Usyk: Number 2 Heavyweight in the world.

[candyslim]

People that outright dog the rating system don't realize all of the difficulties involved.

Changing a minimal factor can have a massive effect.... strange matchmaking, variances in fight lengths, the amount of weight divisions, the bad decisions.. on and on.. a fighter takes his rating into his/her next fight.. on and on.. and don't forget inactivity.

I don't doubt it. I did acknowledge what a difficult task it must be in my earlier post but that bit got pruned. I'm sure the combined efforts of Einstein, Newton and Hawking couldn't produce a formula we wouldn't find fault with.

[candyslim]

LMAO! Caldo is a figment of our imaginations.

I have never met Caldo but on the basis of gender alone I can assure you he bears no resemblance whatever to any of those fig-leaves of my imagination.

Damnit I meant figments

That left a bad mental picture!

Glad to be of service

[Lackeos]

Oogiebe suggests (if I haven't misunderstood) that there should be corrections made to the errant results occasionally thrown up by the system. That sounds appealing but as soon as you introduce an element of human bias, however benign, sensible, and well intentioned, you are contaminating the essential purity of the system which makes BoxRec's rankings the best by virtue of their impartiality .You do see that don't you?

There are some ways I've thought of that might be able to sensibly, semi-manually correct certain systematic aberrations.
-When you have a fluke injury like Vitali's shoulder against Chris Byrd, Vyacheslav Glazkov's knee against Charles Martin, or Eddie Chambers's arm against Tomasz Adamek; you probably shouldn't award full points for that. If fights could be tagged as "tainted by moderate fluke injury" and "tainted by severe fluke injury," then the formula could account for this and award less than full points. The only subjective factor would be in assigning the tags when applicable.
-Sometimes there are corrupt judges' decisions, and the severely inaccurate scorecards affect the accuracy of the boxrec point exchange. One way to adjust for this is to take a poll of which fighter won by how many points, and have the formula use the median score as a scorecard override. We could automatically trigger a poll for every 4-star+ fight that goes to decision. There would still be flaws in this, including bias and low participation, but I honestly trust the median boxrec score a lot more than I trust the judges.
-Amateur standouts should be seeded with non-zero boxrec ratings. This wouldn't change the game very much, but at least it would guarantee greater short-term accuracy, and it would reduce the extent to which talented amateurs are siphoning-off points from their journeyman opponents (did Zack Page really deserve to lose 35% of his points for losing to 3-0 Luis Ortiz? Was that really more of a reflection of how bad Page was or how good Ortiz was?).
-Also note that once an inaccuracy affects one fighter's rating, it is going to propagate down and affect the ratings of each of that fighter's opponents henceforth. Since Adamek's point gain against Chambers wasn't correct, Adamek's inflated rating resulted in inflating the ratings of future opponents like Glazkov, Szpilka, and Molina; and THEIR future opponents like Charles Martin and Adam Kownacki. If we had tagged that fight as tainted by moderate fluke injury, it could have fixed a lot of downstream inaccuracies.

My suggestion for initial boxrec rating seeding would be take the single highest number from these accomplishments (no adding).
-Olympic gold: 16 points
-World amateur gold: 15 points
-Olympic silver: 14 points
-World amateur silver: 13 points
-Olympic semi-finals: 12 points
-World amateur semi-finals: 11 points
-Olympic quarter-finals: 8 points
-World amateur quarter-finals: 7 points
-International tournament gold medal: 6 points
-Olympic or world amateur round of 16: 4 points
-International tournament any medal: 4 points
-National champion, large talent pool: 2 points
-National champion, any country: 1 point

[JCS]

Oogiebe suggests (if I haven't misunderstood) that there should be corrections made to the errant results occasionally thrown up by the system. That sounds appealing but as soon as you introduce an element of human bias, however benign, sensible, and well intentioned, you are contaminating the essential purity of the system which makes BoxRec's rankings the best by virtue of their impartiality .You do see that don't you?

