[GOAT] Which kind of domination?

[Nickognito]

Hi everyone. I'm a new user and I recently found this very interesting forum with great analysis for the GOAT debate.

I would like to congrat everyone, but most of all Wuornos and his great rankings.
I would like to ask him if he could kindly share his enormous statisical material with results, DOT and ELO rankigs, etc. It would be great.

I would like to ask him aso what he thinks about a bayesian elo rating for tennis (http://remi.coulom.free.fr/Bayesian-Elo/#links)

I think that we expect the greater player of the time to have different kind of domination:
- domination in a single match
- domination in one year
- domination in few years
- domination in a decade or more

The GOAT in not necessarily the best player in each category, but I think he has to be of the best.

Vines, or Hoad are between the best in the single match, but not for years.
McEnroe is great in the single match , in one year, and maybe in the span of five years, but we think that his career is too short at this level.
Lendl was dominating for one year and for a span of three years like few players did, and he was competitive at a very high levale for more than a decade, but maybe he's not as competitive in the single match, where he's maybe not better than McEnroe, Becker or even Edberg and Wilander.

So, I think that the question 'it's better to dominate five years and retire or to win slams for 15 years without such domination?' is incorrect.
The GOAT must be one of the best in each category, and a ranking must have a 'value' for each category.
So maybe the DOT ratings are a wonderful way to let us know a kind of domination of a player (maybe the more important), bit i think that it has to be integrated with other different kind of ratings.

What do you think about it?

Just to tell something about me: I'm italian, i'm debating this kind of questions for years in an italian forum, and so it's very intersting to be here. I subjectively think that Pancho Gonzales is maybe the GOAT, even if I am a Pete Sampras fan.

Regards,

Cristiano

[SgtJohn]

Hi Cristiano, and welcome to the forum,

Glad you are interested in the GOAT debate as well! If there were interesting elements of discussion on your italian forum that you think is worth translating for us, you are more than welcome, by the way!

I think your idea is interesting, but mostly adapted for the Open Era...
You mention Vines or Hoad as great one-match players; the problem is, this reputation comes mainly from a few quotes from other greats (Tilden about Vines, Rosewall or Laver about Hoad). I'm not ready to trust quotes by people who had friendly relationships with the players involved. I mean I think they were *probably* among the very greatest when they were 'on', but without more objective proof, I won't base any rankings on this... The only thing that could provide, if not proof, material for discussion is video, and pre-Open era players' clips are too rare for this.
About one-year greatness, like you I think that Wuornos' rankings are interesting, but as far as I know, his database only goes so far as 1968 or so, and is not totally complete before 1971. But such a complete database of every matches played is necessary to assert a player's one-year performance in a precise way, and it can't be properly done before the 70's.

As I'm interested in global, 1877-2007, rankings for tennis, I have to conclude that the only criteria we can rely on are the most basic: 'True Majors' won, and year-end ranking (it might not be possible to objectively compare performances in different years, but it is for two players from the same year). Of course, if someone could provide me a book with a list of every tennis match played from 1877 (more like some hundred books then!), I'd be happy to get along with the ELO or DOT rankings, but it's unfortunately not possible.
That's why, I think, the only way to compare players from different eras is to look at their whole career, and nothing else...

Jonathan

[Nickognito]

Good point, John. Wuornos seems to be more interested in a statistical objective-only kind of debate: I think we cannot do this for pre-open tennis, and he's right to do it only for open tennis.

I think that to consider just majors and years at number one is a too poor statistic for this question. It's a good one, but it's too poor. The problem is that maybe the three best players of all times played in the pro-tennis era, and one of the first five, maybe, in a non-competitive era. So I think i'ts impossible to find stats who tell us who is the GOAT. It's maybe possible to do it with subjective considerations based upon different kinds of objective stats.

Anyway, I think the the best player of the year 2007 is the player who has the best results in matches between the first 100-130 players of 2007.
And so i think that the GOAT is the player who would have the best results with the first 100-130 players of all times. So I believe that head-to-heads with other players (>130 all times) do not count for our purpose.

Challengers and futures are non important for the first positions of 2007 ranking.
So I think that only grand slam tournaments (and maybe master series or similar) are important in this debate.

I think that Dot Rankins or Elo rankings are wonderful because they assign a minimum value to wins over bad players.

With some imagination, we can maybe 'build' some imaginary slam for the pro-tennis era, based upon pros results before becoming pro, and head-to-heads between them. Jack Kramer 'imagined' the winners of this hypothetical slams in his book 'The Game'.

Anyway, it would be a subjective way to solve the question, obviously.

regards

cristiano

[noeledmonds]

I think it is important to emphasise the importantce of versitility along with dominance. This is what, in my subjective opinion, makes Laver stand out from the rest as the greatest.

Sampras, Federer, McEnroe, Tilden* and Connors** have failed to win the French Open. Borg failed to win the US Open. Lendl and Rosewall*** failed to win Wimbledon. Gonzales failed to win the French Open or the French Pro equivalent.

*Tilden rarely made the journey to play the French Open and he did win the French Pro in 1934
** Connors was banned from the French Open during his best years
*** Rosewall played the best part of his career unable to play Wimbledon due to the pro/amateur divide

Then if we look at the dominance of the players who achieved the Career Grand Slam, Laver his well above the rest. Perry was dominant but ultimately less so than his rival Budge. Budge himself was very dominant, but only for a short period of a few years. Emerson was dominant, but mainly due to the fact he remained amateur when his supiror contemperies were professional. Agassi had an up and down career and was inconsistant throughout.

I believe that Tilden is the one who comes cloeset to Laver. Tilden's dominance was supiror to Laver. Tilden maintained a winning percentage of over 90% for over 7 years. However Laver was dominant across all four surfaces for nearly as long and of course achieved the Grand Slam twice and the Pro Slam once. Many foget that Laver was past his best when he won his final Grand Slam in 1969.

[Moose Malloy]

Connors was banned from the French Open during his best years

He was only banned from the French one year('74). The other years he skipped it('75-'7 because he was ****ed off at having been banned in '74.

