Ratings - please read before commenting - Archived

[Lennox]

www.independentworldboxingrankings.com

Streets ahead from accurately placing fighters Dave. A lot more accurate for betting.

The ratings highlight Juli Giner being better than a 6-1 fighter tonight, the weekend Rene Alvarado beating Shafikov and Ogawa beating Farmer both with the 8-1 odds also.

[StarLord25]

The same one on the ratings description page and fighter A being Truax and B being Degale. I use the values they had prior to their fight.

The winner is always rated higher than the loser. This rule wasn't documented in the description. I added it as rule 26.

The how exactly does one work out the extra points to make that happen.

[StarLord25]

www.independentworldboxingrankings.com

Streets ahead from accurately placing fighters Dave. A lot more accurate for betting.

The ratings highlight Juli Giner being better than a 6-1 fighter tonight, the weekend Rene Alvarado beating Shafikov and Ogawa beating Farmer both with the 8-1 odds also.

I've heard those rankings are better but what makes them better?

[DaveyMac]

www.independentworldboxingrankings.com

Streets ahead from accurately placing fighters Dave. A lot more accurate for betting.

The ratings highlight Juli Giner being better than a 6-1 fighter tonight, the weekend Rene Alvarado beating Shafikov and Ogawa beating Farmer both with the 8-1 odds also.

I agree that this is a very solid objective attempt at rankings and I've been supportive of it, but it is not more accurate at predicting results. The Boxrec points (if you know how to use them and I'm not sure the actual Boxrec predictor does) are uncanny at predicting winners and are an extremely useful tool not just for gambling (which I don't do) but matchmaking as well.

[Lennox]

Davey - You have other measuring bars to factor, like the ranking of the best opponent beaten within a certain time frame, so you can see if a fighter is ascending or in decline. If a fighter has recent losses and recent wins his position is quite firm, where as a fighter who is ranked much higher than any opponent he has ever beaten is more unknown. This is often seen in unbeaten prospects, it does not mean the unbeaten prospect won't win but often you can get odds worth betting. The Ward-Giner fight was much more even than 13-2, you could have cashed out at 3-2 in the 5th round. I thought about cashing out but was greedy and got what I deserved in the next round! On Saturday I 10 folded my money. Betting on seasoned fighters close in the rankings to box a draw pays as well.

[Lennox]

www.independentworldboxingrankings.com

Streets ahead from accurately placing fighters Dave. A lot more accurate for betting.

The ratings highlight Juli Giner being better than a 6-1 fighter tonight, the weekend Rene Alvarado beating Shafikov and Ogawa beating Farmer both with the 8-1 odds also.

I've heard those rankings are better but what makes them better?

The points system involves taking into account other losses in a time period, so boxers can be moved by the performance of another boxer. You could be a 10-0 fighter and not have fought for 6 months and move up if a fighter you beat beats a ranked fighter. The points he gains are shared with you so you always have 1 more point. It works in reverse too.
Simply it is about beating ranked fighters, fighting people non top 100 is really not worth much once you are a decent fighter. As you can see from the rankings most 25-0 fighters have only fought a few ranked top 100 some have gone over 40-0 and not fought one.

[StarLord25]

www.independentworldboxingrankings.com

Streets ahead from accurately placing fighters Dave. A lot more accurate for betting.

The ratings highlight Juli Giner being better than a 6-1 fighter tonight, the weekend Rene Alvarado beating Shafikov and Ogawa beating Farmer both with the 8-1 odds also.

I've heard those rankings are better but what makes them better?

The points system involves taking into account other losses in a time period, so boxers can be moved by the performance of another boxer. You could be a 10-0 fighter and not have fought for 6 months and move up if a fighter you beat beats a ranked fighter. The points he gains are shared with you so you always have 1 more point. It works in reverse too.
Simply it is about beating ranked fighters, fighting people non top 100 is really not worth much once you are a decent fighter. As you can see from the rankings most 25-0 fighters have only fought a few ranked top 100 some have gone over 40-0 and not fought one.

But doesn't Boxrec also factor in a guy who fighter A beat beating another top fighter to A's points?
And for Boxrec fights with that wide a difference I notice no points are gained.

[Lennox]

Boxrec have their way, but fighters seem to get much higher with Boxrec without fighting meaningful fighters, more emphasis is put on wins, fighters that lose seem to drop more on Boxrec. Ali Raymi made the top 10 with lots of meaningless wins.