There are some ways I've thought of that might be able to sensibly, semi-manually correct certain systematic aberrations.
-When you have a fluke injury like Vitali's shoulder against Chris Byrd, Vyacheslav Glazkov's knee against Charles Martin, or Eddie Chambers's arm against Tomasz Adamek; you probably shouldn't award full points for that. If fights could be tagged as "tainted by moderate fluke injury" and "tainted by severe fluke injury," then the formula could account for this and award less than full points. The only subjective factor would be in assigning the tags when applicable.
-Sometimes there are corrupt judges' decisions, and the severely inaccurate scorecards affect the accuracy of the boxrec point exchange. One way to adjust for this is to take a poll of which fighter won by how many points, and have the formula use the median score as a scorecard override. We could automatically trigger a poll for every 4-star+ fight that goes to decision. There would still be flaws in this, including bias and low participation, but I honestly trust the median boxrec score a lot more than I trust the judges.
-Amateur standouts should be seeded with non-zero boxrec ratings. This wouldn't change the game very much, but at least it would guarantee greater short-term accuracy, and it would reduce the extent to which talented amateurs are siphoning-off points from their journeyman opponents (did Zack Page really deserve to lose 35% of his points for losing to 3-0 Luis Ortiz? Was that really more of a reflection of how bad Page was or how good Ortiz was?).
-Also note that once an inaccuracy affects one fighter's rating, it is going to propagate down and affect the ratings of each of that fighter's opponents henceforth. Since Adamek's point gain against Chambers wasn't correct, Adamek's inflated rating resulted in inflating the ratings of future opponents like Glazkov, Szpilka, and Molina; and THEIR future opponents like Charles Martin and Adam Kownacki. If we had tagged that fight as tainted by moderate fluke injury, it could have fixed a lot of downstream inaccuracies.

My suggestion for initial boxrec rating seeding would be take the single highest number from these accomplishments (no adding).
-Olympic gold: 16 points
-World amateur gold: 15 points
-Olympic silver: 14 points
-World amateur silver: 13 points
-Olympic semi-finals: 12 points
-World amateur semi-finals: 11 points
-Olympic quarter-finals: 8 points
-World amateur quarter-finals: 7 points
-International tournament gold medal: 6 points
-Olympic or world amateur round of 16: 4 points
-International tournament any medal: 4 points
-National champion, large talent pool: 2 points
-National champion, any country: 1 point

The injury / scorecard thing is a sketchy game... it's something Lennox promotes w/ regards to his system, but unsure how he figures which ones to adjust and which ones not to... also, there's no way complete coverage is possible, so there will be holes.

As for your "starter rating" idea, I actually convinced Martin to implement such a system... and you could see it in Lomachenko's debut fight... but I'm not sure it is still in effect and it is a pro result accelerator more than it is a prior experience placement.

In Ortiz's case, his first four fights were decisive wins over boxers that were higher rated then he was, going into the fight. In my MMA system, I have a progressive accelerator for fighters who have decisive victories over opponents rated higher than them if they have no prior loss/draw/close win (debut exempted)... might be something worth looking into here.

[candyslim]

I rather like your suggestions Lackeos since this is altogether different from massaging the inconvenient rating placements.

I've nothing against suggesting improvements to the formula but I defer to those clever fellows who have devised and refined the system from its inception, to decide whether your proposals are worth implementing, and whether further refinement of those proposals is desirable.

[Ricky]

The HW division is just a freak show full of bums. Usyk is a mile better than everyone, just smaller. Looking forward to seeing him compete at HW.

[greg]

...bums or no bums, I'll still reserve my judgement till I see him in action against a top 10 guy..

[Enlightened-One]

According to BoxRec, if Usyk were to move to HW tomorrow, his points would be converted x.58. That would make him the number two ranked heavyweight behind AJ. Do you agree? Why? I certainly do not.

BoxRec currently rates Luis Nery as being sixth on their pound-for-pound ratings.

Up until recently, BoxRec also rated Jeff Horn amongst the pound-for-pound elite (they ranked him in 9th place).

I really don’t think I need to say anything more than that to make my point any clearer.

But as soon as you take points from someone because you don't think he deserves them, you're stripping away the uniqueness of the points system and making it another "here's what I think" list.

The various rules and calculations used by the BoxRec ratings system have always been flawed, but have fortunately continued to be refined, which means they’re changing on an organic basis.

For sure, it’s inevitable that there will be flaws with any mathematics-based points rating system, which is likely to result in several anomalies, but I’m still pretty confident that it’ll improve over time.