[Nickognito]

I agree with you, Noel, and this is the main reason for not judging 'my' Sampras the best ever.

Anyway, Laver has never won with the best, young Rosewall on Clay, Federer should have won without Nadal (but maybe not, with a Kuerten or Courier too) and so opponents are very important.

For Kramer it was different. In the '40 and early '50, american players do not care of playing on clay. And for Tilden too (Tilden anyway won the French later in his career). So the importance of the 4 slam has never be the same. And surfaces are not always the same too. Wimbledon grass now is different, for example.

That's another problem. Imagine an alltime 4 grand slam tournaments with the best 128 players ever, and imagine Sampras or Kramer winning 3 of 4 slam and losing in the 1st round in the French: who will be the best, at the end of the imaginary year?.

Do you think that versality is important for an one-year ranking? And why for an all.time ranking do you think it's more important?

So, I agree that versality is important, but maybe not so important like you say..

Regards,

c.

[Wuornos]

Originally Posted by Nickognito

Hi everyone. I'm a new user and I recently found this very interesting forum with great analysis for the GOAT debate.

I would like to congrat everyone, but most of all Wuornos and his great rankings.
I would like to ask him if he could kindly share his enormous statisical material with results, DOT and ELO rankigs, etc. It would be great.

I would like to ask him aso what he thinks about a bayesian elo rating for tennis (http://remi.coulom.free.fr/Bayesian-Elo/#links)

I think that we expect the greater player of the time to have different kind of domination:
- domination in a single match
- domination in one year
- domination in few years
- domination in a decade or more

The GOAT in not necessarily the best player in each category, but I think he has to be of the best.

Vines, or Hoad are between the best in the single match, but not for years.
McEnroe is great in the single match , in one year, and maybe in the span of five years, but we think that his career is too short at this level.
Lendl was dominating for one year and for a span of three years like few players did, and he was competitive at a very high levale for more than a decade, but maybe he's not as competitive in the single match, where he's maybe not better than McEnroe, Becker or even Edberg and Wilander.

So, I think that the question 'it's better to dominate five years and retire or to win slams for 15 years without such domination?' is incorrect.
The GOAT must be one of the best in each category, and a ranking must have a 'value' for each category.
So maybe the DOT ratings are a wonderful way to let us know a kind of domination of a player (maybe the more important), bit i think that it has to be integrated with other different kind of ratings.

What do you think about it?

Just to tell something about me: I'm italian, i'm debating this kind of questions for years in an italian forum, and so it's very intersting to be here. I subjectively think that Pancho Gonzales is maybe the GOAT, even if I am a Pete Sampras fan.

Regards,

Cristiano

First of all I would like to thank you for your kind words Cristiano.

I'm not sure what data you would like to see from me in relation to DOT or Elo ratings, but if you let me know I'll see what I can do.

The article on Bayesian Elo is very interesting. One of the main assumptions of the ELO system, and it is a flawed assumption, is that the standard deviation in the level of performance from match to match is 200 points and this is taken for all players. Of course this is not true and some players may have a SD of 100 and others 300. I think the Bayesan system goes some way to rectifying this. I have no doubts that this system is superior, but I have two concerns that that are not really concerned with the accuracy.

1. I have worked as a statistician for many years and one thing that occurs over and over again and where the layman often loses his way, is the question of fitness for purpose and gold plating. By this I mean much statistical modelling can be subjected to the law of diminishing return. You can add in an additional attribute to an analysis carry out all the work to incorporate this element and the impact on the output is minimal. You often find you can get to the 80% accuracy mark by using 20% of the data available and with a relatively simplified system. I would therefore worry that the Bayesian System may yield marginally better results but at what cost?

2. While being independent and free from human opinion and iterference both ELO and Bayesian Rating systems struggle with inactivity. What do you do when a player becomes inactive. E.g. Amelie Mauresmo at the moment. How long do you include them in the rankings? Even worse, what do yoiu do when they return after a year out like Lindsay Davenport. Do you allow her to keep the same rating, surely her quality of play will have declined during the inactivity etc.

You are absolutely correct that domination varies with time period. Single match domination is not the same as longer term. The ability to hand a double or triple doughnut to another player is talking about a completely different subject to being able to dominate by winning against a large number of opponents over a set period of time.

For me the sample of a single match is to small and rather meaningless. I seem to recall someone posting on this site previously regarding a perfect match, i.e 6-0 6-0 and all 12 games being love game. For me this probably says more about the quality of opposition than it does about the quality of the winner.

Single year dominance is an altogether different matter. For me this period is again to small to be completely meaningful. If you just forget about the number of matches and how many wins and just think about the number of losses, the numbers is very small when considering top players within a single year. When comparing two players you can be talking about one who lost 4 matches in a year compared with one who lost 5 and this diffference is enough to effect the overall strike rate and consider one player superior.

A few years. This for me is the most significant factor. It is enough evidence for luck to have been largely eliminated while not being so long as it favours players with only long careers.

Domination over more than 10 years. I don't really call this domination. I tend to refer to it as career achievement. I don't like this as it undervalues players who have had a short career at the top. E.g Monica Seles. However players with along career have had the opportunity to dominate for shorter periods. Failure to do this means to me that they have had a long time nigglking away at the a level just short of the top of the game. Just my view.

For me the best indicator of standard of play or greatness is always going to be dominance over a few years with adjustments for quality of opposition and Tournamentvent. Hence DOT Ratings.

Take care and keep posting.

Tim

[Nickognito]

Thank you Tim.

I agree with you with the Dot, but i think that great results like winning a slam in years of no-domination need to be included in some way in the ranking.

About the single-match-performances, I think there's no a statistical way of consider it. Nevertheless, I think it's quite strange if we think that the 'Greater Player of All TImes' , at his best, would be beaten by 20 other players at their best, for example.

About the data I'd like you see, I mean, I have all data of the open tennis tournaments, but I don't have any excel, or access file with them, and If i can have them from you it would be great. Maybe you have dot rankings for every year too, or , to speak honestly, I would like to receive any statistical work you can share. I know that maybe i'm asking too much, so in that case, don't care. Anyway I can assure you that it's just a personal interest, and that I would not use any of your statistics without your authorization. If you are interested , you can contact me by email, If not, it doesn't matter.

