Wuornos
Professional
As it has been difficult to find acceptance with the Elo rating system because of its relative complexity I have devised a simplified system that enables the comparison of players across era.
Like the Elo system this new methodology assumes the standard of the general population remains constant as this seemed to gain general acceptance due to the belief that players need to be measured against their contemporaries and their comparative levels of success within their time being the criteria on which they are judged.
The Elo system was prone to criticism on the following two points
1. People compared it within the confines of volumes of achievement within a single year. I believe this was due to the fact that tennis fans are used to thinking in this way and were not accustomed to a system that simply monitored quality of results rather than volume. I have therefore changed the system to be volume based, i.e. play for only six months of the year and take the consequences of less points and a lower rating. This results in faster changes in rankings and the current ratings are now more comparable with the ATP / WTA.
2. Of course if point 1 was all there was to the new system then people could simply look at the ATP / WTA ratings. However, the new system is self adjusting in that it places a weighted value on each event depending on the players present and their current rating. This should not have to great an impact on modern rankings but has a greater impact on historical events in that it reduces the value of winning majors like the Australian in the late 1970s and early 1980s. This helps provide a more valid rating for people from this era when compared with their contemporaries. The devaluing of the Australian results in the inflation of the other majors during this time as it is important to keep the volume of points available at a constant.
3. Peak ratings are used to show a players standard as career volumes can be seriously marred for outstanding players due to early career finishes etc.
4. While there are similarities between the points shown and the official ATP / WTA rankings these will differ due to the automated weighting of events based on participants. I felt it was important to maintain this process as even now winning an event which does not include Justine Henin or Roger Federer in the field must attract less points than events that do etc. This approach has been criticised in the past as having a form of rating feedback, i.e. Henin and Federer would benefit as their events would always have more weighting awarded. This is true and it does make it slightly more difficult for others to catch them. However, the emphasis is on slightly, and the benefits of being able to automatically devalue past events that were not considered as important as they may be today far outweighs this minor skewing of ouput.
In summary then think ATP WTA ranking system with a self adjusting points table for each event depending on the current rating of players present. I hope that explains this simple system well enough, but please ask if you have any questions.
As always I would welcome feedback from anyone and everyone, both good and bad, especially if there are comparisons between these data and those calculated under the none volume based Elo system.
Ok first of all the men’s current top 10 remember the points will be different as the sytem will decide that some events are deserving of more points than the official system, while some will receive less.
1. Roger Federer 11,701
2. Rafael Nadal 8,054
3. Novak Đoković 7,356
4. Nikolay Davydenko 5,354
5. Andy Roddick 4,962
6. Fernando González 4,911
7. Richard Gasquet 4,171
8. David Ferrer 3,579
9. Tommy Haas 3,473
10. Carlos Moyà 3,463
Previously in the Elo ratings current top 10 were players like David Nalbandian and Lleyton Hewitt. These players are much further down the rankings now due to time limit of 1 year despite David Nalbandian’s recent success.
As you can see the point count is markedly different to the current ATP rankings due the event weighting, however each of the top 5 are the same and of the 6 – 10 positions three are the same with only James Blake and Tommy Robredo being missing in fact Tommy Robredo is 11th on my list with 3,450. While my list includes Richard Gasquet who is 11th on the official ATP list. So some differences but not major.
In the womens rankings we have the following current points:
1. Justine Henin 8,209
2. Serena Williams 6,061
3. Venus Williams 5,743
4. Ana Ivanović 5,407
5. Svetlana Kuznetsova 5,274
6. Maria Sharapova 5,073
7. Jelena Janković 3,699
8. Marion Bartoli 3,617
9. Anna Chakvetadze 3,550
10. Nicole Vaidišová 3,212
Here we can see Kuznetsova and Janković appearing lower down in my rankings than they do in the official list. This is because the weighting awarded by the automated system seems to be proportionately lower based on rating of participant for minor events when compared with majors. Hence Janković who never shows well at the big events suffers proportionately as has Kuznetsova. Venus Williams who missed part of this year due to injury made up much ground due to the heavy weighting given to Wimbledon. However, it is interesting to note that Wimbledon was considered the weakest of the women’s majors this year with the French being the strongest event. While the order may be slightly different in these ranking compared with the offical WTA data, 9 of the top 10 players are the same. Only Nicole Vaidišová in 10th place differs from the official list which has Daniela Hantuchova in the final spot.
