Some GOAT questions for chaog and others

urban said:

I think, John your analysis makes itself a bit too easy. Rosewall didn't win everything, maybe due to his long absence, but anyway. There are valid arguments, that Gonzales wasn't the Nr. 1 player on the pro tournament series in 1958 and 1959, when Segdman and Hoad respectively had probably the edge, with betetter head to head and more important title wins. And as i said, for the pre open era, you have to evalute all parts of a career, both the amateur and pro stages. By completely dismissing all amateur results, you cannot get a clear picture. And Federer still has to win a French, to be thoroughly dominant. His 3 majors a year is nice and good, but on the other side, he did miss a Grand Slam in every year.

Click to expand...

Your analysis seems to point to Laver as the GOAT. EVERY other player has at least one gap in his resume. Maybe Gonzalez or Rosewall could have won a career Slam if they had been allowed to compete in the 50's or early 60's. Maybe Tilden would have beaten the French Musketeers on clay if he had had the opportunity. I don't think that we can assume these things, however.

I'm reasonably comfortable with the selection of Laver as the GOAT. The difficult part of the assessment would be to rank the next 4-5 players after him.

Urban, your points about Tilden and Gonzales are most helpful. They make me wonder whether anyone has ever been thoroughly dominant by these exacting standards. I guess Tilden in 1921 would qualify. Laver in 1968 would not. As for Laver in 1967 and 1969, or Rosewall in 1962-1963, I don't know. They seem to have won all the biggest events on all the different surfaces, but also to have lost many matches (Laver lost more matches in 1969 than Federer did in 2004-2006 combined).

Steve132, I don't think that these facts lead inexorably to the conclusion that Laver is the GOAT. Even if we count it against Tilden and Gonzales that they didn't sufficiently prove themselves on clay, that would have to be weighed against the fact that they were at or near the top and winning important tournaments for far longer than was Laver, and that Tilden lost much less often than did Laver. Moreover, if we're going to say that Tilden and Gonzales have decisive holes in their resumes due to circumstances outside their control (if indeed these circumstances were outside their control), then what we're really saying is that no one before 1960 can be considered for GOAT because the opportunities to achieve it didn't exist before 1960. If that's the case, then we might as well just admit that we can't compare the pre-1960 players with the post-1960 players.

And if it's the absence of holes that matters, what is the hole in Rosewall's resume? He seems to have been as dominant and versatile as Laver. Failing to win Wimbledon as an amateur, when neither he nor his opponents at the tournament were the world's best players, isn't a hole in my book.

I'd put Laver (barely) ahead of Rosewall due to Laver's longer tenure at #1, which he achieved via direct head-to-head competition with Rosewall. But I don't know how to compare Laver to the best from the early 1920s, the mid-1950s, or the mid-2000s.

urban said:

I think, John your analysis makes itself a bit too easy. Rosewall didn't win everything, maybe due to his long absence, but anyway. There are valid arguments, that Gonzales wasn't the Nr. 1 player on the pro tournament series in 1958 and 1959, when Segdman and Hoad respectively had probably the edge, with better head to head and more important title wins. And as i said, for the pre open era, you have to evalute all parts of a career, both the amateur and pro stages. By completely dismissing all amateur results, you cannot get a clear picture. And Federer still has to win a French, to be thoroughly dominant. His 3 majors a year is nice and good, but on the other side, he did miss a Grand Slam in every year.

Click to expand...

Hi urban,

I was just wondering about your views concerning 1959... How do you come to the idea of Hoad being #1?? I don't currently have an opinion on this, I'm still going through McCauley's book, so I'm just curious about it...
Thanks,
Jonathan

John123 said:

Urban, your points about Tilden and Gonzales are most helpful. They make me wonder whether anyone has ever been thoroughly dominant by these exacting standards. I guess Tilden in 1921 would qualify. Laver in 1968 would not. As for Laver in 1967 and 1969, or Rosewall in 1962-1963, I don't know. They seem to have won all the biggest events on all the different surfaces, but also to have lost many matches (Laver lost more matches in 1969 than Federer did in 2004-2006 combined).

Click to expand...

Well, in seeking out the "most dominant" players we certainly have to look to the pre-World War I era. Willie Renshaw was undefeated for many years in a row; however, in some of those years he only played one match. Laurie Doherty, in the early 1900s, had some undefeated seasons (or sometimes with only a single loss), and he played a much fuller tournament schedule. Of course, tennis was not the same competitive, international event that it would later become, which needs to be taken into account, but these guys were still great players with some formidable rivals, and in the context of their own times no player since has been more dominant than either Renshaw or Doherty.

On Segdman and Hoad in 1958 and 1959, there are some valid comments and rankings (not by me), which argue in favor of both. If you put 4 major pro events in 1958 with all top players competing: Sedman won Wembley and the Austalian pro, Gonzales the Forest Hills pro round robin, and Rosewall the French pro. (The US pro hadn't a very good field). Sedgman had also a 4-2 head to head in important events over Gonzales. In 1959, there were several lists and experts, who put Hoad on top. Hoad was the 'race-leader' in a 14 pro tournament series over the whole world, and he had beaten Gonzales in the important Forest Hills final. In the eyes of most pros, Gonzales however remained Nr.1, for his head-to-head series over Hoad in 1958. In the mentality of pros then, it was like the old boxing system, when a champion had to be dethroned in a title match.
It shows the need to study all these years carefully,to come to differenciate results and to contextualize these results. I am not in a race here, to speculate over goats, Laver certainly doesn't need my help in this question, nor i want to penalize someone for his failings. Who i am, to try such a thing? But for dominant performance, i would rank a Grand Slam indeed as pretty dominant. Percentages are only half the truth. If you ask McEnroe or Federer, to exchange their high percentages in 1984 or 2005 for the missing links and lost important matches (RG and Davis Cup for Mac, Australia and RG for Federer), what would they answer? Mac is still dreaming about these losses.

urban said:

But for dominant performance, i would rank a Grand Slam indeed as pretty dominant.