Regards,

c.

[SgtJohn]

Originally Posted by Nickognito

I think that to consider just majors and years at number one is a too poor statistic for this question. It's a good one, but it's too poor. The problem is that maybe the three best players of all times played in the pro-tennis era, and one of the first five, maybe, in a non-competitive era. So I think i'ts impossible to find stats who tell us who is the GOAT. It's maybe possible to do it with subjective considerations based upon different kinds of objective stats.

I agree with you that it is poor statistic, but my point is precisely that you might generate good statistic for the Open Era because we have a wealth of data, but you can't do any statistics at all for the previous period. This is due to:
1. The obvious lack of data.
2. The constant change in what people call greatness.

Let me elaborate on that last point. Using data kindly provided to me by Karoly Mazak, I computed a list of players per number of titles won. That would seem like a very useful stat...unfortunately it's not! The number of events a player could play has changed too much with time for us to draw any conclusion...While Trabert was winning 15+ events on the 52-weeks amateur circuit, Gonzales won only a handful simply because there were too few pro events, for example. Gonzales won 92 events by my count, and Laver approximately twice more, but can we deduce anything regarding their level of play from this fact?
Similarly, you could argue that number of tournaments won is not that important, that win/loss record is more of a useful stat...That would be wrong too: head-to-head might have been significant in the 50s when 2-men or 4-men tours were a norm, but not at all in the mid and late 60s, when even dominant players suffered many defeats, simply because they played every event they could (see Laver's 1967). They didn't 'take care' of their W/L record as much as Federer does, or as Mac or Lendl did when the ATP rankings were based on it.
'Majors' won and year-end rankings are the only two factors I could think of that were constantly considered as indicating greatness.

Originally Posted by Nickognito

With some imagination, we can maybe 'build' some imaginary slam for the pro-tennis era, based upon pros results before becoming pro, and head-to-heads between them. Jack Kramer 'imagined' the winners of this hypothetical slams in his book 'The Game'.

Anyway, it would be a subjective way to solve the question, obviously.

That's what I was trying to do for some time, though it doesn't require 'imagination', but rather many documents, and trying to have an objective point of view, though I totally admit it's impossible not to have much subjectivity in such a study.

In case it is of interest for you, I will copy my *current* list in the next post. It's NOT an 'official' list I try to show here, just my point of view, I totally understand if many of you disagree on many details in it, I just copy it to show that the whole approach is at least sensible.
There are 4 'Majors' for all years (except the war years and 1877-1880), and at least one major on clay or a slow surface from 1899 on (I chose this date because it is the first time an event on clay was seriously considered as an international tournament, when the 'Championships of Europe' were organized in Bad Homburg).

[SgtJohn]

(Sorry about the terrible layout, I had to copy this from MS Word directly, and it didn't keep any of my page settings...)