Now onto the men’s peak annual points for the open era. Please bear in mind these points are adjusted in such a way as to nullify the different emphasis on events for each era. They use the same automated event weighting that so closely replicated the official rankings above.
1. Rod Laver 11,966 Achieved in calendar year of 1969
2. Roger Federer 11,701 Current points
3. John McEnroe 10,749 Achieved in year ending after the US open of 1984
4. Ivan Lendl 10,372 Achieved in the year ending after the Australian open of 1988
5. Pete Sampras 10,040 Achieved in the year ending after the Australian open of 1994
6. Mats Wilander 9,905 Achieved in the year ending after the French open of 1988
7. Andre Agassi 9,880 Achieved in the year ending after the Australian open of 2001
8. Jim Courier 9,370 Achieved in the year ending after the French open of 1992
9. Björn Borg 9,109 Achieved in the year ending after Wimbledon of 1979
10. Jimmy Connors 8,705 Achieved in the year ending after the Australian open of 1975
As stated in my commentary relating to Elo calculations, a single year is very short space of time and I am concerned looking at some of these data that the sample time frame is to small and the ‘luck’ element may have played to great a part. I would especially welcome peoples views on this and on a further concern which is Borg’s position at number 9 in the rankings.
Now for the peak Women’s points of the open era.
1. Martina Navratilova 12,642 Achieved 1986 (inc Australian open of Dec 1985)
2. Steffi Graf 12,417 Achieved in Calendar year of 1988
3. Monica Seles 12,377 Achieved in year up to the end of the Australian open 1993
4. Serena Williams 12,274 Achieved in year up to the end of the Australian open 2003
5. Margaret Smith Court 12,190 Achieved in year up to the end of the Australian open 1971
6. Chris Evert 11,714 Achieved in year up to the end of the French open 1985
7. Justine Henin 11,144 Achieved in year up to the end of the Australian open 2004
8 Martina Hingis 10,621 Achieved in year up to the end of the Australian open 1998
9. Hana Mandlíková 10,232 Achieved in year up to the end of Wimbledon 1981
10. Billie Jean King 10,209 Achieved in Calendar year of 1972
So there it is. What most posters on tis forum seemed to consider the best indicator of performance. Points calculated over a single year, automated adjustment for differences in relative values of event cross era and the system based on a cumulative volume of points gathered.
Which do you prefer, Elo or this?
Take care all
Tim
Like the Elo system this new methodology assumes the standard of the general population remains constant as this seemed to gain general acceptance due to the belief that players need to be measured against their contemporaries and their comparative levels of success within their time being the criteria on which they are judged.
The Elo system was prone to criticism on the following two points
1. People compared it within the confines of volumes of achievement within a single year. I believe this was due to the fact that tennis fans are used to thinking in this way and were not accustomed to a system that simply monitored quality of results rather than volume. I have therefore changed the system to be volume based, i.e. play for only six months of the year and take the consequences of less points and a lower rating. This results in faster changes in rankings and the current ratings are now more comparable with the ATP / WTA.
2. Of course if point 1 was all there was to the new system then people could simply look at the ATP / WTA ratings. However, the new system is self adjusting in that it places a weighted value on each event depending on the players present and their current rating. This should not have to great an impact on modern rankings but has a greater impact on historical events in that it reduces the value of winning majors like the Australian in the late 1970s and early 1980s. This helps provide a more valid rating for people from this era when compared with their contemporaries. The devaluing of the Australian results in the inflation of the other majors during this time as it is important to keep the volume of points available at a constant.
3. Peak ratings are used to show a players standard as career volumes can be seriously marred for outstanding players due to early career finishes etc.
4. While there are similarities between the points shown and the official ATP / WTA rankings these will differ due to the automated weighting of events based on participants. I felt it was important to maintain this process as even now winning an event which does not include Justine Henin or Roger Federer in the field must attract less points than events that do etc. This approach has been criticised in the past as having a form of rating feedback, i.e. Henin and Federer would benefit as their events would always have more weighting awarded. This is true and it does make it slightly more difficult for others to catch them. However, the emphasis is on slightly, and the benefits of being able to automatically devalue past events that were not considered as important as they may be today far outweighs this minor skewing of ouput.
In summary then think ATP WTA ranking system with a self adjusting points table for each event depending on the current rating of players present. I hope that explains this simple system well enough, but please ask if you have any questions.
As always I would welcome feedback from anyone and everyone, both good and bad, especially if there are comparisons between these data and those calculated under the none volume based Elo system.
