Ratings - please read before commenting - Archived

[computerrank]

[
I don't understand the point possible an example please ???

It is just the answer to your division adjustment proposal.

Regarding 2016 I would give following annual weights:

Heavyweight-19 boxers with 150+ points (=10 or more) => annual weight is 100 points
Cruiserweight-12 boxers with 150+ points (=10 or more) => annual weight is 100 points
Light Heavyweight-19 boxers with 150+ points (=10 or more) => annual weight is 100 points
Super Middleweight-17 boxers with 150+ points (=10 or more) => annual weight is 100 points
Middleweight-22 boxers with 150+ points (=10 or more) => annual weight is 100 points
Light Middleweight-18 boxers with 150+ points (=10 or more) => annual weight is 100 points
Welterweight-36 boxers with 150+ points (=10 or more) => annual weight is 100 points
Light Welterweight-21 boxers with 150+ points (=10 or more) => annual weight is 100 points
Lightweight-34 boxers with 150+ points (=10 or more) => annual weight is 100 points
Super Featherweight-18 boxers with 150+ points (=10 or more) => annual weight is 100 points
Featherweight-29 boxers with 150+ points (=10 or more) => annual weight is 100 points
Super Bantamweight-15 boxers with 150+ points (=10 or more) => annual weight is 100 points
Bantamweight-16 boxers with 150+ points (=10 or more) => annual weight is 100 points
Super Flyweight-13 boxer swith 150+ points (=10 or more) => annual weight is 100 points
Flyweight-11 boxers with 150+ points (=10 or more) => annual weight is 100 points
Light Flyweight-5 boxers with 150+ points => annual weight is 50 points
Minimumweight- 4 boxers with 150+ points => annual weight is 40 points

[computerrank]

These are the ratings for the last rule set:

- 100 rating points limit instead of 150 in order to better integrate the lower weights with lower level

- annual year's end rank in division with best performance
-- maximum rank is 20
-- 100 points / rank * weight
-- weight = number of division top 10 with rating higher than 100 points / 10
-- annual year's end rank points only if there was a ranked best opponent in the same year


- annual best opponent rank converted to bout division a time of bout
-- maximum opponent rank is 20
-- 100 points / (rank_opp-1) * weight_opp ; when rank_opp is 1 or 2 then use 2
-- weight_opp = number of bout division top 10 with rating higher than 150 points / 10 at time of bout


multiple opponent ranks per year not regarded - I tested this feature, but this inflated the ratings for the 1930ies to 1970ies, where boxers had more fight per year.


http://backup.boxrec.com/~martin/ann_tops_5.txt
http://backup.boxrec.com/~martin/ann_tops_years_5.txt
http://backup.boxrec.com/~martin/bouts_ranks_5.txt
http://backup.boxrec.com/~martin/all_ti ... _men_5.pdf
http://backup.boxrec.com/~martin/all_ti ... vies_5.pdf

[mike1989]

These are the ratings for the last rule set:

- 100 rating points limit instead of 150 in order to better integrate the lower weights with lower level

- annual year's end rank in division with best performance
-- maximum rank is 20
-- 100 points / rank * weight
-- weight = number of division top 10 with rating higher than 100 points / 10
-- annual year's end rank points only if there was a ranked best opponent in the same year


- annual best opponent rank converted to bout division a time of bout
-- maximum opponent rank is 20
-- 100 points / (rank_opp-1) * weight_opp ; when rank_opp is 1 or 2 then use 2
-- weight_opp = number of bout division top 10 with rating higher than 150 points / 10 at time of bout


multiple opponent ranks per year not regarded - I tested this feature, but this inflated the ratings for the 1930ies to 1970ies, where boxers had more fight per year.


http://backup.boxrec.com/~martin/ann_tops_5.txt
http://backup.boxrec.com/~martin/ann_tops_years_5.txt
http://backup.boxrec.com/~martin/bouts_ranks_5.txt
http://backup.boxrec.com/~martin/all_ti ... _men_5.pdf
http://backup.boxrec.com/~martin/all_ti ... vies_5.pdf

I like this a rating. question: why Vladimir Klitschko in 2003 not assessed

[computerrank]

I like this a rating. question: why Vladimir Klitschko in 2003 not assessed

Waldimir Klitschko had no top 20 opponent in 2003 and in 2001.