2007
AO Federer (Gonzalez)
RG Nadal (Federer)
Wim Federer (Nadal)
USO Federer (Djokovic)

2006
AO Federer (Baghdatis)
RG Nadal (Federer)
Wim Federer (Nadal)
USO Federer (Roddick)

2005
AO Safin (Hewitt)
RG Nadal (Puerta)
Wim Federer (Roddick)
USO Federer (Agassi)

2004
AO Federer (Safin)
RG Gaudio (Coria)
Wim Federer (Roddick)
USO Federer (Hewitt)

2003
AO Agassi (Schuettler)
RG Ferrero (Verkerk)
Wim Federer (Philippoussis)
USO Roddick (Ferrero)

2002
AO Johansson (Safin)
RG Costa (Ferrero)
Wim Hewitt (Nalbandian)
USO Sampras (Agassi)

2001
AO Agassi (Clément)
RG Kuerten (Corretja)
Wim Ivanisevic (Rafter)
USO Hewitt (Sampras)

2000
AO Agassi (Kafelnikov)
RG Kuerten (Norman)
Wim Sampras (Rafter)
USO Safin (Sampras)

1999
AO Kafelnikov (Enqvist)
RG Agassi (Medvedev)
Wim Sampras (Agassi)
USO Agassi (Martin)

1998
AO Korda (Rios)
RG Moya (Corretja)
Wim Sampras (Ivanisevic)
USO Rafter (Philippoussis)

1997
AO Sampras (Moya)
RG Kuerten (Bruguera)
Wim Sampras (Pioline)
USO Rafter (Rusedski)

1996
AO Becker (Chang)
RG Kafelnikov (Stich)
Wim Krajicek (Washington)
USO Sampras (Chang)

1995
AO Agassi (Sampras)
RG Muster (Chang)
Wim Sampras (Becker)
USO Sampras (Agassi)

1994
AO Sampras (Martin)
RG Bruguera (Berasategui)
Wim Sampras (Ivanisevic)
USO Agassi (Stich)

1993
AO Courier (Edberg)
RG Bruguera (Courier)
Wim Sampras (Courier)
USO Sampras (Pioline)

1992
AO Courier (Edberg)
RG Courier (Korda)
Wim Agassi (Ivanisevic)
USO Edberg (Sampras)

1991
AO Becker (Lendl)
RG Courier (Agassi)
Wim Stich (Becker)
USO Edberg (Courier)

1990
AO Lendl (Edberg)
RG Gomez (Agassi)
Wim Edberg (Becker)
USO Sampras (Agassi)

1989
AO Lendl (Mecir)
RG Chang (Edberg)
Wim Becker (Edberg)
USO Becker (Lendl)

1988
AO Wilander (Cash)
RG Wilander (Leconte)
Wim Edberg (Becker)
USO Wilander (Lendl)

1987
AO Edberg (Cash)
RG Lendl (Wilander)
Wim Cash (Lendl)
USO Lendl (Wilander)

1986
RG Lendl (Pernfors)
Wim Becker (Lendl)
USO Lendl (Mecir)
Boca West Lendl (Wilander)

1985
AO Edberg (Wilander)
RG Wilander (Lendl)
Wim Becker (Curren)
USO Lendl (McEnroe)

1984
AO Wilander (Curren)
RG Lendl (McEnroe)
Wim McEnroe (Connors)
USO McEnroe (Lendl)

1983
AO Wilander (Lendl)
RG Noah (Wilander)
Wim McEnroe (Lewis)
USO Connors (Lendl)

1982
RG Wilander (Vilas)
Wim Connors (McEnroe)
USO Connors (Lendl)
Masters Lendl (McEnroe)

1981
RG Borg (Lendl)
Wim McEnroe (Borg)
USO McEnroe (Borg)
Masters Lendl (Gerulaitis)

1980
RG Borg (Gerulaitis)
Wim Borg (McEnroe)
USO McEnroe (Borg)
Masters Borg (Lendl)

1979
RG Borg (Pecci)
Wim Borg (Tanner)
USO McEnroe (Gerulaitis)
Masters Borg (Gerulaitis)

1978
Philadelphia Connors (Tanner)
RG Borg (Vilas)
Wim Borg (Connors)
USO Connors (Borg)

1977
RG Vilas (Gottfried)
Wim Borg (Connors)
USO Vilas (Connors)
Masters Connors (Borg)

1976
Philadelphia Connors (Borg)
RG Panatta (Solomon)
Wim Borg (Nastase)
USO Connors (Borg)

1975
RG Borg (Vilas)
Wim Ashe (Connors)
USO Orantes (Connors)
Masters Nastase (Borg)

1974
WCT Finals Newcombe (Borg)
RG Borg (Orantes)
Wim Connors (Rosewall)
USO Connors (Rosewall)

1973
WCT Finals Smith (Ashe)
RG Nastase (Pilic)
USO Newcombe (Kodes)
Masters Nastase (Okker)

1972
WCT Finals Rosewall (Laver)
RG Gimeno (Proisy)
PSW Los Angeles Smith (Tanner)
USO Nastase (Ashe)

1971
AO Rosewall (Ashe)
Rome Laver (Kodes)
Wim Newcombe (Smith)
USO Smith (Kodes)

1970
Sydney Laver (Rosewall)
Wim Newcombe (Rosewall)
USO Rosewall (Roche)
Barcelone Santana (Laver)

1969
AO Laver (Gimeno)
RG Laver (Rosewall)
Wim Laver (Newcombe)
USO Laver (Roche)

1968
RG Rosewall (Laver)
Wimbledon Laver (Roche)
PSW Los Angeles Laver (Rosewall)
USO Ashe (Okker)

1967
Wembley Laver (Rosewall)
World Pro Laver (Rosewall)
Wimbledon Pro Laver (Rosewall)
US Pro Laver (Gimeno)

1966
Wembley Laver (Rosewall)
Barcelona Gimeno (Rosewall)
New York MSG Pro Rosewall (Laver)
US Pro Laver (Rosewall)

1965
French Pro Rosewall (Laver)
Milan Pro Gimeno (Rosewall)
Wembley Laver (Gimeno)
US Pro Rosewall (Laver)

1964
Noordwijk Pro Gimeno (Rosewall)
French Pro Rosewall (Laver)
Wembley Laver (Rosewall)
US Pro Laver (Gonzales)

1963
French Pro Rosewall (Laver)
Wembley Rosewall (Hoad)
US Pro Rosewall (Laver)
Italian Pro Rosewall (Laver)

1962
French Pro Rosewall (Gimeno)
Wembley Rosewall (Hoad)
Stockholm Pro Rosewall (Gimeno)
Wimbledon Laver (Mulligan)

1961
Geneva Pro Gonzales (Rosewall)
Copenhagen Pro Gonzales (Olmedo)
French Pro Rosewall (Gonzales)
Wembley Rosewall (Hoad)

1960
Victorian Pro Hoad (Rosewall)
World Cham’p Tour Gonzales (Rosewall)
French Pro Rosewall (Hoad)
Wembley Rosewall (Segura)

1959
Melbourne Pro Sedgman (Gonzales)
French Pro Trabert (Sedgman)
Los Angeles Pro RR Gonzales (Hoad)
Tourn. Of Champ. Hoad (Gonzales)

1958
Australian Pro Sedgman (Trabert)
French Pro Rosewall (Hoad)
Wembley Sedgman (Trabert)
Tourn. Of Champ. Gonzales (Rosewall)

1957
Australian Pro Segura (Sedgman)
Wembley Rosewall (Segura)
Tourn. Of Champ. Gonzales (Sedgman)
RG Davidson (Flam)

1956
French Pro Trabert (Gonzales)
Wembley Gonzales (Sedgman)
Tourn. Of Champ. Gonzales (Sedgman)
Wimbledon Hoad (Rosewall)

1955
US Pro Gonzales (Segura)
US Pro Hard Court Gonzales (Segura)
Scarborough Pro Gonzales (Segura)
RG Trabert (Davidson)

1954
Australian Pro Sedgman (Segura)
US Pro Gonzales (Sedgman)
US 4-men Tour Gonzales (Segura)
Roland Garros Trabert (Larsen)

1953
New York Indoor Kramer (Sedgman)
Scarborough Pro Segura (Sedgman)
Wembley Sedgman (Gonzales)
Paris Pro Sedgman (Gonzales)

1952
Phil. Indoor Pro Gonzales (Segura)
Wembley Gonzales (Kramer)
US Pro Segura (Gonzales)
Wimbledon Sedgman (Drobny)

1951
Phil. Indoor Pro Kramer (Gonzales)
Berlin Pro Segura (Earn)
Wembley Gonzales (Segura)
US Pro Segura (Gonzales)

1950
Phil. Indoor Pro Gonzales (Kramer)
Wembley Gonzales (Van Horn)
Paris Pro Indoors Segura (Kramer)
US Pro Segura (Kovacs)