Ok first of all the men’s current top 10 remember the points will be different as the sytem will decide that some events are deserving of more points than the official system, while some will receive less.
1. Roger Federer 11,701
2. Rafael Nadal 8,054
3. Novak Đoković 7,356
4. Nikolay Davydenko 5,354
5. Andy Roddick 4,962
6. Fernando González 4,911
7. Richard Gasquet 4,171
8. David Ferrer 3,579
9. Tommy Haas 3,473
10. Carlos Moyà 3,463
Previously in the Elo ratings current top 10 were players like David Nalbandian and Lleyton Hewitt. These players are much further down the rankings now due to time limit of 1 year despite David Nalbandian’s recent success.
As you can see the point count is markedly different to the current ATP rankings due the event weighting, however each of the top 5 are the same and of the 6 – 10 positions three are the same with only James Blake and Tommy Robredo being missing in fact Tommy Robredo is 11th on my list with 3,450. While my list includes Richard Gasquet who is 11th on the official ATP list. So some differences but not major.
In the womens rankings we have the following current points:
1. Justine Henin 8,209
2. Serena Williams 6,061
3. Venus Williams 5,743
4. Ana Ivanović 5,407
5. Svetlana Kuznetsova 5,274
6. Maria Sharapova 5,073
7. Jelena Janković 3,699
8. Marion Bartoli 3,617
9. Anna Chakvetadze 3,550
10. Nicole Vaidišová 3,212
Here we can see Kuznetsova and Janković appearing lower down in my rankings than they do in the official list. This is because the weighting awarded by the automated system seems to be proportionately lower based on rating of participant for minor events when compared with majors. Hence Janković who never shows well at the big events suffers proportionately as has Kuznetsova. Venus Williams who missed part of this year due to injury made up much ground due to the heavy weighting given to Wimbledon. However, it is interesting to note that Wimbledon was considered the weakest of the women’s majors this year with the French being the strongest event. While the order may be slightly different in these ranking compared with the offical WTA data, 9 of the top 10 players are the same. Only Nicole Vaidišová in 10th place differs from the official list which has Daniela Hantuchova in the final spot.
Now onto the men’s peak annual points for the open era. Please bear in mind these points are adjusted in such a way as to nullify the different emphasis on events for each era. They use the same automated event weighting that so closely replicated the official rankings above.
1. Rod Laver 11,966 Achieved in calendar year of 1969
2. Roger Federer 11,701 Current points
3. John McEnroe 10,749 Achieved in year ending after the US open of 1984
4. Ivan Lendl 10,372 Achieved in the year ending after the Australian open of 1988
5. Pete Sampras 10,040 Achieved in the year ending after the Australian open of 1994
6. Mats Wilander 9,905 Achieved in the year ending after the French open of 1988
7. Andre Agassi 9,880 Achieved in the year ending after the Australian open of 2001
8. Jim Courier 9,370 Achieved in the year ending after the French open of 1992
9. Björn Borg 9,109 Achieved in the year ending after Wimbledon of 1979
10. Jimmy Connors 8,705 Achieved in the year ending after the Australian open of 1975
As stated in my commentary relating to Elo calculations, a single year is very short space of time and I am concerned looking at some of these data that the sample time frame is to small and the ‘luck’ element may have played to great a part. I would especially welcome peoples views on this and on a further concern which is Borg’s position at number 9 in the rankings.
Now for the peak Women’s points of the open era.
1. Martina Navratilova 12,642 Achieved 1986 (inc Australian open of Dec 1985)
2. Steffi Graf 12,417 Achieved in Calendar year of 1988
3. Monica Seles 12,377 Achieved in year up to the end of the Australian open 1993
4. Serena Williams 12,274 Achieved in year up to the end of the Australian open 2003
5. Margaret Smith Court 12,190 Achieved in year up to the end of the Australian open 1971
6. Chris Evert 11,714 Achieved in year up to the end of the French open 1985
7. Justine Henin 11,144 Achieved in year up to the end of the Australian open 2004
8 Martina Hingis 10,621 Achieved in year up to the end of the Australian open 1998
9. Hana Mandlíková 10,232 Achieved in year up to the end of Wimbledon 1981
10. Billie Jean King 10,209 Achieved in Calendar year of 1972
So there it is. What most posters on tis forum seemed to consider the best indicator of performance. Points calculated over a single year, automated adjustment for differences in relative values of event cross era and the system based on a cumulative volume of points gathered.
Which do you prefer, Elo or this?
Take care all
Tim
Last edited:
