fighters who are likely to beat higher ranked fighters. If you think someone is lower than they should be it is because they have avoided quality opposition. They are the finest ratings.

[Lennox]

Boxrec ratings are awful.

http://www.premierboxingorganisation.com

Produced monthly. You won't find the Shafikov/Commey type situations. These are linear based so YOU rise up in the ratings by beating quality. In Boxrec you can jump up by beating poor opponents. Their formula is dire.

How is Wladimir Klitschko the 'real' heavyweight champion?

Tyson Fury was #1 when he beat WK and Wladimir went to #2. When Tyson Fury abdicated and said he could not box everybody moves up a rank. Wladimir is not considered inactive because he had a fight scheduled and the cancellation was not his fault.

[Covfefe]

Boxrec ratings are awful.

http://www.premierboxingorganisation.com

Produced monthly. You won't find the Shafikov/Commey type situations. These are linear based so YOU rise up in the ratings by beating quality. In Boxrec you can jump up by beating poor opponents. Their formula is dire.

How is Wladimir Klitschko the 'real' heavyweight champion?

Tyson Fury was #1 when he beat WK and Wladimir went to #2. When Tyson Fury abdicated and said he could not box everybody moves up a rank. Wladimir is not considered inactive because he had a fight scheduled and the cancellation was not his fault.

Keep it vacant then.

[candyslim]

Boxrec ratings are the worst ... except for everyone else's, to paraphrase whoever the hell it was. Seriously, any formula is bound to throw up some very curious results, but if you dispense with the formula then of course the results become subjective and unscientific. I guess we have to live with the anomalies.

[tiny_acres]

Boxrec ratings are awful.

http://www.premierboxingorganisation.com

Produced monthly. You won't find the Shafikov/Commey type situations. These are linear based so YOU rise up in the ratings by beating quality. In Boxrec you can jump up by beating poor opponents. Their formula is dire.

Your ratings suck too.

I spit coffee with this quote.
Too funny