1949
Scarborough Kramer (Budge)
Wembley Kramer (Riggs)
Barcelona Kramer (Segura)
Forest Hills Gonzales (Schroeder)

1948
Melbourne Pro Kramer (Pails)
US Claycourt Pro Kovacs (Lufler)
US Pro Kramer (Riggs)
Davis Cup Schroeder (Parker)

1947
US Indoor Pro Riggs (Budge)
Buffalo Pro Kovacs (Riggs)
US Pro Riggs (Budge)
RG Asboth (Sturgess)

1946
Southern Pro Budge (Riggs)
Miami Beach Van Horn (Kovacs)
US Pro Riggs (Budge)
US Pro Hard Court Riggs (Budge)

[SgtJohn]

1945
US Pro Claycourt Van Horn (Nogrady)
US Pro Hardcourt Riggs (Budge)

1943
US Pro Barnes (Nogrady)

1942
US Pro Budge (Riggs)

1941
US Pro Perry (Skeen)
Forest Hills RR Perry (Budge)

1940
US Pro Budge (Perry)
US Open Budge (Barnes)

1939
Southport Nusslein (Tilden)
French Pro Budge (Vines)
Wembley Budge (Nusslein)
US Pro Vines (Perry)

1938
Davis Cup Budge (Bromwich)
RG Budge (Menzel)
Wimbledon Budge (Austin)
Forest Hills Budge (Mako)

1937
Davis Cup Budge (Austin)
RG Henkel (Austin)
Wimbledon Budge (Von Cramm)
Forest Hills Budge (Von Cramm)

1936
Davis Cup Perry (Quist)
Roland Garros Von Cramm (Perry)
Wimbledon Perry (Von Cramm)
Forest Hills Perry (Budge)

1935

Davis Cup Perry (Austin)
Roland Garros Perry (Von Cramm)
Wimbledon Perry (Von Cramm)
Wembley Vines (Tilden)

1934
Davis Cup Austin (Perry)
Roland Garros Von Cramm (Crawford)
Wimbledon Perry (Crawford)
Wembley Vines (Tilden)

1933
Davis Cup Perry (Cochet)
Roland Garros Crawford (Cochet)
Wimbledon Crawford (Vines)
Forest Hills Perry (Crawford)

1932
Davis Cup Borotra (Vines)
RG Cochet (De Stefani)
Wimbledon Vines (Austin)
Forest Hills Vines (Cochet)

1931
Davis Cup Cochet (Austin)
RG Borotra (Boussus)
Wimbledon Wood (Shields)
Forest Hills Vines (Lott)

1930
Davis Cup Cochet (Tilden)
Roland Garros Cochet (Tilden)
Wimbledon Tilden (Allison)
Forest Hills Doeg (Shields)

1929
Davis Cup Cochet (Tilden)
Roland Garros Lacoste (Borotra)
Wimbledon Cochet (Borotra)
Forest Hills Tilden (Hunter)

1928
Davis Cup Cochet (Tilden)
Roland Garros Cochet (Lacoste)
Wimbledon Lacoste (Cochet)
Forest Hills Cochet (Hunter)

1927
Davis Cup Lacoste (Tilden)
Roland Garros Lacoste (Tilden)
Wimbledon Cochet (Borotra)
Forest Hills Lacoste (Tilden)

1926
Davis Cup Johnston (Lacoste)
Roland Garros Cochet (Lacoste)
Wimbledon Borotra (Kinsey)
Forest Hills Lacoste (Borotra)

1925
Davis Cup Johnston (Tilden)
Roland Garros Lacoste (Borotra)
Wimbledon Lacoste (Borotra)
Forest Hills Tilden (Johnston)

1924
Davis Cup Tilden (Richards)
Olympics (Paris) Richards (Cochet)
Wimbledon Borotra (Lacoste)
Forest Hills Tilden (Johnston)

1923
Davis Cup Tilden (Anderson)
HC World Champ’s Johnston (Washer)
Wimbledon Johnston (Hunter)
Forest Hills Tilden (Johnston)

1922
Davis Cup Johnston (Tilden)
HC World Champ’s Cochet (de Gomar)
Wimbledon Patterson (Lycett)
Forest Hills Tilden (Johnston)

1921
Davis Cup Johnston (Tilden)
HC World Champ’s Tilden (Washer)
Wimbledon Tilden (Norton)
Forest Hills Tilden (Johnston)

1920
Davis Cup Johnston (Tilden)
HC World Champ’s Laurentz (Gobert)
Wimbledon Tilden (Patterson)
Forest Hills Tilden (Johnston)

1919
Aus. vs USA match Tilden (Johnston)
US Clay Johnston (Tilden)
Wimbledon Patterson (Brookes)
Forest Hills Johnston (Tilden)

1918
US Clay Tilden (Garland)

1917
Newport Murray (Niles)

1916
Newport Williams (Johnston)

1915
Newport Johnston (McLoughlin)

1914
Davis Cup McLoughlin (Wilding)
HC World Champ’s Wilding (Salm)
Wimbledon Brookes (Wilding)
Newport Williams (McLoughlin)

1913
Davis Cup Parke (McLoughlin)
HC World Champ’s Wilding (Gobert)
Wimbledon Wilding (McLoughlin)
Newport McLoughlin (Williams)

1912
Davis Cup Parke (Brookes)
Monte-Carlo Wilding (Moore)
Wimbledon Wilding (Gore)
Newport McLoughlin (Johnson)

1911
Davis Cup Brookes (Heath)
Monte-Carlo Wilding (Decugis)
Wimbledon Wilding (Ropert Barrett)
Newport Larned (McLoughlin)

1910
Monte-Carlo Decugis (Ritchie)
Wimbledon Wilding (Gore)
Longwood Bowl Larned (McLoughlin)
Newport Larned (Bundy)

1909
Davis Cup Brookes (Wilding)
Monte-Carlo Alexander (L Doherty)
Wimbledon Gore (Ritchie)
Newport Larned (Clothier)

1908
Davis Cup Wright (Wilding)
Monte-Carlo Wilding (Eaves)
Wimbledon Gore (Ropert Barrett)
Newport Larned (Wright)

1907
Davis Cup Brookes (Gore)
Monte-Carlo Ritchie (L Doherty)
Wimbledon Brookes (Gore)
Newport Larned (LeRoy)

1906
Davis Cup L Doherty (S Smith)
Monte-Carlo L Doherty (Eaves)
Wimbledon L Doherty (Riseley)
Newport Clothier (Wright)

1905
Davis Cup L Doherty (S Smith)
Monte-Carlo L Doherty (Ritchie)
Wimbledon L Doherty (Brookes)
Newport Wright (Ward)

1904
Davis Cup L Doherty (R Doherty)
Monte-Carlo R Doherty (Ritchie)
Wimbledon L Doherty (Riseley)
Newport Ward (Clothier)