Click to expand...

Urban, everything you wrote makes a lot of sense. Out of curiosity, how do you feel about Rosewall in 1962-1963?

I've now concluded that Connors is the GOAT. I just watched the YouTube of him beating Mac at Wimbledon.

I don't see how anyone could win anything with that godawful T-2000 racquet. Much less Wimbledon, using it to beat Mac with a Dunlop Maxply. The T-2000 has got to be the worst racquet of all time.

Give Federer one and he would quit after one game.

John asked me about Rosewall. I have written some posts about Rosewall's style and position in the game, which i very admire. Agreed, Rosewall is always a bit underrated, as a sort of Raymond Poulidor of tennis. His longevity was phenomenal, no other player (including Gonzales and Tilden) had that great successes over that long period of time. As an amateur, he was never the Nr.1, always a bit behind Trabert and Hoad. And in his own mind, the Wimbledon losses to Nielsen, who was his nemesis, going for the middleline volley, Drobny, Hoad and later Newcombe hurt him quite severely. He later critizised Hopman, to give him the false defensive tactic against old lefty Drobny. On the pro front, after his rough baptism against Gonzales in 1957, he was in the group of top 3-4 pros, before taking over the mantle of pro king in 1961. In 1962 he was dominant as a pro, but it was a bit of a transition year on the pro tour: Gonzales was not playing, Hoad suffering from a bad back and/or a loss of form, Sedgman,Trabert and Segura aging, and Gimeno still learning his trade. I still think, that the amateurs in 1962, who had regrouped in the early 60s, with Laver, Emerson, Santana, and behind them a group of Fraser, Osuna, Krishnan and others, in their depth were not much behind the pros that year. In 1963, Rosewall dominated the new pro Laver in his baptism year, at least in the first half of the year. In the second half, it was much closer, with a head to head of 4-3. Rosewall won the big prizes of the pros, Wembley and US pro, and the French pro indoors in a great five setter over Laver, but on European clay on the summer tour, Laver was the Nr.1 pro, winning the big clay events at Kitzbühel and Scheveningen, with Gimeno winning at Geneva. Since the middle 60s he settled for Nr.2, but he was always a factor at the top. In 1970, with a strange, unclear situation at the top, he has a valid claim for Nr.1. All in all, Rosewall was the most excellent over a long period of time, but undoubtedly Nr. 1 for 2-3 years. His crafty, effortless, classic style made him steady like no other, but he could be outplayed by players with more power and improvising talent.

Regarding Rosewall, and also the question of "true" majors: I sometimes think it is easiest and fairest to highlight the three biggest events of each era, rather than four. This is because from the early 1900s to about 1938, clearly the most important events were Wimbledon, the US Championships, and the Davis Cup.* Once Budge toured as a pro against Vines and Perry in 1939, the pro game could finally claim most of the top players, and so from 1939-67 the three pro majors were probably the most consistently significant contests: Wembley, the US Pro, and French Pro. (There were other important pro events on a year-to-year basis, but these three were easily the most "traditional.") Finally, in the Open Era, allowing for the complications of the early 1970s, I think it can be reasonably asserted that Wimbledon, the US Open, and the French Open have been the three biggest events on a more-or-less consistent basis. The Australian is now perceived by many as being on par with the others, but this was not the case for many years in the 1970s and early '80s, and even in the late 1980s players valued it a bit beneath the other three. (Here I would cite a players' poll that ranked the majors as follows: 1. Wimbledon, 2. French Open, 3. US Open, 4. Australian Open.)

So, allowing for these rough chronological divisions and tallying up each player's "big three" titles, what do we find? Perhaps not surprisingly, Rosewall is No. 1, with only Tilden close behind. This makes sense, as Rosewall had an exceptionally long career and the ability to win on all surfaces, giving him more good opportunities than most to amass big titles. Here is how the complete top ten would look given this methodology:

1. Rosewall (17)
2. Tilden (16)
3. Gonzales (12)
-. Laver
-. Sampras
6. Borg (11)
-. Perry
8. Budge (10)
-. L. Doherty
10. Federer (9)

Though I usually argue for Tilden and Laver, rather than Rosewall, as the co-GOATs, to a certain extent I must bow to this methodology. The top ten players it produces are exactly the same players I would choose in a more subjective sense; it also complements SgtJohn's research (which is based on a more fluid four-tournament approach), though he puts Sampras a bit lower and Laurie Doherty a bit higher. There are other factors that play into a player's greatness, of course, but if you are looking for a particularly efficient and easily defensible way of ranking players, then I think this one is hard to beat. And as you can see, Federer is only about halfway to reaching the totals of the top two!

*I should be clear about what constitutes giving a player credit for "winning" the Davis Cup: in my mind, this means winning both Challenge Round singles matches for a victorious team. Thus Tilden gets credit for 1920-25, but not '26, when he suffered a loss to Lacoste and it was the other Bill--Little Bill Johnston--who effectively won the cup for America.

John123 said:

B. Rosewall won everything on every surface and was thoroughly dominant for 2-plus years, and he maintained the ability to win majors and stay close to the top for an eternity.

Click to expand...

As much as I admire Rosewall, I don't think this is entirely true. Muscles never won Wimbledon, although he was an excellent grass-court player, getting to the AELTAC finals four times: 1954, 1956, 1970, 1974.

Newbies should note the first and last dates--20 years apart, surely a record there! How good was Jaroslav Drobný in 1974? (I wonder if Federer will be in the Wimbledon finals in 2023?)

chaognosis said:

1. Rosewall (17)
2. Tilden (16)
3. Gonzales (12)
-. Laver
-. Sampras
6. Borg (11)
7. Budge (10)
-. L. Doherty
-. Perry
10. Federer (9)

Though I usually argue for Tilden and Laver, rather than Rosewall, as the co-GOATs, to a certain extent I must bow to this methodology. The top ten players it produces are exactly the same players I would choose in a more subjective sense; it also complements SgtJohn's research (which is based on a more fluid four-tournament approach), though he puts Sampras a bit lower and Laurie Doherty a bit higher. There are other factors that play into a player's greatness, of course, but if you are looking for a particularly efficient and easily defensible way of ranking players, then I think this one is hard to beat. And as you can see, Federer is only about halfway to reaching the totals of the top two!