That being said, I don’t believe that the BoxRec ratings system is currently in a good enough state to carry any prestige amongst fight fans, but this perception may change in a few years’ time when far more progress has been made.

To be honest, I’m not really interested in the “uniqueness” of the BoxRec ratings system, because it’s currently not informative enough to fulfil any real purpose. And I don’t believe that my opinion is in the minority either.

There are a hell of a lot of good things about the BoxRec website, but their ratings system is not one of them.

[ewenhay]

Boxing is decided in the ring not on a computer.

If Mayweather had moved up to heavyweight the computer would likely have ranked him number one. The system doesn't work well for fighters changing weights.

It's simple. If Usyk beats a top ten heavyweight first time out then that is where he should be ranked. He shouldn't automatically get a ranking at a new weight until he fights someone.

[jamamb]

i think u could rate him at hw. he has beaten multiple guys who have beaten career hws and wouldve prob had a hw rating from 20-30 (hunter now for ustinov, bakole, kiladze, etc and briedes then for ktfoing charr). huck gave pov hell though years before. and usyk did technically fight at hw once lol! the top cruisers are all at least good enough for the top 35 or so at hw if they moved up imo. beating them isnt totally meaningless to a hw ranking

but a really high rating he has to beat someone legit first imo, at least a top 10 or so

[greg]

..I was just wondering if that could be possible to keep USYK (or anyone else for that matter) UNRANKED till he actually makes his first move and only THEN depending on the result rank him at HW using either the old algorithm or probably a new readjusted one...

[Lennox]

number 6 he comes in with the PBO.

every heavyweight that comes up from cruiserweight will carry the weight jump differently, that aside any fighter going up a weight is different some really lose a lot of power others don't.

you have to have a formula though.

[JCS]

number 6 he comes in with the PBO.

every heavyweight that comes up from cruiserweight will carry the weight jump differently, that aside any fighter going up a weight is different some really lose a lot of power others don't.

you have to have a formula though.

Do you track division-specific points for fighters?

[candyslim]

I don't see the harm in being a bit mean in converting points when changing divisions. It seems most criticism applies to little guys being overrated in the higher division for example Lomachenko's points translating to an impressive heavyweight ranking as mentioned earlier.

I don't recall any howls of protest because a fighter moving up isn't rated highly enough in the higher division, then maybe I've not been paying attention. Usyk hasn't fought at heavyweight as a pro. If he was unrated at heavyweight until he fights there, or if his cruiser points translated to say number twenty, that would not reflect his ability, but it would reflect his (lack of) achievement at the higher weight.

Once he started on building a resume in the new division his ranking would soon find its correct level wouldn't it?

[Lennox]

number 6 he comes in with the PBO.

every heavyweight that comes up from cruiserweight will carry the weight jump differently, that aside any fighter going up a weight is different some really lose a lot of power others don't.

you have to have a formula though.

Do you track division-specific points for fighters?

Jason - not really sure what you mean? our formula is the same for every boxer though we do have different rules from each division, ie going from 105 up to 108 there is no penalty same up to 115.

[JCS]

number 6 he comes in with the PBO.

every heavyweight that comes up from cruiserweight will carry the weight jump differently, that aside any fighter going up a weight is different some really lose a lot of power others don't.

you have to have a formula though.

Do you track division-specific points for fighters?

Jason - not really sure what you mean? our formula is the same for every boxer though we do have different rules from each division, ie going from 105 up to 108 there is no penalty same up to 115.

I'm trying to understand if it is strictly a conversion, or if accomplishments for fighters are separately kept per division...

So in other words, if Usyk had success at HW a few years ago, a move to HW would not penalize him as much as the actual Usyk who's all CW.

[Lennox]

no it is just a formula when he moves up + the result of the new achievement at heavyweight.

If he has tags with other fighters that is different but more appropriate if he dropped back down.

[oogiebe]

My understanding is that Usyk will get ranked at heavyweight only after he fights at HW. That's when the CW points are converted. Yes?