1903
Davis Cup L Doherty (R Doherty)
Monte-Carlo R Doherty (Riseley)
Wimbledon L Doherty (Riseley)
Newport L Doherty (Larned)

1902
Monte-Carlo R Doherty (Hillyard)
Irish L Doherty (S Smith)
Wimbledon L Doherty (A Gore)
Newport Larned (R Doherty)

1901
Champ’s of Europe Decugis
Irish R Doherty (L Doherty)
Wimbledon Gore (R Doherty)
Newport Larned (Wright)

1900
Paris Olympics L Doherty (Mahony)
Irish R Doherty (Gore)
Wimbledon R Doherty (S Smith)
Newport Whitman (Larned)

1899
Champ’s of Europe Mahony (R Doherty)
Irish R Doherty (Mahony)
Wimbledon R Doherty (Gore)
Newport Whitman (Paret)

1898
Irish Mahony (Eaves)
Northern L Doherty (Hobart)
Wimbledon R Doherty (L Doherty)
Newport Whitman (Davis)

1897
Irish Eaves (Baddeley)
Northern Baddeley (R Doherty)
Wimbledon R Doherty (Mahony)
Newport Wrenn (Eaves)

1896
Irish Baddeley (Mahony)
Northern Baddeley (Mahony)
Wimbledon Mahony (Baddeley)
Newport Wrenn (Hovey)

1895
Irish Pim (Eaves)
Northern Baddeley
Wimbledon Baddeley (Eaves)
W Newton Int’l RR Pim (Mahony)

1894
Irish Pim (Chaytor)
Northern Baddeley (Pim)
Wimbledon Pim (Baddeley)
Brit. Indoors Mahony

1893
Irish Pim (E Renshaw)
Northern Pim (Baddeley)
Wimbledon Pim (Baddeley)
Brit. Indoors Mahony (Meers)

1892
Irish E Renshaw (Lewis)
Northern Pim (Barlow)
Wimbledon Baddeley (Pim)
Brit. Indoors Meers (Lewis)

1891
Irish Lewis (Pim)
Northern Pim (Baddeley)
Wimbledon Baddeley (Pim)
Brit. Indoors Lewis

1890
Irish Lewis (Hamilton)
Northern Pim (Hamilton)
Wimbledon Hamilton (W Renshaw)
Brit. Indoors Lewis

1889
Irish Hamilton (E Renshaw)
Northern Hamilton (Mahony)
Wimbledon W Renshaw (E Renshaw)
Brit. Indoors Lewis (Barlow)

1888
Irish E Renshaw (Hamilton)
Northern Hamilton (Grove)
Wimbledon E Renshaw (Lawford)
Brit. Indoors Lewis (Meers)

1887
Irish E Renshaw (Lawford)
Northern Grove (Chaytor)
Wimbledon Lawford (E Renshaw)
Brit. Indoors Lewis (EL Williams)

1886
Irish Lawford (Hamilton)
Northern Grove (Dwight)
Wimbledon W Renshaw (Lawford)
Brit. Indoors EL Williams (Lawford)

1885
Irish Lawford (Chatterton)
Northern Dwight (Stewart)
Wimbledon W Renshaw (Lawford)
Brit. Indoors Lawford (Ross)

1884
Irish Lawford (E Renshaw)
Northern Stewart (Wilberforce)
Wimbledon W Renshaw (Lawford)
Scottish Gamble (Horn)

1883
Irish E Renshaw (W Renshaw)
Northern Wilberforce (Richardson)
Wimbledon W Renshaw (E Renshaw)
Prince’s Champ. Lawford (E Renshaw)

1882
Irish W Renshaw (Richardson)
Northern Richardson (E Renshaw)
Wimbledon W Renshaw (E Renshaw)
Prince’s Champ. E Renshaw (Lubbock)

1881
Irish W Renshaw (Lawford)
Northern Richardson (Browne)
Wimbledon W Renshaw (Hartley)
Prince’s Champ. W Renshaw (Mulholland)

1880
Irish W Renshaw (Goold)
Northern Richardson (Fairlie)
Wimbledon Hartley (Lawford)
Prince’s Champ. Lawford (Lubbock)

1879
Irish Goold (Barry)
Wimbledon Hartley (Goold)
Cheltenham W Renshaw (Goold)

1878
Wimbledon Hadow (S Gore)

1877
Wimbledon S Gore (Marshall)

[Nickognito]

great! Thank you very much

c.

[Moose Malloy]

John, thanks so much for posting that. I've been wondering about it since you mentioned it many threads ago. What is the leader board for the 'majors' now? Also do you have a count on 'masters series' like events? I would love to see scores on some of these matches.

1972
WCT Finals Rosewall (Laver)
RG Gimeno (Proisy)
PSW Los Angeles Smith (Tanner)
USO Nastase (Ashe)

I see you made some changes from the previous thread.

http://72.29.176.21/showthread.php?t=135592

imo, W had a much better field that year, & the final(Smith-Nastase) is widely considered one of the best matches of all-time.

[SgtJohn]

Originally Posted by Moose Malloy

John, thanks so much for posting that. I've been wondering about it since you mentioned it many threads ago. What is the leader board for the 'majors' now? Also do you have a count on 'masters series' like events? I would love to see scores on some of these matches.



I see you made some changes from the previous thread.

http://72.29.176.21/showthread.php?t=135592

imo, W had a much better field that year, & the final(Smith-Nastase) is widely considered one of the best matches of all-time.

Hi Moose,

Thanks for the kind words!