*I should be clear about what constitutes giving a player credit for "winning" the Davis Cup: in my mind, this means winning both Challenge Round singles matches for a victorious team. Thus Tilden gets credit for 1920-25, but not '26, when he suffered a loss to Lacoste and it was the other Bill--Little Bill Johnston--who effectively won the cup for America.

Click to expand...

Hi chaog, and thank you for your contribution!

This is an interesting approach, mainly through the fact that it lets us avoid all the "4th event" debates. Still, I feel that the 4-Majors-a-year tradition is constitutive of the game of tennis, and would have a hard time abandoning it...(plus you can make some players very unhappy with your choice of events...think Agassi! ).
I think selecting the "Pro Slams" as the true majors is quite reliable, but not totally: the US Pro was often very depleted, and I don't select any of these tournaments in my list from 1956 to 1962.

Actually, I've recently been reshaping my old list, mainly thanks to the McCauley book I just acquired. In this new list I chose to do something I had been aiming at for a long time: including a clay or slowcourt event every year, acknowledging the significance of this surface in the game, and letting us compare Federer and Sampras fairly with the past players... So you can see how your 3-Major approach isn't compatible with this, as it can't include RG for all the pre-WW2 era...

Here is the Majors tally that I obtained:

Rosewall 21
Gonzales 20
Laver 19
L Doherty 16
Tilden 15*
Sampras 14
Budge, Borg 13
W Renshaw, Cochet, Perry, Federer 12
Lendl 11
Johnston, Connors 10
Lewis, Pim 9
Larned, Lacoste, Agassi 8


*You might be surprised about Tilden's result...It comes from the fact that I re-examined in more details my attribution of Davis Cup wins, and that I decided that only one player would benefit from such a win each year. This may seem unfair at times, but giving this honor to both players of the winning team would lead to 5-majors years, wich is dissatisfying too...

In all of their DC victories, except in 1923 and 1926, "Big and Little Bill" went undefeated. If you have a close look at their respective records, you can note that Johnston often did better than Tilden. I went through the NY Times archive and found that in the articles for these Davis Cup finales, the writers remarked that, though he was the national Champion, Tilden's game has been more uneven than Johnston's. For instance in 1921 (Tilden's great year), Johnston easily beat Kumagae and Shimidzu (top 10 players at the time) in 3 short sets, and Tilden was extended by both of them, almost suffering a defeat in one of these matches...it did not go unnoticed, and it seems that the Americans considered Johnston as the true 'winner' of the David Cup.
Consequently I decided to attribute the 1920, 21, 22, 25 and 26 Davis Cup victories to Johnston, who barely lost a set during all these years. 1923 (Johnston beaten by Anderson) and 1924 (Richards is Tilden's teammate, and Big Bill is very dominant this time anyway) go to Tilden.


Jonathan

Fascinating as always, Jonathan. I would recommend the three-tournament approach only as a sort of shorthand, a substitute for those who don't have the time, knowledge, or available resources to do your sort of comprehensive historical analysis. While our results diverge a bit more now, they are still remarkably close when you think about it. Roughly the same groups emerges on top, and in both cases Rosewall is at No. 1. I am still not satisfied without taking into account some sort of dominance factor, but I've yet to find a system that works really well in that regard. (Perhaps you have some thoughts? I have relied on expert rankings in the past, though it is perhaps dubious to mix statistics with subjective observation.) Again, I presented a sort of shorthand approach in an update to my earlier post, but in no means do I consider it definitive... though it has made me reevaluate Perry's career!

Chaog, this is fascinating. But you don't explain how you came up with the second (updated) list -- i.e., how you placed total career majors on the same scale with percentage of majors won in order to create a ranking that combines them. What weight did you give to each factor? Depending on the weighting, could your formula actually penalize someone for winning a major (e.g., Federer at Wimbledon in 2003) by lowering the player's percentage of major victories per year?

I just educated myself a bit about Perry's career as a result of your list. It looks like he was incredibly impressive in the mid-1930s, but that your metric creates a distortion in his favor in two ways -- both of which I'm sure you already know. First, as with Tilden (and as noted by Jonathan), it favors Perry because none of the three majors you include were played on clay. And second, it doesn't take into account the fact that Perry won those majors without playing Vines, who had banished himself to the pros. So Perry ends up as the best player of all time on that list, even though it's at least possible that he wasn't even the best player during his own prime years.

Still, the list is very, very interesting, and I think it should be used as a piece of the puzzle in identifying the GOAT.

chaognosis said:

Fascinating as always, Jonathan. I would recommend the three-tournament approach only as a sort of shorthand, a substitute for those who don't have the time, knowledge, or available resources to do your sort of comprehensive historical analysis. While our results diverge a bit more now, they are still remarkably close when you think about it. Roughly the same groups emerges on top, and in both cases Rosewall is at No. 1. I am still not satisfied without taking into account some sort of dominance factor, but I've yet to find a system that works really well in that regard. (Perhaps you have some thoughts? I have relied on expert rankings in the past, though it is perhaps dubious to mix statistics with subjective observation.) Again, I presented a sort of shorthand approach in an update to my earlier post, but in no means do I consider it definitive... though it has made me reevaluate Perry's career!

Click to expand...

It is right that our lists are similar, though the method is different. Actually, from the first list I had made a year ago to this one, including all of the intermediary steps, Rosewall and Gonzales have always been the two leaders, which clearly means something. Of course, the choice of including clay tournaments downgraded Pancho to #2 and helped Rosewall to make it to #1.