The IWBR way is more a fighter loses and falls to a position of a gatekeeper, or his last win (factored by time). You can't be ranked ahead of a fighter if you recently lost to him, so groups of fighters can move together. With Boxrec, if you lose a fight you are ranked beow that fighter but if you fight the next day against a not great opponent you can go ahead. That would not happen in a subjective system and for many destroys computerised ratings.

The perfect system might be a top 100 computerised with the top 10 subjectively picked from say the computerised top 15. Both Boxrec and IWBR have things like Arthur Abraham ranked too high, Jeff Horn at #1 is not the best welterweight. The argument to that I suppose is Jeff Horn beat the #1 and the sport would not be exciting if everyone knew the exact places of the fighters.

[computerrank]

... Ali Raymi made the top 10 with lots of meaningless wins. ...

This was corrected with an adapted seeding system in the current system. Ali Raymi only reached #45 with it.

[computerrank]

The winner is always rated higher than the loser. This rule wasn't documented in the description. I added it as rule 26.

The how exactly does one work out the extra points to make that happen.

This doesn't depend on extra points.
The winner is simply set a bit above the mean of both pre-bouts ratings.
The loser a bit below it.

[StarLord25]

Ok.
I for one prefer computerised rankings with improvements overtime. They're more objective and less prone to corruption. The suggestions here help to improve Boxrec. Horn may not be the best in our eyes but he did beat the guy with the best resume at 147.

computerrank, the main complaint I've noticed about Boxrec is how points are kept when fighters move to different divisions and are automatically ranked in the top 10.

[StarLord25]

Bump

[computerrank]

Ok.
I for one prefer computerised rankings with improvements overtime. They're more objective and less prone to corruption. The suggestions here help to improve Boxrec. Horn may not be the best in our eyes but he did beat the guy with the best resume at 147.

computerrank, the main complaint I've noticed about Boxrec is how points are kept when fighters move to different divisions and are automatically ranked in the top 10.

Boxers changing weight division are not automatically ranked in the top 10 of the new division.
They are set into the new division depending on their weight advantage or weight disadvantage in relation to the old division.
Win-loss statistics teach us, that converting the ratings in relation of old division upper weight limit and new division upper weight limit raised to 3rd power is best predictive for wins.
So a Cruiserweight boxer moving up to Heavyweight will keep 58 percent of his rating.
A Flyweight boxer moving up to Super Flyweight will keep 92 percent of his rating.

[StarLord25]

Ok. Thank you for clearing things up. I respect your ranking systems because they come closest to ranking fighters based off what they're doing and I trust them more than sanctioning body nonsense.

[StarLord25]

By the way, have you guys ever considered archiving your top 10-15 ratings per month?

[StarLord25]

Another question, so what if a fighter is moving down a division how do the points for that work?

[computerrank]

By the way, have you guys ever considered archiving your top 10-15 ratings per month?

As new and old bouts are added every day, the ratings even for the past are changing every day.

So there it makes no sense to archive the ratings.

[computerrank]

Another question, so what if a fighter is moving down a division how do the points for that work?

At a lower weight division a boxer has a higher rating.

E.g. the rating of a boxer at Flyweight is 108 percent of his rating at Super Flyweight.

[StarLord25]

What is the logic behind that type of higher rating? And you guys do archive your top 10 rankings annually though. If it were monthly it could be done every 4 weeks or at the end of each month.

[DaveyMac]

www.independentworldboxingrankings.com

Streets ahead from accurately placing fighters Dave. A lot more accurate for betting.

The ratings highlight Juli Giner being better than a 6-1 fighter tonight, the weekend Rene Alvarado beating Shafikov and Ogawa beating Farmer both with the 8-1 odds also.

Ogawa-Farmer was much closer to the even money bet you would have expected it to be using Boxrec points.

[computerrank]

What is the logic behind that type of higher rating? And you guys do archive your top 10 rankings annually though. If it were monthly it could be done every 4 weeks or at the end of each month.

- When a boxer is moving to a higher division he is inferior regarding his common weight. When a boxer moves down he is superior regarding his common weight.
- we do not archive our top 10 annual ratings - we calculate them from scratch every year - these are some 160 annual rankings for 17 weight divisions = which sums to some 2600 rankings
- it would make too much effort to do this monthly

[StarLord25]

Ok. I've made a database arguing your top 15 fighters per division at the end of each month. I used excel and database.

[mike1989]

hi Martin!
what happened with the ratings why many boxers somewhere disappeared????

[StarLord25]

I've noticed Crawford is ranked #1 at 147 before even having a fight there.

[computerrank]

hi Martin!
what happened with the ratings why many boxers somewhere disappeared????

who disappeared - BoxRec ID please?