A points ranking system to legitimise boxing rankings!!!!!!!

[yaltamaltadavid]

In the last month the initial rankings were settled, and there's been further debate and clarification about the system's methods. The first updates are now posted for heavyweight and welterweight, and the rest will be posted by the 1st. I do need to note that the rankings should not be taken as authoritative in the early stages, but could take a full three years to flesh out, though sooner if ranked fighters fight each other more often.

The TBRB recently added their rules in Spanish, so I'm thinking of soon adding our rules in Spanish, French, German, Russian, Tagalog, Japanese, and Thai (haha).

Take a look, and if you think objective rankings are worthwhile leave any constructive criticism you have. If you support the project as it stands, we're still looking for more help maintaining the rankings. I'm hoping this can be shared between 17 people, with one person managing each division. Again, if you're dead set against the idea, please keep your comments here on the Scene forum.

http://worldboxingrankings.proboards.com/

[crold1]

In the last month the initial rankings were settled, and there's been further debate and clarification about the system's methods. The first updates are now posted for heavyweight and welterweight, and the rest will be posted by the 1st. I do need to note that the rankings should not be taken as authoritative in the early stages, but could take a full three years to flesh out, though sooner if ranked fighters fight each other more often.

The TBRB recently added their rules in Spanish, so I'm thinking of soon adding our rules in Spanish, French, German, Russian, Tagalog, Japanese, and Thai (haha).

Take a look, and if you think objective rankings are worthwhile leave any constructive criticism you have. If you support the project as it stands, we're still looking for more help maintaining the rankings. I'm hoping this can be shared between 17 people, with one person managing each division. Again, if you're dead set against the idea, please keep your comments here on the Scene forum.

http://worldboxingrankings.proboards.com/

While there is some interesting stuff here, I see right away at Heavyweight something that makes me say, "yup, formulas just as problematic as people."

Chris Arreola -3

Bermane Stiverne - 9

Stiverne just kicked Arreola's ass a few months ago. 9-3 and 10-2 type scores. Dropped him. Arreola doesn't have a win of any note since. Mitchell sucks. People can SEE that. They also SAW Stiverne. Formulas don't see. In a low output (meaning guys don't fight that often) sport, recent events have currency. This isn't tennis where people play 10-12 tourneys a year. What Arreola did 3 years ago doesn't matter near as much as STIVERNE KICKED HIS ASS and no rating that doesn't reflect that is 'right.'

I've flirted with formulas for AT ratings, for divisionals too (as I noted when you emailed me before). Ultimately, I concluded they come with their own problems. This is a perfect example.

I wish you good luck. A math offset to people can have its advantages but it's not an answer either. Like the college football BCS, things like this ultimately piss people off because a formula is betraying their eyes. Since people ultimately choose the tools they utilize, it dooms the system.

[yaltamaltadavid]

While there is some interesting stuff here, I see right away at Heavyweight something that makes me say, "yup, formulas just as problematic as people."

Chris Arreola -3

Bermane Stiverne - 9

Stiverne just kicked Arreola's ass a few months ago. 9-3 and 10-2 type scores. Dropped him. Arreola doesn't have a win of any note since. Mitchell sucks. People can SEE that. They also SAW Stiverne. Formulas don't see. In a low output (meaning guys don't fight that often) sport, recent events have currency. This isn't tennis where people play 10-12 tourneys a year. What Arreola did 3 years ago doesn't matter near as much as STIVERNE KICKED HIS ASS and no rating that doesn't reflect that is 'right.'

I've flirted with formulas for AT ratings, for divisionals too (as I noted when you emailed me before). Ultimately, I concluded they come with their own problems. This is a perfect example.

I wish you good luck. A math offset to people can have its advantages but it's not an answer either. Like the college football BCS, things like this ultimately piss people off because a formula is betraying their eyes. Since people ultimately choose the tools they utilize, it dooms the system.