I've been thinking about the dominance factor for a long time, and it's not an easy question, to say the least. Your rescaled rankings are a good start, but I think their problem is that they rely too much on our choice of events...
An example: I could easily decide that, in spite of the war, the US Pro '41 was still a major event for this year (probably the only one). Then Perry's dominance factor is not anymore ((1936-1933+1)/8 )=2, but (1941-1933+1)/9=1 ! By simply awarding Fred with an additional event, I just crushed his leadership in terms of 'rescaled' events.
You see what I mean: the problem with this formula is that a player is actually hurt in the rankings if he wins a major late in his career, though it should be considered as an achievement. Borg is actually the ultimate player for this system: he retired after winning his last slam. Now imagine he managed to win RG during his 90's comeback. He would've had a terrible 'rescaled' ranking then. But the point of the rankings is to judge dominance, and this extra win did not make his dominance in the 70s any less significant...

So I think we have to come up with an evolution of your system. The problem we identified is caused by the 'extra' events that a player wins after his prime, and that don't affect the dominance of his youth. What I propose is to consider only the years when a player was #1 (after all, that is what dominance is about), and to calculate 2 factors: the first one is similar to yours: (# of majors won during the #1 years)/(# of such years), I would call it 'efficiency' factor. The second one is just the number of majors won during the years at #1, it's more of a true 'dominance' factor, as it counts the big tournaments a player has been able to win while dominating his peers.

Here are the corresponding lists (computed with my own list of majors and #1, so of course fairly subjective):

Dominance factor (years at #1)

Laver 18 (1964-1970)
L Doherty 15 (1902-1906)
Tilden 11 (1920-1925)
Gonzales 11 (1952, 1954-1957)
Rosewall 11 (1960-1963)
Federer 11 (2004-2007)**
Sampras 10 (1993-1998 )
W Renshaw 9 (1881-1886)
Budge 9 (1937-1939)*
Borg 9 (1978-1980)
Cochet 8 (1928-1931)
Perry 8 (1934-1936)
R Doherty 7 (1897-1901)
Pim 6 (1893-1895)
Kramer 6 (1948-1951)
Connors 6 (1974, 76, 82)
Lendl 6 (1985-1987)

*I decided not to award #1 rankings for the war years, which of course hurts Budge in these rankings, as he was probably the best player in '40 and '42.

**And counting...


'Efficiency' factor (Slams per #1 year)

Doherty 3
Budge 3
Wilander 3
Rosewall 2,75
Federer 2,75
Perry 2,7
Laver 2,7
Borg 2,25
Gonzales 2,2

(...basically all the rest is between 1 and 2, here are some notable results
Tilden 1,8
Sampras 1,7
W Renshaw 1,5)


These lists are probably not the perfect tool to judge dominance, but I hope they will help, and encourage people to imagine other formulae maybe... I don't have time right now to comment them, I'll send another post later...

Have a nice day everybody,
Jonathan

Of course, one can make a list of 3 or 4, and maybe of the two 2 most important events, or for the one most important event each year, calling that a World Championship. But in all that lists, the unsolved problem is the amateur/pro split. If one ignores the amateur achievements of the pre open players, one virtually nullifies the careers of a Cochet, Perry, Hoad, Emerson. In that scheme, Hoad has won absolutely nothing, less than Costa or Gaudio, is not worthy of a place in the hall of fame.And yet, some experts rate him as the best, regarding his Wimbledon wins and Davis Cup victories.
I think, Raymond Lee in his recent article on tennis week has many of these elements noted above: Absolute numbers of majors and titles, percentages of majors played and won, head to head percentages for career and best 5 year period and so on. One should maybe add categories for one year domination (absolute numbers and percentages), for year end Nr. 1 duration, and to add a longevity parameter, years of consistent ranking in the top ten.

urban said:

Of course, one can make a list of 3 or 4, and maybe of the two 2 most important events, or for the one most important event each year, calling that a World Championship. But in all that lists, the unsolved problem is the amateur/pro split. If one ignores the amateur achievements of the pre open players, one virtually nullifies the careers of a Cochet, Perry, Hoad, Emerson. In that scheme, Hoad has won absolutely nothing, less than Costa or Gaudio, is not worthy of a place in the hall of fame.And yet, some experts rate him as the best, regarding his Wimbledon wins and Davis Cup victories.

Click to expand...

Well, the amateur/pro split is not an unsolved problem, it seems...The point of such a list is to choose the 4 biggest events each year. There is of course a lot of subjectivity in such an approach, but the choice of events for every year reflects the relative levels of the amateur and pro players.
The first pro events I include in my list are for the year 1939, so my count does not nullify the careers of Cochet or Perry at all, but it kind of hurts Vines's career. I made an 'ideological' choice that tennis is a tournament sport, and thus never counted one of these big 2-men pro tours, that Vines took part to, as 'Majors'... There were pro tournaments before 1939, but they usually didn't involve the best players (Nusslein single-handedly dominated the tournament circuit for some years), and generally speaking, the pro field with Vines, Nusslein and the aging Tilden couldn't be compared to the amateurs Perry, Budge, Von Cramm, Austin, etc..
After the war until 1968, most of my 'Majors' are pro events, but *not* every one of them. In my opinion, only some amateurs had seasons when their level of play could be compared to the pros'. That would be: Kramer '47, Gonzales '49, Sedgman '52, Hoad '56. Their great amateur year is not nullified, as I award them with one major for that year. (Other great years include Laver '62 and Emerson '64, but the depth and number of tournaments there were in the pro circuits these years prevented me to award the amateur tour with a Major).
Do you think it is unfair and 'nullifying' their career to not give them more? Let's take the example of Hoad. After all, he only won 4 amateur majors. I personally think that his 1956 level was on a par with the best pros. He dominated Rosewall during most of this year, and Rosewall went on to score some big wins against Gonzales in 1957...
To acknowledge this, and also due to the fact that there were few pro tournaments in 1956, I include Wimbledon 1956 in my list of true majors. Hoad's victories in the list are also the Forest Hills TOC in 1959 and the Victorian Pro in 1960. That's 3 majors. Does he deserve a lot more? You imply that this is the case by referring to the opinion of many people that Hoad was 'one of the greatest', etc. His peak performance might have been amazing, but let's try to keep an objective look at his career: in 1956 he was great, but Gonzales, Sedgman, Segura were at the top too, so giving him more than 1 major for that year seems exaggerating. 1957 was not a good year, even by amateur standards, only his brilliant Wimbledon victory could save it...but does it have a 'true major' value, and after Rosewall's departure to the pro tour, were there many worthy rivals for Hoad? In 1958 he had a hard time adapting to the pro format, and barely made it into the top 5. 1959 was his great year, he's easily a co-no1 with Pancho, maybe the #1 but with a slight margin. His only big victory was the TOC. 1960, 1961, 1962 were good years, he was probably 3rd after Rosewall and Gonzales, but not more....
Maybe I'm being a little provocative, but I don't think with this career Hoad should be credited with much more than 3 majors.