[SportsRatings]

i think u could rate him at hw. he has beaten multiple guys who have beaten career hws and wouldve prob had a hw rating from 20-30 (hunter now for ustinov, bakole, kiladze, etc and briedes then for ktfoing charr). huck gave pov hell though years before. and usyk did technically fight at hw once lol! the top cruisers are all at least good enough for the top 35 or so at hw if they moved up imo. beating them isnt totally meaningless to a hw ranking

but a really high rating he has to beat someone legit first imo, at least a top 10 or so

He also beat Junior Fa and Joe Joyce (ranked #56 and #38 currently at Heavyweight) in the World Series of Boxing.

[oogiebe]

i think u could rate him at hw. he has beaten multiple guys who have beaten career hws and wouldve prob had a hw rating from 20-30 (hunter now for ustinov, bakole, kiladze, etc and briedes then for ktfoing charr). huck gave pov hell though years before. and usyk did technically fight at hw once lol! the top cruisers are all at least good enough for the top 35 or so at hw if they moved up imo. beating them isnt totally meaningless to a hw ranking

but a really high rating he has to beat someone legit first imo, at least a top 10 or so

He also beat Junior Fa and Joe Joyce in the Olympics, ranked #56 and #38 currently at Heavyweight.

Yes, but that was six years ago. He probably can beat Fa right now.

[oogiebe]

From an interview with Paulie Malignaggi:

"They talk about him going to the heavyweight division. I’ve seen him in person, he’s not a small guy. He’s about 6’2″ or 6’3″, so maybe he can fill out for the heavyweight division a little bit. He’ll never be a big big heavyweight. But I think he can turn into a decent size heavyweight and compete," Malignaggi said.

"You’ll probably want to match him against some heavyweights who are a little on the smaller side but are still pretty good. Maybe like a Fres Oquendo or Carlos Takam or someone like that. Based on how he looks in those fights, and if he’s able to win, and if he’s able to win looking good or if he struggles, you judge if he’s going to be a heavyweight or if he goes back to cruiserweight."

Oquendo or Takam? Seems most posters see Usyk in top 20 right now. I fancy the idea of seeing how he does against mid-size guys before putting him up against the real big boys of the division. But maybe not those guys. I don't see him struggling against either. Oquendo? OYE!

[ewenhay]

I think he should just go straight in with Joshua and go for it.

He'd have the surprise factor on his side.

If he wins he's taken the heavyweight division by storm.

If he loses he's made a lot of dough and can ponder his next move, whether that be heavy, cruiser or putting his feet up

[oogiebe]

I think he should just go straight in with Joshua and go for it.

He'd have the surprise factor on his side.

If he wins he's taken the heavyweight division by storm.

If he loses he's made a lot of dough and can ponder his next move, whether that be heavy, cruiser or putting his feet up

He has no need to risk that. He can take on a 10-20 HW ranked fighter and if he doesn't obliterate them, he has the option of moving back to CW with his "0" still intact.

[ewenhay]

I think he should just go straight in with Joshua and go for it.

He'd have the surprise factor on his side.

If he wins he's taken the heavyweight division by storm.

If he loses he's made a lot of dough and can ponder his next move, whether that be heavy, cruiser or putting his feet up

He has no need to risk that. He can take on a 10-20 HW ranked fighter and if he doesn't obliterate them, he has the option of moving back to CW with his "0" still intact.

I hear you.

But he could risk losing out on a crack at the heavyweight title and the money that comes with it?

He might just fancy doing what Roy Jones did, except against a tougher belt holder. He doesn't seem short of confidence

[oogiebe]

I think he should just go straight in with Joshua and go for it.

He'd have the surprise factor on his side.

If he wins he's taken the heavyweight division by storm.

If he loses he's made a lot of dough and can ponder his next move, whether that be heavy, cruiser or putting his feet up

He has no need to risk that. He can take on a 10-20 HW ranked fighter and if he doesn't obliterate them, he has the option of moving back to CW with his "0" still intact.

I hear you.

But he could risk losing out on a crack at the heavyweight title and the money that comes with it?

He might just fancy doing what Roy Jones did, except against a tougher belt holder. He doesn't seem short of confidence

He is long on integrity it seems. If that's so, he won't take the RJJ route. Who knows...his team will figure it out. He's been well-managed so far and he's taken on all comers, so like I said. Who knows.