Hi Cliff,

As I mentioned, starting out is tricky and it will take a while for the system to flesh out, so Arreola's position shouldn't yet be considered reflective of the system. But with the system this 'could' be one of those occasions when a fighter soon after moves ahead of the fighter he recently lost, because Arreola's body of success is likely greater. Head to head, Stiverne definitely proved he's better. But I think consistency is what should be most rewarded. And Stiverne hasn't beaten another top-20 heavy, right? But Arreola at least fought a couple fringe 20-ish guys, and lost to top opposition. Stiverne I think should have to prove himself against another decent opponent before we know his win over Arreola wasn't merely because of stylistic advantage. Again I can't say for sure where the system would place them had it been firmly in place, I'm guessing from looking at their records that Arreola could be slightly ahead, but Stiverne could very well be ahead because the better opposition Arreola fought was more than three years previous.

If consistent success is accepted as the prime criteria, I don't think it invalidates the system if there are occasions when a fighter even moves below the fighter he just beat. With the three year cycle, there could be occasions similar to what's very occasionally happened in tennis... after Nadal beat Ferrer in the French Open final, he actually moved BELOW Ferrer in the rankings. No one's saying that invalidates the tennis rankings, and I don't think similar rankings would invalidate strictly quantifiable rankings for boxing.

If that happens even 3% of the time, I agree this system won't work. Yes, the low output for boxing is the main difficulty with objective rankings. I also found there are difficulties because only 30 are ranked and rankings could be less accurate the further down you go, but I think objective rankings would be accurate if basically every active boxer was ranked. I admit the fractured structure of boxing makes this impossible, so objective rankings I don't think can currently be as accurate as they could be. But I actually have a bit of a wild solution as to how a single international body for boxing like FIFA or the ATP could work to correct this, and how in that case objective rankings would work from top to bottom, which I'll detail on forums soon (it wouldn't include regularly scheduled fights however, as that would never work for boxing).

Though much more wouldn't be fair, I think anything less than a three year ranking period is not sufficient. I don't know if you gave your experiment that long. What I definitely did find is earlier fights can't be backgraded, because of course each fight would need to be graded based on both fighters' ranking at the time. So I think it would be impossible to numerically grade ATG rankings. Unless you somehow started in 1900 and proceeded to do progressive monthly rankings updates since then!

Also, it's exactly that a formula obfuscates the rankings that's a major reason I'm against the boxrec rankings (and of course because their specific criteria is flawed). I'm no stats-obsessed engineering-type, and the system I'm proposing is clean and simple and more like the standings in a sports league or the ATP rankings than those American football rankings, which I took a quick look at when others mentioned them. You can look at a sports league's standings and know how each team needs to do to reach a certain position or to make the playoffs, and our ranking system is about as simple.

While I don't know much more about either baseball or American football, I do know that the MLB plays the most games of any sports league and the NFL plays the fewest, about ten times less. Yet the NFL seems to go by the same simple wins and losses criteria to determine standings. At least FIFA makes an effort to adjust records when comparing national teams who don't play each other. Shouldn't that mean the NFL standings aren't sufficiently reflective of success? And if the NFL is able to get by with a system so similar to the MLB's, shouldn't boxing with its similarly proportional sample size compared to tennis should be able to use a similar system to the ATP's without compromising itself significantly?

If this project proves in three or so years to not work, then it doesn't work. But till fair objective rankings are proven impossible, I'm going to give it a go. If it works close to how I think it will, I think these or similar rankings will be what can truly clear up the confusion with rankings and championships.

[yaltamaltadavid]

My main problem with existing rankings and the reason I thought of this idea is that the most recent win is overrated. I have difficulty accepting that one fight is a significant reflection of a boxer's impact on a division.

[crold1]

My main problem with existing rankings and the reason I thought of this idea is that the most recent win is overrated. I have difficulty accepting that one fight is a significant reflection of a boxer's impact on a division.

Arreola has never really beaten anyone of quality. Getting his ass handed to him by Stiverne is a reality. Stiverne won. He moves ahead. If the formula doesn't reflect that, the formula doesn't work. Trust me. I have tried formulas to VERY mixed results. See the jr. Division Top 20s I did a few years back. Some were good. Some were NOT. Anything that comes up with Arreola 3 isn't right.

Not saying don't keep at it. Maybe there is a perfect recipe. I think you'll be frustrated. Boxing is what it is. It's a what did you do for me lately enterprise. That dictates what people want to see. Again, ATP doesn't work here. No one fights enough. You need 3 year old results to get to a 'body of work.'