I have read Raymond Lee's article. I disagree with many things in here. All his statistics are based on the assumption that 'Big tournaments' include the amateur slam + the 'pro slams':

1) Let's have a look at the field of, say, the 1964 Australian Open...Calling this a big one when Laver, Rosewall and Gonzales were out there playing matches for the ages seem like a joke.

2) though it was called a part of the Pro Slam, the US Pro was NOT a great tournament, at least from 1956 to 1962, and was a terrible tournament from 1960 to 1962, the farthest thing from a 'Big one' you could imagine.

3) this count leaves out tournaments that were obviously major tournaments with incredible all-star fields, such as every Forest Hills TOC, LA Masters Round Robin, or a handful of Australian Pros.

Plus, Mr Lee uses the number of tournaments won as a significant factor. This is totally ridiculous for anyone who knows how much the yearly number of opportunities to play a tournament is a parameter that has been changing through time....

Jonathan

deleted..............

Regarding the amateur vs. professional problem in analysing pre-open era years. Here is a possible partial solution. Rather than allocating 4 tournaments accross the amatuer and professional calander might it be fairer to allocate 4 tournaments for both amateur and professional, but then to give the different calanders different relative proportions. For example, if you allocate 75% of value to the professional tour in 1962 and 25% for the amateur tour. Laver's 4 majors would give him the full 25% of the amateur tour. However it is fairer than allocating 1 grand slam to the amateur tour because in years where the grand slams are split (e.g. in 1967 when Emerson won 2 slams and Newcombe won 2 slams) you can give the players both half the proportion of the sector you allocate to amateur players. So rather than giving 1 major out of the 4 in 1967 you effectivly give half a major to Emerson and half to Newcombe. Obviously you can adjust the proportions on a yearly basis depending on the strength and depth of the respecitve amateur and professional fields.

I think, Noel Edmonds is right. One has to find a binding formula, that gives both sectors of tennis in the pre open era their due. To chose sometimes only pro events, sometimes only amateur events, sometimes some amateur events, opens up a can of subjective choices.
Of course, as i have written in some posts here, the pro circuit pre open wans't that solid and deep structured, as it is today. The US pro had a slump in the late 50s, and early 60s, that is quite well known, but it is the oldest and most continious of all pro events, and in the most cases, it has to be regarded as important. The 1948 event pretty much set a standard, also since 1964, when it turned to Boston, it revived the pro game in the US. The French pro' status is too to be analyzed. It was mostly played end September, begin October, one week before Wembley, the real biggie in the eyes of all pros, and turned from clay to indoor from 1963 to 1968, when it was played at RG again. In terms of continuity and public acclaim, Wembley and US pro were in most cases the most prestigious events on the old pro tour.

The one nagging suspicion I have about all these counts is that the three players who almost always come out on top -- Gonzales, Rosewall, and Laver -- are all from roughly the same era. Now, it could be simple coincidence that the three greatest players of all time were contemporaries of one another. Or, and I think more likely, it could be that this methodology strongly favors the professional game of the 1950s and '60s. This is probably best explained by the nature of the pro tour and the events they played: while Wembley, the US Pro, etc., often had draws that included the best players in the world, these draws were nevertheless very small when compared with the traditional majors on the amateur circuit, much less the huge draws of the open majors today. The question, then, is whether a "true major" can even be defined by the quality of the players involved, or whether there is something essential about playing through and against a large field of opponents. Also, there is the more nebulous matter of "prestige" or "grandeur" associated with the traditional majors -- and the fact that these events gripped the public attention -- that was mostly lacking from even the very best pro events. On the eve of the open era, Gonzales himself expressed concerns about how the pros would now fare after so many years of playing in a somewhat incestuous format well outside the limelight. Of course, the pros did all right for themselves in those early open years, but nevertheless I do think we have to look at these pro records with something of a critical eye.

chaognosis said:

The one nagging suspicion I have about all these counts is that the three players who almost always come out on top -- Gonzales, Rosewall, and Laver -- are all from roughly the same era. Now, it could be simple coincidence that the three greatest players of all time were contemporaries of one another. Or, and I think more likely, it could be that this methodology strongly favors the professional game of the 1950s and '60s. This is probably best explained by the nature of the pro tour and the events they played: while Wembley, the US Pro, etc., often had draws that included the best players in the world, these draws were nevertheless very small when compared with the traditional majors on the amateur circuit, much less the huge draws of the open majors today.

Click to expand...

Good point chaog, I already thought about this earlier in fact...The fact that possibly the 3 greatest players of the game have been contemporary and regularly played each other is amazing. But is it much more unlikely than 2 players passing the 14-Open Slams bar within 6 or 7 years of each other, as it seems might happen soon?

