Ratings - please read before commenting

[SportsRatings]

All-time ratings rewarding good results against top 10 opponents only.

40 4 128 George Foreman Heavyweight
1973-01-22 Heavyweight 2 W 4 Joe Frazier
1974-03-26 Heavyweight 2 W 6 Ken Norton
1976-01-24 Heavyweight 2 W 10 Ron Lyle
1976-06-15 Heavyweight 2 W 8 Joe Frazier

41 2 122 Michael Moorer Heavyweight
1990-04-28 Light Heavyweight 8 W 10 Mario Melo
1994-04-22 Heavyweight 10 W 3 Evander Holyfield

Seeing this part makes me wonder why George's KO of Moorer isn't listed in his top 10 opponent wins?

[computerrank]

All-time ratings rewarding good results against top 10 opponents only.

40 4 128 George Foreman Heavyweight
1973-01-22 Heavyweight 2 W 4 Joe Frazier
1974-03-26 Heavyweight 2 W 6 Ken Norton
1976-01-24 Heavyweight 2 W 10 Ron Lyle
1976-06-15 Heavyweight 2 W 8 Joe Frazier

41 2 122 Michael Moorer Heavyweight
1990-04-28 Light Heavyweight 8 W 10 Mario Melo
1994-04-22 Heavyweight 10 W 3 Evander Holyfield

Seeing this part makes me wonder why George's KO of Moorer isn't listed in his top 10 opponent wins?

Moorer wasn't in the top 10 at that time, although he had defeated Foreman.

[margaret thatcher]

surely the unbeaten moorer coming off a win over holyfield was in the top 10?

[computerrank]

surely the unbeaten moorer coming off a win over holyfield was in the top 10?

Moorer was #10 when defeating Holyfield #3. But when defeated by Foreman he was no longer in the top 10. Moorer was in decline since 1992 and the Holyfield win was just an outlier.

[margaret thatcher]

10 beating 3, and then not being top 10 in his next fight 7 months later seems very interesting! is it that whole rating them based on future results thing?

[gregregegg]

surely the unbeaten moorer coming off a win over holyfield was in the top 10?

Moorer was #10 when defeating Holyfield #3. But when defeated by Foreman he was no longer in the top 10. Moorer was in decline since 1992 and the Holyfield win was just an outlier.

So let me get this strait. Moorer was number 10, then he beat number 3. Then 6 months later when he fought forman, moorer was ranked outside of the top 10? how is that possible.

Also, how was he on decline since 1992... arguably his 4 best wins were from 1994-1997.

I appreciate the boxrec way of using an algorithm rather than an opinion, but i feel the current version of the algorithim is a bit broken...

[Freiheit]

The rating for this boxer seems very very high, can you please investigate
https://boxrec.com/en/proboxer/818883

[computerrank]

10 beating 3, and then not being top 10 in his next fight 7 months later seems very interesting! is it that whole rating them based on future results thing?

yes

[margaret thatcher]

ah, okay

[computerrank]

surely the unbeaten moorer coming off a win over holyfield was in the top 10?

Moorer was #10 when defeating Holyfield #3. But when defeated by Foreman he was no longer in the top 10. Moorer was in decline since 1992 and the Holyfield win was just an outlier.

So let me get this strait. Moorer was number 10, then he beat number 3. Then 6 months later when he fought forman, moorer was ranked outside of the top 10? how is that possible.

Also, how was he on decline since 1992... arguably his 4 best wins were from 1994-1997.

I appreciate the boxrec way of using an algorithm rather than an opinion, but i feel the current version of the algorithim is a bit broken...

The Boxrec rating sees him more succesfull with clear wins in his Light Heavyweight and beginning Heavyweight period.

His best bouts in career in 1994-1997 were against
- Holyfield - yes
- Foremen - no - loss against nearly equal rated
- Schulz - no - close win against lower rated
- Botha - no - against lower rated
- Bean - no - close win against lower rated

[computerrank]

The rating for this boxer seems very very high, can you please investigate
https://boxrec.com/en/proboxer/818883

His rating is based on his win against Edi Delibaltaoglu. His losses against debuting Leon Gavanas and debuting Granit Shala do not reduce his rating,because both went well after that.