[mike1989]

These are the ratings for the last rule set:

- 100 rating points limit instead of 150 in order to better integrate the lower weights with lower level

- annual year's end rank in division with best performance
-- maximum rank is 20
-- 100 points / rank * weight
-- weight = number of division top 10 with rating higher than 100 points / 10
-- annual year's end rank points only if there was a ranked best opponent in the same year


- annual best opponent rank converted to bout division a time of bout
-- maximum opponent rank is 20
-- 100 points / (rank_opp-1) * weight_opp ; when rank_opp is 1 or 2 then use 2
-- weight_opp = number of bout division top 10 with rating higher than 150 points / 10 at time of bout


multiple opponent ranks per year not regarded - I tested this feature, but this inflated the ratings for the 1930ies to 1970ies, where boxers had more fight per year.


http://backup.boxrec.com/~martin/ann_tops_5.txt
http://backup.boxrec.com/~martin/ann_tops_years_5.txt
http://backup.boxrec.com/~martin/bouts_ranks_5.txt
http://backup.boxrec.com/~martin/all_ti ... _men_5.pdf
http://backup.boxrec.com/~martin/all_ti ... vies_5.pdf

only one problem remained, the rating of the fighters from the top 20. this is 95% of the total rating.
there is a proposal: in return the square of the maximum value of Ratha boxer shield waters for achievements : for winning the fight 12 rounds - 1 point 10 rounds - point 0.8 8 0.6 point rounds 6 point rounds 0.4 4 0.2 rounds point win in the title fight not a world champion win 2 point draw and lose 1 point battle for the title of world champion win 3 bluff draw and lose 2 Pont.

[mike1989]

2014 Mikey Garcia Super Featherweight and not Featherweight Orlando Salido top 2because Vasyl Lomachenko and not plus 50 plus 100

[computerrank]

2014 Mikey Garcia Super Featherweight and not Featherweight Orlando Salido top 2because Vasyl Lomachenko and not plus 50 plus 100

- Mikey Garcia's only bout from January 2014 was at super featherweight - right
- Orlando Salido was at super featherweight in both his 2014 bouts???

Something goes wrong ...

[computerrank]

only one problem remained, the rating of the fighters from the top 20. this is 95% of the total rating.
there is a proposal: in return the square of the maximum value of Ratha boxer shield waters for achievements : for winning the fight 12 rounds - 1 point 10 rounds - point 0.8 8 0.6 point rounds 6 point rounds 0.4 4 0.2 rounds point win in the title fight not a world champion win 2 point draw and lose 1 point battle for the title of world champion win 3 bluff draw and lose 2 Pont.

Your proposal is to give every opponent a weight depending on the bout result? Nice idea

I will try with the common Boxrec weights:

- clear decision factor = for the winner 1 for KO, TKO and clear decisions on points down to 0.5+ for close wins on points
- clear decision factor = for the draws = 0.5
- clear decision factor = for the loser 0.5- for close losses down to 0

- bout weight depending on rounds boxed for decisions on points
- weight = rounds boxed / 12

[mike1989]

2014 Mikey Garcia Super Featherweight and not Featherweight Orlando Salido top 2because Vasyl Lomachenko and not plus 50 plus 100

- Mikey Garcia's only bout from January 2014 was at super featherweight - right
- Orlando Salido was at super featherweight in both his 2014 bouts???

Something goes wrong ...

http://boxrec.com/media/index.php?title=Fight:1861656

[mike1989]

only one problem remained, the rating of the fighters from the top 20. this is 95% of the total rating.
there is a proposal: in return the square of the maximum value of Ratha boxer shield waters for achievements : for winning the fight 12 rounds - 1 point 10 rounds - point 0.8 8 0.6 point rounds 6 point rounds 0.4 4 0.2 rounds point win in the title fight not a world champion win 2 point draw and lose 1 point battle for the title of world champion win 3 bluff draw and lose 2 Pont.