The pro draws were usually 16-men for the big tournaments, which is indeed much smaller than the 128 format used in the Slams for ages (except the Australian). Still, are these three first rounds that significant?? With today's depth and international game, a big upset of a top player is not unheard of. But in the 60s, the top-tenners would encounter very little trouble getting to the quarters. Only the matches involving the big shots really mattered, and I don't think anyone could suggest that Gonzales, Rosewall or Laver would've lost many matches against random top-100ers, had they played these extra first rounds...
For the year 1968, Laver is 22-1 against amateurs. Rosewall (34 years old) is 17-1. Gonzales (38 ) is 10-3.
Laver's losses in 1968-1970 slams came to Rosewall, Okker, Drysdale, Ralston, that is fellow pros, that he could have encountered in any small pro field, which he frequently did in the WCT tournaments.
Rosewall won 4 big-fielded Slams in the Open Era, the last at age 38.
Gonzales, nearly 40, reached an RG semi (lost to Laver), the US quarter (lost to Okker, 2 rounds of 16 (lost to Roche and Ashe), and won 2 Los Angeles tournaments, with their 64-field (the last at age 43).

So I think the stats speak by themselves: these old pros, all past their prime, even Laver, had not much trouble with the 128-men events and the upsets they made possible. They were defeated sometimes of course, but by top-tenners, and nearly always in the last rounds of the tournaments...

Jonathan

SgtJohn said:

So I think the stats speak by themselves: these old pros, all past their prime, even Laver, had not much trouble with the 128-men events and the upsets they made possible. They were defeated sometimes of course, but by top-tenners, and nearly always in the last rounds of the tournaments...

Jonathan

Click to expand...

The challenge of the 128-man event is that it requires tremendous stamina. So while it may be unlikely for someone like Laver to lose in the opening rounds of such an event there would certainly be some carry-over of fatigue. No doubt, preparing to play a 128-man event is different than preparing for a 16-man event. It is not out of the question to presume that the results of these pro events would have been somewhat different with the current format of majors that we are used to.

A good example would be this year's Wimbledon in regards to three players and all semifinalists: Nadal, Djokovic and Gasquet. Each of these guys would have probably fared much better with a smaller field, just due to the fact that they would not have had to deal with as difficult and inconsistent a schedule. The mileague of a seven-round event (if one makes it to the final) creates unprecedented complications. In the case of Gasquet and Djokovic these guys were completely spent going into the semi not only due to the length of their quarter final matches but also due to the weather delays and the havoc those wreacked on the first four rounds. A smaller pro event with the same top seeds could have yielded a completely different result (for one, it would have had a healthy Tommy Haas facing Roger Federer).

I'm not sure whether all of this renders the pro championships as lesser accomplishments - not necessarily. But the notion of a major as a test of endurance certainly isn't something that would apply as evenly to a pro championship that the likes of Rosewall were accustomed to.

The weather conditions is one of the biggest factors to test a player's endurance. It's a bigger factor in a major than in a 16-man pro event. The typical heat of a French Open or French Pro event is another factor that becomes that much more impactful over a two-week major than a shorter 16-man event. Rosewall piled up a load of French Pros, but something tells me that he would have had fewer French Opens just for that fact in particular - winning the French Open is like triathlon; like an ultimate test on the body. The French Pro falls short of that.

Good reasoning. Lots to think about, of course.

Thank you also for pointing out an error in my own tallies. Given the three-tournament approach and my own division of amateur/pro eras after 1938, Perry should be credited with 11 wins, tying him with Borg in sixth place overall. Almost every system I have ever tried results either with Perry and Budge tied, or with Perry just narrowly ahead--not exactly the conventional wisdom. Your own research of course has them close as well, but gives Budge the edge, which would probably raise fewer eyebrows. I am curious exactly which tournaments you are counting for the years between World War I and World War II... these to me were probably the most dramatic and exciting years in tennis history!

Pro and amateur tennis in the pre open era had different formats and inviroments, but the guys who came on top in both, showed, that they could adjust to these different formats. That a particular period brings up high scores in one sector, is not only a phenomenon of the 50s and 60s. In the 70s and 80s, players had astonishing high percentages in career matches won/lost: Borg, Connors, Lendl, McEnroe all have over 80-81 %, which is in fact higher than the actual stand of Federer. On the simple major count, one would expect much higher numbers for these guys, and yet only Borg has over 10. Contrary, Sampras has a high major score, but his match percentage is lower than that of all these older guys.
Maybe these factors are related to time and circumstances: For the 70s, even old Laver had a percentage of ca. 80%, while he didn't better in his prime on the old pro circuit. The old pros played all they could for small money, even if they couldn't barely walk. In the 70s computer rankings system, percentages counted. A Connors in the 70s, when feeling unwell, often stepped out of a tournament, before losing a match, so his loss wasn't counted for the percentage. Sampras could afford many match losses, because of the 14 tournaments a year rule, they didn't count for the computer ranking. Federer now seems to be more careful about his percentage. In the last years, he stepped out of half a dosend events, he was scheduled to play, simply citing fatigue.
So i am sceptical on all statistical attempts, as cleverly and thoroughly they are made. Lee has ten factors, and he regards absolute numbers of titles (why not) and percentages (and even tour wins of the old pro circuit of the 50s). I would add some, as i wrote above, especially years for Nr.1. But the best approach in my view, is still a hermeneutical approach: Study the career of a player carefully, with all parts of a career, put it in historical context and relate the results to the three parameter: dominance, longevity and versatility.

urban said:

So i am sceptical on all statistical attempts, as cleverly and thoroughly they are made. Lee has ten factors, and he regards absolute numbers of titles (why not) and percentages (and even tour wins of the old pro circuit of the 50s). I would add some, as i wrote above, especially years for Nr.1. But the best approach in my view, is still a hermeneutical approach: Study the career of a player carefully, with all parts of a career, put it in historical context and relate the results to the three parameter: dominance, longevity and versatility.

Click to expand...