1972 was a difficult year to judge (as I said, it's just my current list, I can still change my mind, and people can have their own take on the question). The first thing is, for this new list I choose to select a clay event every year. '72 was a very poor year 'clay-wise', so that the best tournament of the year was the quite depleted Roland Garros. That leaves us with 3 fast-court tournaments to choose...
Secondly, 1972 was another year when the WCT pros (Laver, Rosewall, Newcombe, Roche, Ashe, etc.) and the others (Smith, Nastase, Orantes, Gimeno, etc.) were almost all the time separated. Wimbledon was a victim of this: except Smith ('71 runner-up),*none* of the winners or runner-up from 1968-1971 played. We can't say it had a great field then, and it was a non-WCT events (no WCTers admitted).
The question is: were the fields of the US Open, PSW and WCT Finals better?
-The US Open had the only great field in this year, period.
-The PSW had a good, mixed (WCT players and amateurs) field, but some of the greats were not there (Laver & Newcombe).
-This is why I wanted to select the Dallas WCT, because otherwise there would've been only one event featuring Laver and Newcombe, which seems unfair to their level.

The last debate is between the PSW and Wimbledon. Though Los Angeles had a better field (Ashe and Rosewall were there), on prestige alone, I admit that Wimbledon could easily replace it...
I agree the Smith-Nastase final is considered as one of the greatest, but I try to think more of how significant the tournaments seemed a priori, before actual play takes place. A slam is not a slam because it features great matches, but because it involves great players who come and play due to the reputation of the tournament. Wimbledon '02 or Roland Garros '03 were very boring events, but they are still slams...

* * *

The current leader-board: (10+ only)

Rosewall 21
Gonzales 19
Laver 18
L Doherty 15
Tilden 15
Sampras 14
Budge 13
Borg 13
W Renshaw 12
Cochet 12
Federer* 12
Perry 11
Lendl 11
Pim 10
R Doherty 10
Johnston 10
Connors 10


About a similar 'Masters-Series' list, I have one as well, except for diverse reasons, I worked on a 'Super-14' rather than 'Super-9' basis, which inflates the results of current players too. I think this list is informative, but not nearly as useful as the Majors list, because the best often did not have enough opportunity to play them. Today, Federer plays in a dozen valuable events apart the Slams. In the late pro years ('64-'67), the pro tour had much developed, and Laver had many opportunities too (10 out of 14 events I selected for these years are pro events). But, for instance, for the 50s, the 40s, etc. I could not select more than a handful of pro events,sometimes none...The rest were amateur events.
So of course the list is biased towards the players who *could* play in events. I think such bias almost doesn't exist with the Majors list.

(More clearly maybe with an example: in 1954-1955, Gonzales was really dominant. In my major list, he won 5 majors during these 2 years, which seems coherent with modern dominance standards (Federer, etc.). During the same amount of time, he won 1 'masters-series' events, even if he was almost undefeated in 1955. That seems incoherent with our standards, if Gonzales had played in 14 'masters-series', he would have won a lot, lot more...)

Anyway, here is the 'Masters-Series' leaderboard for those interested (but don't forget the 'disclaimer' above! ). If some are interested, I could send more details or a whole list...

Laver 42
Tilden 38
Rosewall 36
McEnroe 35
Lendl 33
Ritchie 26
Connors 26
Gonzales 25
Wilding 25
Becker 24
L Doherty 24
Federer 22
Borotra 21
Agassi 21
Emerson 21
Borg 21
Sampras 21


Good night!
Jonathan

[FedForGOAT]

Tim,

I really liked your explanation for why domination over a time period of a few years is the most meaningful. I was wondering however what you'd have to say about McEnroe. he completely dominated in 84', but not as much in other years. how would you rank him using the DOt system? how would you, subjectively, based on pure opinion, rate him? I think that based on pure potential, he is one of the top, if not the verry best. but he underachieved, IMO, and that hurt his rating. How would you rate him on the list of all-time greats?

[Nickognito]

without the doubles, I suppose

c.

[Steve132]

Originally Posted by Wuornos

First of all I would like to thank you for your kind words Cristiano.

I'm not sure what data you would like to see from me in relation to DOT or Elo ratings, but if you let me know I'll see what I can do.

The article on Bayesian Elo is very interesting. One of the main assumptions of the ELO system, and it is a flawed assumption, is that the standard deviation in the level of performance from match to match is 200 points and this is taken for all players. Of course this is not true and some players may have a SD of 100 and others 300. I think the Bayesan system goes some way to rectifying this. I have no doubts that this system is superior, but I have two concerns that that are not really concerned with the accuracy.

1. I have worked as a statistician for many years and one thing that occurs over and over again and where the layman often loses his way, is the question of fitness for purpose and gold plating. By this I mean much statistical modelling can be subjected to the law of diminishing return. You can add in an additional attribute to an analysis carry out all the work to incorporate this element and the impact on the output is minimal. You often find you can get to the 80% accuracy mark by using 20% of the data available and with a relatively simplified system. I would therefore worry that the Bayesian System may yield marginally better results but at what cost?

2. While being independent and free from human opinion and iterference both ELO and Bayesian Rating systems struggle with inactivity. What do you do when a player becomes inactive. E.g. Amelie Mauresmo at the moment. How long do you include them in the rankings? Even worse, what do yoiu do when they return after a year out like Lindsay Davenport. Do you allow her to keep the same rating, surely her quality of play will have declined during the inactivity etc.

You are absolutely correct that domination varies with time period. Single match domination is not the same as longer term. The ability to hand a double or triple doughnut to another player is talking about a completely different subject to being able to dominate by winning against a large number of opponents over a set period of time.

For me the sample of a single match is to small and rather meaningless. I seem to recall someone posting on this site previously regarding a perfect match, i.e 6-0 6-0 and all 12 games being love game. For me this probably says more about the quality of opposition than it does about the quality of the winner.

Single year dominance is an altogether different matter. For me this period is again to small to be completely meaningful. If you just forget about the number of matches and how many wins and just think about the number of losses, the numbers is very small when considering top players within a single year. When comparing two players you can be talking about one who lost 4 matches in a year compared with one who lost 5 and this diffference is enough to effect the overall strike rate and consider one player superior.

A few years. This for me is the most significant factor. It is enough evidence for luck to have been largely eliminated while not being so long as it favours players with only long careers.