[computerrank]

All-time ratings rewarding good results against top 10 opponents only. Only best result against same opponent as before. At most 200 points for defeating #1 or #2.

Now additional points for each year in top 10. At most 100 points for each year #1.

yeFirst glance for all men and all divisions with more than 100 points:

[gregregegg]

Moorer was #10 when defeating Holyfield #3. But when defeated by Foreman he was no longer in the top 10. Moorer was in decline since 1992 and the Holyfield win was just an outlier.

So let me get this strait. Moorer was number 10, then he beat number 3. Then 6 months later when he fought forman, moorer was ranked outside of the top 10? how is that possible.

Also, how was he on decline since 1992... arguably his 4 best wins were from 1994-1997.

I appreciate the boxrec way of using an algorithm rather than an opinion, but i feel the current version of the algorithim is a bit broken...

The Boxrec rating sees him more succesfull with clear wins in his Light Heavyweight and beginning Heavyweight period.

His best bouts in career in 1994-1997 were against
- Holyfield - yes
- Foremen - no - loss against nearly equal rated
- Schulz - no - close win against lower rated
- Botha - no - against lower rated
- Bean - no - close win against lower rated

The opponents were lower rated than him. But higher rated than his pre 1992 wins? Surely?

[computerrank]

All-time ratings with extended scope - max top 100 per division - updated - please compare with current all-time ratings and give me you response

- points for top 100 opponents defeated - max 200 points for defeating #1 or #2 ... 2 points for defeating #100
- points for annual top 100 rankings in division - max 100 points for annual #1 ... 1 points for #100
- rating division is division with most points for defeated opponents

[computerrank]

# improved rating algorithm for boxers with less 6 bouts
# improved all_time ratings based on ranks of 10 best defeated opponents and own ranks of 10 best years
# - avoids overrating of current boxers
# - based on ranks in annual top 1000
# - ranks in neighbor divisions are rated lower
# - ranks depreciated when top 1000 are not filled completely in initial times

[pugilisticspecialist]

# improved rating algorithm for boxers with less 6 bouts
# improved all_time ratings based on ranks of 10 best defeated opponents and own ranks of 10 best years
# - avoids overrating of current boxers
# - based on ranks in annual top 1000
# - ranks in neighbor divisions are rated lower
# - ranks depreciated when top 1000 are not filled completely in initial times

What is the prediction ratio?

[SportsRatings]

# - ranks in neighbor divisions are rated lower

? I this when one fighter is ranked in two divisions?

[computerrank]

# improved rating algorithm for boxers with less 6 bouts
# improved all_time ratings based on ranks of 10 best defeated opponents and own ranks of 10 best years
# - avoids overrating of current boxers
# - based on ranks in annual top 1000
# - ranks in neighbor divisions are rated lower
# - ranks depreciated when top 1000 are not filled completely in initial times

What is the prediction ratio?

The prediction ratio has not changed. It is for the all-time ratings.

[computerrank]

# - ranks in neighbor divisions are rated lower

? I this when one fighter is ranked in two divisions?

This is when a fighter got points in other divisions than the division is assigned to.

[pugilisticspecialist]

# improved rating algorithm for boxers with less 6 bouts
# improved all_time ratings based on ranks of 10 best defeated opponents and own ranks of 10 best years
# - avoids overrating of current boxers
# - based on ranks in annual top 1000
# - ranks in neighbor divisions are rated lower
# - ranks depreciated when top 1000 are not filled completely in initial times

What is the prediction ratio?

The prediction ratio has not changed. It is for the all-time ratings.

Can I see all-time career peak ratings for heavyweight, please?

[computerrank]

What is the prediction ratio?

The prediction ratio has not changed. It is for the all-time ratings.

Can I see all-time career peak ratings for heavyweight, please?

There is no algorithm for that any more. You can go through the records of the Boxers and collect the peak ratings.