Of course I agree with you, urban. None of the list I ever made was intended to be a kind of "GOAT-formula", every one of us has his own, evolving "GOAT-list", that is for a large part based on subjective appreciation of a player's career, supposed peak performance, or even character.
But I think making lists, coming up with statistical studies is, firstly, fun (at least for some people), and also a good way to learn about tennis history, and maybe to change points of view... All in all, the stats cannot give us the truth, but at least they can get us a bit closer to it.

Jonathan

chaognosis said:

I am curious exactly which tournaments you are counting for the years between World War I and World War II... these to me were probably the most dramatic and exciting years in tennis history!

Click to expand...

Well at first, my inclination towards symmetry and regularity prompted me to select 4 events a year even for these periods, but I had to conclude it was not right, that the field was way too damaged by the wars for me to call 'majors' some very depleted tournaments.

I finally chose 4 majors for 1915-1918 (so as to keep a kind of regularity though, the war years counting as 1 year).

The US Championships for 1915, 1916, 1917 count as Majors, as they were not too depleted and featured some of the best American players (Johnston, McLoughlin, Williams). As my idea was to choose one slowcourt for every year, I had to choose one among my 4 tournaments. It is the 1918 US Clay court championships, won by Tilden. It is kind of a 'moral' choice too: I found an archive article from a December 1918 issue of the NY Times, commenting on Murray's designation as the best player of the year. The journalist suggested it might be a bit unfair as Tilden was excellent throughout the year, and his only loss came in the US Finals to Murray, when he was suffering physical problems, it seems...
My winners for the WW1 era are then Johnston, Williams, Murray, Tilden, which seems a fair repartition considering their respective level of play during these years...

My 'majors' for WW2 are:

1940: US Pro and the US Open
1941: US Pro and Forest Hills RR
1942: US Pro
1943: US Pro
1944: None
1945: US Pro Hardcourt and US Pro Claycourt

Jonathan

Jonathan: Not to put words in Chaog's mouth, but I think that when he wrote "the years between World War I and World War II" he meant 1919-1939, rather than the years during each World War.

I'm also curious about which tournaments you're counting during the 1950s and 1960s.

I think that your list, plus the efficiency factor, is one of the most valuable devices I've seen for rating players. Of course it's true that these things have limitations, but that doesn't mean they have no value.

Chaog and Urban make very valuable points, repectively, that the major wins of the 1950s and 1960s might well be inflated due to small fields whereas the match-win percentages of the 1970s and 1980s might well be inflated due to some other consideration. But I think Urban could be stretching it too far to say that a disparity in era explains the gap between Federer and Sampras in winning percentage. Those two played about the same number of matches in their prime years, and both were striving to retain the #1 ranking. The difference says something about their respective dominance and consistency.

A slight amendment: Upon re-reading Urban's post, I see his point about Sampras's losses not counting toward the rankings. This makes sense, due to the fact that the number of countable tournaments has apparently increased between the 1990s and 2000s. So maybe this does explain the disparity in win percentage between Sampras and Federer. But I don't know if I'm convinced entirely. If Sampras didn't care about winning those extra tournaments, then why did he enter them? Could appearance fees really have been worth the tiring travel, the losses, and the risk of injury?

John123 said:

A slight amendment: Upon re-reading Urban's post, I see his point about Sampras's losses not counting toward the rankings. This makes sense, due to the fact that the number of countable tournaments has apparently increased between the 1990s and 2000s. So maybe this does explain the disparity in win percentage between Sampras and Federer. But I don't know if I'm convinced entirely. If Sampras didn't care about winning those extra tournaments, then why did he enter them? Could appearance fees really have been worth the tiring travel, the losses, and the risk of injury?

Click to expand...

Sampras entered a lot of lesser tournaments to maintain his #1 ranking. Notice that once he began to let go of the standing circa '99 he stopped playing stuff like Lyon.

I think, i was misunderstood. I never wanted to make a comparison between Sampras and Federer.When Sampras had to play more tournaments in Europe in 1998, to stay Nr.1, he did. But his early round exits at RG and other clay venues, didn't count in the computer ranking score, because there was the 14 (+ 4) best tournament rule in the 90s, which only counted. Muster in 95 played some additional events, which didn't count on the ranking. In contrast, in the 70s, the ranking system was based on percentages: To win 8out of 10, was better than to win 15 out of 20. Today in the race, all majors and masters count, as far i have understood. On the other hand, Federer seems to think about losing and protecting his percentage, although the race is not based on percentage ranking, as in the 70s.So i understand, that he frequently steps out of scheduled events, not because of injuries, but simply citing fatigue. It is alright, given the schedule, but on the old Kramer pro circuit, they had to play with the head under their arms, simply because of the need for money.

I may be wrong, but wouldn't Sampras' early round exits at RG count? Didn't the majors automatically count? What they probably would not count would be his early exits from, say, Rome.

John123 said:

Jonathan: Not to put words in Chaog's mouth, but I think that when he wrote "the years between World War I and World War II" he meant 1919-1939, rather than the years during each World War.

Click to expand...

OK, sorry chaog, I misread your post.

About the pro/amateur split, as I said earlier, I don't give the pros any major before 1939. I understand that this is very unfair to Vines, Perry, Nusslein and Tilden, clearly. But this come from the lack of good tournaments involving all of these players. I'll go in some details:
-1931-1933: there is no real competition for Tilden, the pro field cannot be compared to the amateurs with Cochet, Vines, Crawford and Perry.

-1934-1936: Vines's only rival in tours is the aging Tilden. The only tournaments with a good field (ie Vines, Tilden and Nusslein at least) listed by McCauley are the French Pro 1935 and the Wembley tournaments, but Bowers and others strongly oppose the fact there were any Wembley events in 1934 or 1936. If I believe Bowers, and I acknowledge the fact that the best claycourters were amateurs, that leaves us with one good pro tournament for this period: Wembley 1935. I'm ready, on second thought to replace the weakest fastcourt amateur Major with Wembley for that year...