Your proposal is to give every opponent a weight depending on the bout result? Nice idea

I will try with the common Boxrec weights:

- clear decision factor = for the winner 1 for KO, TKO and clear decisions on points down to 0.5+ for close wins on points
- clear decision factor = for the draws = 0.5
- clear decision factor = for the loser 0.5- for close losses down to 0

- bout weight depending on rounds boxed for decisions on points
- weight = rounds boxed / 12

additional poin should be in title and championship bouts that will help to objectively establish the ranking top 20

[computerrank]

only one problem remained, the rating of the fighters from the top 20. this is 95% of the total rating.
there is a proposal: in return the square of the maximum value of Ratha boxer shield waters for achievements : for winning the fight 12 rounds - 1 point 10 rounds - point 0.8 8 0.6 point rounds 6 point rounds 0.4 4 0.2 rounds point win in the title fight not a world champion win 2 point draw and lose 1 point battle for the title of world champion win 3 bluff draw and lose 2 Pont.

Your proposal is to give every opponent a weight depending on the bout result? Nice idea

I will try with the common Boxrec weights:

- clear decision factor = for the winner 1 for KO, TKO and clear decisions on points down to 0.5+ for close wins on points
- clear decision factor = for the draws = 0.5
- clear decision factor = for the loser 0.5- for close losses down to 0

- bout weight depending on rounds boxed for decisions on points
- weight = rounds boxed / 12

additional poin should be in title and championship bouts that will help to objectively establish the ranking top 20

mike1989,

Boxrec ratings per rule and good reason do not go for titles. Titles are in the hands of genuine corrupt organisations ... which not give a penny for objectivity.

[mike1989]

mike1989,

Boxrec ratings per rule and good reason do not go for titles. Titles are in the hands of genuine corrupt organisations ... which not give a penny for objectivity.

I agree. so you try to apply this idea in return for the square root of the maximum rating ????

[mike1989]

all the same, to complete objectivity I need the shield rating of the annual water charge and all the best for the vanquished opponents to do the top 20. ??? what do you think ???

[computerrank]

New release:

- division assignment corrected

- annual year's end rank in division with best performance
-- maximum rank is 20
-- 100 points / rank * weight
-- weight = number of division top 10 with rating higher than 100 points / 10
-- annual year's end rank points only if there was a ranked best opponent in the same year

- annual best opponent rank converted to bout division a time of bout
-- maximum opponent rank is 20
-- 100 points / (rank_opp-1) * weight_opp ; when rank_opp is 1 or 2 then use 2
-- weight_opp = number of bout division top 10 with rating higher than 150 points / 10 at time of bout

- clear decision factor = for the winner 1 for KO, TKO and clear wins on points
- clear decision factor = 0.99 .. 0.51 for close wins on points
- clear decision factor = for the draws = 0.5
- clear decision factor = 0.49 .. 0.01 for close losses on points
- clear decision factor = for the loser 0 for KO, TKO and clear losses on points

- bout weight depending on rounds boxed for decisions on points
- weight = rounds boxed / 12


http://backup.boxrec.com/~martin/ann_tops_6.txt
http://backup.boxrec.com/~martin/ann_tops_years_6.txt
http://backup.boxrec.com/~martin/bouts_ranks_6.txt
http://backup.boxrec.com/~martin/all_ti ... _men_6.pdf
http://backup.boxrec.com/~martin/all_ti ... ight_6.pdf

[mike1989]

New release:

- clear decision factor = for the winner 1 for KO, TKO and clear wins on points
- clear decision factor = 0.99 .. 0.51 for close wins on points
- clear decision factor = for the draws = 0.5
- clear decision factor = 0.49 .. 0.01 for close losses on points
- clear decision factor = for the loser 0 for KO, TKO and clear losses on points
- bout weight depending on rounds boxed for decisions on points
- weight = rounds boxed / 12

http://backup.boxrec.com/~martin/ann_tops_6.txt
http://backup.boxrec.com/~martin/ann_tops_years_6.txt
http://backup.boxrec.com/~martin/bouts_ranks_6.txt
http://backup.boxrec.com/~martin/all_ti ... _men_6.pdf
http://backup.boxrec.com/~martin/all_ti ... ight_6.pdf

I don't see the application here of this piece???
a little can not understand the changes ???
Example: why Lomachenko is so little point for Orlando Salido ???
As naschet shield annual ranking points as not all top 20 ???

[computerrank]

New release:

- clear decision factor = for the winner 1 for KO, TKO and clear wins on points
- clear decision factor = 0.99 .. 0.51 for close wins on points
- clear decision factor = for the draws = 0.5
- clear decision factor = 0.49 .. 0.01 for close losses on points
- clear decision factor = for the loser 0 for KO, TKO and clear losses on points
- bout weight depending on rounds boxed for decisions on points
- weight = rounds boxed / 12

http://backup.boxrec.com/~martin/ann_tops_6.txt
http://backup.boxrec.com/~martin/ann_tops_years_6.txt
http://backup.boxrec.com/~martin/bouts_ranks_6.txt
http://backup.boxrec.com/~martin/all_ti ... _men_6.pdf
http://backup.boxrec.com/~martin/all_ti ... ight_6.pdf

I don't see the application here of this piece???
a little can not understand the changes ???
Example: why Lomachenko is so little point for Orlando Salido ???
As naschet shield annual ranking points as not all top 20 ???

applied to the Lomachenko loss SD Lomachenko by judges: Jack Reiss 112-116, Oren Shellenberger 113-115, Levi Martinez 115-113 - 12 rounds

Salido #3 featherweight = 50 points
clear decision factor for 330:334: point difference / number of judges / number rounds = 4 / 3 / 12 = 0.11

as Lomachenko lost - so his weight is 0.50 - 0.11 = 0.39 only

so he gets 50 * 0.39 = 19 points.

[computerrank]

Correction (20+ opponent ranks caused points)

Now new standings including all top 20 opponents in bouts_rank_7.txt

http://backup.boxrec.com/~martin/ann_tops_7.txt
http://backup.boxrec.com/~martin/ann_tops_years_7.txt
http://backup.boxrec.com/~martin/bouts_ranks_7.txt
http://backup.boxrec.com/~martin/all_ti ... _men_7.pdf
http://backup.boxrec.com/~martin/all_ti ... ight_7.pdf

[mike1989]

New release:

I don't see the application here of this piece???
a little can not understand the changes ???
Example: why Lomachenko is so little point for Orlando Salido ???
As naschet shield annual ranking points as not all top 20 ???

applied to the Lomachenko loss SD Lomachenko by judges: Jack Reiss 112-116, Oren Shellenberger 113-115, Levi Martinez 115-113 - 12 rounds

Salido #3 featherweight = 50 points
clear decision factor for 330:334: point difference / number of judges / number rounds = 4 / 3 / 12 = 0.11

as Lomachenko lost - so his weight is 0.50 - 0.11 = 0.39 only

so he gets 50 * 0.39 = 19 points.

Salido is No. 2 and not No. 3. because 100 point????
The formula that you propose is too complicated. I propose a more simple and clear option for the average user:
12rounds=1; 10rounds= 10/12=0,83; 8rounds= 8/12= 0,67; 6rounds= 0,5; 4rounds= 0,33.
KO, TKO= 1; UD= 1 if one of the three suddy will give not less than 120-108.
further according to the following formula :
12rounds= 1/36 = -0,027 progeny for every round winner;
10rounds= 1/30 = -0,03 progeny for every round winner*0,83;
8rounds= 1/24 = -0,04 progeny for every round winner*0,67;
6rounds= 1/18 = -0,055 progeny for every round winner*0,5;
4rounds= 1/12 = -0,083 progeny for every round winner*0,33;

[computerrank]

[
Salido is No. 2 and not No. 3. because 100 point????
The formula that you propose is too complicated. I propose a more simple and clear option for the average user:
12rounds=1; 10rounds= 10/12=0,83; 8rounds= 8/12= 0,67; 6rounds= 0,5; 4rounds= 0,33.
KO, TKO= 1; UD= 1 if one of the three suddy will give not less than 120-108.
further according to the following formula :
12rounds= 1/36 = -0,027 progeny for every round winner;
10rounds= 1/30 = -0,03 progeny for every round winner*0,83;
8rounds= 1/24 = -0,04 progeny for every round winner*0,67;
6rounds= 1/18 = -0,055 progeny for every round winner*0,5;
4rounds= 1/12 = -0,083 progeny for every round winner*0,33;

Salido was featheweight #3 on 2014-03-01 - behind MIkey Garcia #1 and Jhonny Gonzales #2

On 2014-03-01 Mikey Garcia was still rated at featherweight because his featherweight bout from 2013-06-15 against Juan Manuel Lopez was valid more than his later super featherweight bouts AND the Lopez bout was within the last 12 months.