Domination over more than 10 years. I don't really call this domination. I tend to refer to it as career achievement. I don't like this as it undervalues players who have had a short career at the top. E.g Monica Seles. However players with along career have had the opportunity to dominate for shorter periods. Failure to do this means to me that they have had a long time nigglking away at the a level just short of the top of the game. Just my view.

For me the best indicator of standard of play or greatness is always going to be dominance over a few years with adjustments for quality of opposition and Tournamentvent. Hence DOT Ratings.

Take care and keep posting.

Tim

Tim: Excellent post of usual.

In general when assessing dominance I tend to follow the baseball sabermetrician Bill James, who basically says that you can assess greatness in two ways - by looking at a player's peak performance and by examining his or her career achievement. Peak performance refers to the player's best few years. For tennis players I think this would be a period of about 3 consecutive years. One year is not enough, since a player can have a season in which everything goes right. Five years would probably be a bit too long, since in the Open era to date no male player has been great (as opposed to very good) for five consecutive years. Federer might become the exception if he duplicates his results of the past four years in 2008.

Career achievement includes everything that a player accomplishes during his or her career. Many rating schemes focus solely on career achievement, since they simply count the number of Slams and year-end No. 1 rankings that a player accumulated during his or her career. This information is vital, because if player A is great for 10 years and B maintains the SAME level of excellence for three years we would want to say that A is a greater player than B. It's not enough, however, because a significant element in evaluating greatness is determining how dominant the player was at his or her very best.

James does not have any simple formula for combining peak performance and career achievement. I don't really have one to suggest either, but I would maintain that a player assessment formula has to take both features into consideration.

Many fans like to ask whether player A could beat player B if both were at their best. This question is much more complex that it initially appears, and can probably only be answered satisfactorily for close contemporaries. For one thing, you would need to choose the surface. For another, you would have to consider matchup effects, e.g Nadal is a bad matchup for Federer but a good one for Blake, who holds a 3-0 head to head advantage against Nadal but has never beaten Federer.

You would probably need to stage three round robin competitions - say, at Roland Garros, Wimbledon and the U.S Open - between contemporaries in order to address these complications. Even then it would not make much sense to ask whether Federer at his best would beat Laver at his best. The game has simply changed too much over the past 40 years, and all we can do is compare their achievements.

To sum up: this post probably contains more questions than answers, but I hope that these questions are at least some of those to be addressed in order to complete player evaluations.

[CyBorg]

Great stuff, Sgt. John. I'm going to save that list in notepad.

Great attention to detail. I can see you've put some time and research into this (no Wimbledon in '73, for example).

Anyway, here is the 'Masters-Series' leaderboard for those interested (but don't forget the 'disclaimer' above! ). If some are interested, I could send more details or a whole list...

Please...

[urban]

Great List, Sgt John. There are some minor questions and considerations about the surface status. I think, you have tried to integrate clay court events in the rating. In 1963, the Rome event was most probably an indoor event, because it was played in late September/ begin October 1963 after the French Pro, which was also played indoors. The best and most significant pro clay event in 1963 was imo Kitzbühel or Nordwijk played in July and August that year. In 1965, the US pro indoor was imo the most prominent pro event, next to the pro majors, and in 1970 the Barcelona event, played in october in the middle of the indoor season, was less significant than the South west Pacific event at LA, or for a clay court event, the Louisville event in August, which had a better field.
I will remark some other disputable things in the future. Its not a general critique, but i put on some considerations in a very difficult data field, which i know myself too well. I appreciate the manner of research very much.

[SgtJohn]

Hi urban,
Thank you very much for your remarks!

-I can't say I'm 100% sure about the Italian Pro... Apparently from what I've read in books or from discussions with other people, most of European continental tennis was on clay until the 70s, with some notable exceptions such as Paris indoors arena or Cannes' Palais des Sports. It's pretty sunny in Rome at the beginning of fall, and there is already a very famous tennis facility so I assumed it was played on clay.
Plus, it was the first year the French Pro was played indoors. That's only guessing, but I thought the pros might have seen the sunnier Rome as a replacement for Paris as a great fall clay event... But again, I could be wrong. That's one of the weak points of my list, and I just have to keep looking for more documents...I will ask the question on wikipedia, as there are knowledgeable people on the forum out there who seem to own many magazines and books from this era.
I asked the same question about Geneva, but they couldn't confirm me anything. It was a great tournament, and I thought about it for my clay event, but I know there is a big sports indoor arena in Geneva (and it rains more than in Rome ), so I thought it was too risky to assume it was on a slow surface.

-the US Pro indor was indeed a huge event in 1965, just as in 1964 by the way, but I had to select a clay event apart from the 3 pro majors (you see this clay business is quite a strong constraint as it forces us to select less significant tournaments only based on their surfaces. That's a choice I made but my former list without this constraint can also be considered, for people who don't think too much of surface as a decisive factor...)
City of Milan Pro is one of the only tournaments of the year that I could assume was on clay (for roughly the same reasons as Rome). It's a small 4-men event, but Gimeno won it by beating Laver and Rosewall, so though I don't like these a posteriori reasonings, I think Gimeno 'earned' this Major.

-1970 was a year when 'Open Tennis' didn't seem like much of a reality anymore. No WCT Pro took part in Roland Garros. Sure the Louisville event featured Newcombe, one of the great player that year, as well as former RG winners Roche, Emerson and Stolle, who were not in Barcelona.
But:
-Barcelona was a 'mixed' event with WCT Pros as well as independents and amateurs. There were the last three years' Roland Garros winners and runner-up: Laver, Rosewall, Kodes and Franulovic, as well as Gimeno and Orantes, other great clay-courters. It would seem unfair to select an event featuring only players banned from RG, instead of one that gave those who did well in Paris a chance to prove themselves against the very best.
-Barcelona was a best-of-five, 64-players draw, as opposed to Louisville, best-of-three, 16-men event...Barcelone really seemed like the 'unoffical clay major' that year, just as Sydney was the unofficial Australian Open.


Thanks, and don't hesitate to make other remarks...
Jonathan