-1937-1938: these could have been great years! Imagine a big tournament circuit involving Perry, Vines, Tilden, Nusslein, etc. But it didn't turn out this way... I have always chosen not to count the 2-men tours as 'majors', I think it is constitutive of the game of tennis to have a minimum of 3 or 4 different opponents to win a big one. Plus, the '37 and '38 tours have not taken place in ideal conditions:both men became quickly mad at each other and demotivated.
Tournament pro tennis was totally separated, with Nusslein dominating the circuit. Only Perry played and won one tournament, the US Pro 1938, but it was terribly depleted (no Nusslein nor Tilden).

So, here is the list:

1919: Wimbledon, Forest Hills, US Clay Championships*, USA v Australia encounter**

*The World Hardcourt Championships in Paris didn't take place that year. Considering the level of play of the US players, the US Clay was then probably the best slowcourt tournament.

**In 1919 the US federation sportingly declined to play the Davis Cup, considerig that American tennis had not been nearly as hurt by the War as elsewhere. Australasia won the DC, with Patterson winning both his singles.
But after that, a friendly encounter was organized against the US. There were 6 matches, 4 singles and 2 doubles, each time with different players. The Americans didn't lose a single match. Johnston beat the old Brookes, and, in the most anticipated match, the US #2 Tilden beat Davis Cup winner Patterson, clearly proving that the US were the actual #1 team of the year. That's why I choose this encounter as a 'Major', and not the Davis Cup, to reward the actual best team. The 'winner' of the event is Tilden, as he beat the #1 player from the opposite team.

1920-1923: Wimbledon, Forest Hills, World HC Championships in Paris, Davis Cup.

1924: Wimbledon , Forest Hills, Olympics in Paris (on clay), Davis Cup

1925-1938: Wimbledon, Forest Hills, Roland Garros, Davis Cup

1939: US Pro, Wembley RR, French Pro, Southport Pro

chaognosis said:

Almost every system I have ever tried results either with Perry and Budge tied, or with Perry just narrowly ahead--not exactly the conventional wisdom. Your own research of course has them close as well, but gives Budge the edge, which would probably raise fewer eyebrows.

Click to expand...

Budge's position depends a lot of how you deal with the war years. He was probably the best player in the world from 1937 to 1942, maybe later, I'm quite confident he could have won between 2 to 4 pro majors a year from 1940 to 1944 , had there not been a war, but, along with Von Cramm and others, he had his career cut off by historical events. He had the peak performance and the dominance to be in the GOAT shortlist, but some of his best years were taken away from him...

About Perry, even if I like him personally through what I have read, I have to recognize that he might be a bit overrated in my list, thanks to the futility of the pro tournaments circuit. For 1934-1936, Vines was at least on a par with him, and in an ideal circuit, would have split majors with Perry.


Have a nice day,
Jonathan


[CORRECTION: according to Bowers, there indeed was a 1934 Wembley tournament, only 1936 is doubtful. So 1934 and 1935 Wembley would be possible pro 'True Majors' for the 30s. Yet it seems very difficult to judge the relative levels of the pros and amateurs for these years. Vines's edge against Perry in later tours hints at a possible superiority of the pros, but the pro-am matches described by Bowers, with young Von Cramm and Henkel destroying Nusslein and Stoefen, the masters of the European pro circuit, proves it is not so simple.]

One player who hasn't been mentioned much in this thread is Jack Kramer. Where does he fit into the GOAT conversation?

Take a look at this thread, really good arguments and comments by SgtJohn, urban and chaognosis.
I did'n read it until now.

And John123, Kramer must be in the GOAT conversation, no doubt.......

CyBorg said:

By the way, I have the Bill James abstract and I am an avid reader. If he was a tennis enthusiast he would definitely be into the likes of Tilden and Budge. He speaks glowingly of players like Honus Wagner - a guy who played 100 years ago.

Sadly, baseball is much more statistically richer, allowing James to build up an excellent argument... or should I say, criteria - the formula is based for the large part as to how much a player helps his team win.

In tennis there are no teams. It's just individuals (unless you count The Davis Cup). Very different.

One thing to add about James is that he is not terribly analytical about players (or selectively analytical, such as his thoughts on Hornsby). Most of his writings are anecdotal (Hal Chase) - fascinating in themselves, but they don't support the criteria. The rankings are based almost entirely on statistics. He also writes two pages for some players and a mere sentence for others. My favorite is his line for Don Mattingly: "100 percent ballplayer. 0 percent bullshlt." That's all.

Click to expand...

A Canadian who likes tennis, hockey, baseball, and particularly Bill James? :shock: I'm flying to Ontario NOW!

TheFifthSet said:

A Canadian who likes tennis, hockey, baseball, and particularly Bill James? :shock: I'm flying to Ontario NOW!

Click to expand...

Cyborg is an all around sports fan. Bill James is brilliant in baseball and he has helped the Red Sox win two championships with his great baseball analysis.

I would also recommend the ESPN Baseball Encyclopedia with Pete Palmer who is also brilliant.

pc1 said:

Cyborg is an all around sports fan. Bill James is brilliant in baseball and he has helped the Red Sox win two championships with his great baseball analysis.

I would also recommend the ESPN Baseball Encyclopedia with Pete Palmer who is also brilliant.

Click to expand...

Bill James is one of my favourite baseball writers, up there with Roger Angell (for a completely different reason though). Truly ahead of his time as a sports analyst, and also extremely well-versed in BB history. Palmer is a gem too, but I find his prose to be a bit uniform. I liked his early stuff the best.

TheFifthSet said:

Bill James is one of my favourite baseball writers, up there with Roger Angell (for a completely different reason though). Truly ahead of his time as a sports analyst, and also extremely well-versed in BB history. Palmer is a gem too, but I find his prose to be a bit uniform. I liked his early stuff the best.

Click to expand...

The thing about Palmer is that he's also excellent in football. The book "The Hidden Game of Football" written by the late great Bob Carroll, Pete Palmer and John Thorn is a classic.