I use the same standard formula as in the current ratings - there is nothing new about it

[mike1989]

[

Salido was featheweight #3 on 2014-03-01 - behind MIkey Garcia #1 and Jhonny Gonzales #2

On 2014-03-01 Mikey Garcia was still rated at featherweight because his featherweight bout from 2013-06-15 against Juan Manuel Lopez was valid more than his later super featherweight bouts AND the Lopez bout was within the last 12 months.

I use the same standard formula as in the current ratings - there is nothing new about it

in Mikey Garcia's last two fights were both super featherweight champion:
http://boxrec.com/media/index.php?title=Fight:1832225
http://boxrec.com/media/index.php?title=Fight:1855038
because for featherweight Mikey Garcia as they say here in Ukraine : ні в які ворота ))))... in any gate ))))
later, Salido №2 )))

how about the fact that lagodny points to the boxers nsitivity without restrictions this will make an adequate rating for boxers below the top 20, and featured opoziciju defeated the shield in the top 20 ???

[computerrank]

in Mikey Garcia's last two fights were both super featherweight champion:
http://boxrec.com/media/index.php?title=Fight:1832225
http://boxrec.com/media/index.php?title=Fight:1855038
because for featherweight Mikey Garcia as they say here in Ukraine : ні в які ворота ))))... in any gate ))))
later, Salido №2 )))

how about the fact that lagodny points to the boxers nsitivity without restrictions this will make an adequate rating for boxers below the top 20, and featured opoziciju defeated the shield in the top 20 ???

Yes, Garcia won 2 title bouts later. But titles do not count at Boxrec ratings. The boxer is rated at the division, where he had his most valid bout within the last 12 months. And on 2014-03-01 this was Garcia's featherweight bout from 2013-06-15.

The ranks regarded are constrained to max top 20. Otherwise the boxers in the 1930ies to 1970ies with their inflated ratings would have an advantage.

[mike1989]

The ranks regarded are constrained to max top 20. Otherwise the boxers in the 1930ies to 1970ies with their inflated ratings would have an advantage.

I propose to charge the soldiers ezhegodnye rating and a reasonable bat of opozicii to do top 20. this will not affect the age of 30-70 years .??????????

[computerrank]

The ranks regarded are constrained to max top 20. Otherwise the boxers in the 1930ies to 1970ies with their inflated ratings would have an advantage.

I propose to charge the soldiers ezhegodnye rating and a reasonable bat of opozicii to do top 20. this will not affect the age of 30-70 years .??????????

Sorry, I do not understand, please explain.

[mike1989]

I propose to charge the soldiers ezhegodnye rating and a reasonable bat of opozicii to do top 20. this will not affect the age of 30-70 years .??????????

Sorry, I do not understand, please explain.

boxer that would get ejegodnyi rating point should beat boxer with the top 20. during the year, such teams can be an average of 20-30 fighters, respectively 20-30 boxers can obtain an annual credit rating point, the rest is not needed?????
I offer annual rating earned by all, and for the defeated opoziciju limited to the top 20 ???

[computerrank]

boxer that would get ejegodnyi rating point should beat boxer with the top 20. during the year, such teams can be an average of 20-30 fighters, respectively 20-30 boxers can obtain an annual credit rating point, the rest is not needed?????
I offer annual rating earned by all, and for the defeated opoziciju limited to the top 20 ???

The current version does exactly this.

Every boxer boxing in top 20 in a close fight gets annual points for his best opponent in top 20.

Additionally every boxer with an annual end rank in top 20 gets additional annual points for this annual end rank, if he boxes an opponent in top20 .

Example:

A boxer with annual end rank #21 and best opponent rank #20 gets points for this opponent only.

A boxer with annual end rank #20 and best opponent rank #20 gets points for his opponent AND for his annual end rank.

A boxer with annual end rank #20 and best opponent rank #21 gets points gets no points at all.