[Lennox]

surely the unbeaten moorer coming off a win over holyfield was in the top 10?

Moorer was #10 when defeating Holyfield #3. But when defeated by Foreman he was no longer in the top 10. Moorer was in decline since 1992 and the Holyfield win was just an outlier.

This is nonsense
With IWBR. Moorer was#4 when he defeated Holyfield who was #1
Moorer was #1 when he was defeated Foreman who was #10 that moved Foreman to #2

Your ratings can't be that far out from ours. Below is a Heavyweight Lineage of World title fights which shows the pre-fight ranking.

https://www.premierboxingorganisation.c ... EAVYWEIGHT

[SportsRatings]

It's probably been a while since he beat anyone decent, but he just got draws vs. Zhang and Hunter and still isn't ranked. Does he need to have a win before he's ranked?

[computerrank]

It's probably been a while since he beat anyone decent, but he just got draws vs. Zhang and Hunter and still isn't ranked. Does he need to have a win before he's ranked?

Jerry Forrest presents himself as #16 WBC Bridgerweight. Bridgerweights (neither Heavyweight nor Cruiserweight) cannot be ranked in the BoxRec schema, which doesn't know Bridgerweight. None of the other major orgs shares that WBC idea.

[marcianofan]

It's probably been a while since he beat anyone decent, but he just got draws vs. Zhang and Hunter and still isn't ranked. Does he need to have a win before he's ranked?

Jerry Forrest presents himself as #16 WBC Bridgerweight. Bridgerweights (neither Heavyweight nor Cruiserweight) cannot be ranked in the BoxRec schema, which doesn't know Bridgerweight. None of the other major orgs shares that WBC idea.

I've long been confused about if and/or why that exclusion is really necessary. I agree with the status quo of ignoring Bridgerweight, but is there any particular reason he can't just be ranked wherever he fits in the rankings of the boxrec-recognized weight class in which his most recent fight would belong if not for the existence of bridgerweight (ie in most cases HW, or in a few like Bilal Laggoune, CW)? I'm not quite understanding whether excluding self-or-WBC-styled Bridgerweights is something you decided to do at an Admin or editorial level, or if it's a matter of the system not recognizing their weight class because none is assigned on the bout sheet. If it's the former, I'd ask that the decision be reconsidered. If the latter, could that not be fixed just by specifying the most logical "real" division in the bout details?

It just seems to me that Jerry Forrest is a heavyweight by any but the random WBC standard, and his being excluded from the heavyweight rankings means the heavyweight rankings are, to some extent, incorrect. I actually did an excel file a few hours ago where I copy and pasted the current heavy and cruiser rankings, and inserted the fighters listed as WBC Bridgerweights rankees on the WBC webstie, using the rankings points each guy appears to have had in the division as of his last fight. It looks like the top 10 are still solid and unchanged, but by the time you get down to the #50 heavyweight, Nathan Gorman....he really ought to be #60 after inserting #13 Oscar Rivas, #23 Evgeny Romanov, #25 Lukasz Rozanski, #30 Jerry Forrest, #31 Alen Babic, #34 Andrey Fedosov, #49 Shigabudin Aliev, #52 Ruslan Fayfer, #54 Petar Milas, and #57 Joey Tambe Djeko (subject to any time-related decay their points might have seen since last fights), To me, when roughly 15% of the top 50 or 60 heavyweights are hidden, and lower-ranked fighters are glorified in their place just because of some random WBC nonsense they might be giving lip service to, that significantly lowers the value of the rankings as they are, and sorta cries out for any remedy that can be made to function.

It's probably a lot more difficult to address from a coding/policy/practical perspective than I realize, but I'm just wondering, I guess, if you see it as a problem worth addressing. I just think that Boxrec behaving as if bridgerweight doesn't exist is good, but that making it look like the fighters themselves don't exist as a consequence of that is very, very bad.

Oh...PS- do you happen to know what Forrest's points at heavyweight would have had him ranked before the fight, if he'd been rankable as a heavyweight?