voicemale, I agree. With the exception of Krajicek, those with winning records against Pete had a winning record by only one match, with relatively few total matches (less than 10) between them.

Safin (4-3)
Hewitt (5-4)
Roddick (2-1)
Bruguera (3-2)
Stich (5-4)

And Krajicek was (6-4) vs Pete.

Seeing as how Safin and Hewitt were nine years younger than Pete, I'd say Pete's records against them is pretty impressive...more so with Roddick being 11 years younger than Pete, and Pete's win against him was in the last match they played together in 2002 in the USO QF.

Stich and Karjicek were Pete's contemporaries, but 5-4 and 6-4 are still very close. You can't make the same claim for the Fed vs Rafa. And although Murray and Fed have only played 7 times (8 if you count the exhibition) Murray's margin over the Fed (5-2, 6-2 if you count the exhibition) is already more significant than Samrpras' competition was over him.

Nalbandian Bane - Solved.
Nadal Bane - Unsolved.
Murray Bane - Unsolved.
Simon Bane - Unsolved thus far in its very early stage.
If we all put our heads together, we might be able to do something similar for the WTA. Interesting stuff, no?

We all remember Sampras's Biggest Bane....his name was Mr. Clay, Mr. Red Clay.

LOL.

Red Clay - Totally Unsolved.

vm, i dont think anyone - the writer of that article and certainly not myself - is claiming that Rafa is the GOAT yet.

To me, the thing that comes across most clearly is that head to head match-ups aren't particularly useful in evaluating player's overall careers. Surely no one is going to claim that Karol Kucera is anywhere near the player Andre Agassi was or that Richard Krajicek was a more worthy rival of Pete Sampras than Agassi. Not only are there players that match-up poorly or well for whatever reason, luck plays a role in these numbers. In some cases the bulk of the matches might happen at a time when a player is recovering from an injury, when a player is past his prime or before his peak, or simply off form...I'm tempted to check to see whether Kucera won a bunch of matches against Agassi during the time he fell out of the top 100. Not only do these affect aggregate numbers but also whether a player appears to have "solved" his rival. For example, Roddick lost six of his first seven matches against Hewitt, but has won the last three. Has Roddick solved Hewitt or is Hewitt just not the player he was a few years ago? I think it's the latter. Then there's surface effects, since we've already argued this ad nauseum re Fed/Nadal, here's another one: do you really think James Blake's winning record against Nadal isn't partly because all the matches were played on hard courts?

If I'm going to choose a player as the greatest of all time, it's far more important to me to look at his whole record--against the whole tour, on all surfaces, and at all tournaments (not just slams)--than it is to look at how he performed in a few head to head match-ups.

The thing is you want to exclude data, it's just that your exclusions are based on whether they support your point rather than issues of statistical sampling. You want to be able to discount Sampras' losses to Hewitt, Roddick and Safin because he was past his prime while still counting his wins over Becker and Edberg when they were past theirs. You still want to be able to count wins against Alex Corretja before he made it to the top 100 and Andres Gomez when he was 32 and out of the top 200.

I am seeking statistically comparable samples, and I'm willing to do a little normalizing of the data to get them. So if I'm going to look at head-to-heads (and I don't think they're particularly useful anyway), I don't think matches played when the players were at very different stages of their careers are meaningful. Toss 'em all...if that means you don't have a direct comparison between the players (like between Federer and Sampras) then so be it. I'd rather use hypothetical data based on meaningful real data than real data that's meaningless (at least for judging overall careers), like the fact that Gisela Dulko is 2-0 vs Martina Navratilova.

Obviously I have no data on Blake vs Nadal on clay (and even on hard courts they haven't played that much), but the contrast in their records on clay makes it really really hard to imagine Blake winning, so I feel pretty comfortable assuming that if they played 30% of the matches on clay, Nadal would win all of them.

If I'm going to look at how Sampras performed against top clay courters, I don't think records against French Open winners guarantees wins over a quality clay courter. I'd be a heck of a lot more frightened by Novak Djokovic or Nikolay Davydenko on clay right now than Gaston Gaudio.

So maybe I'd look at a player's record against players ranked in the top 10 or top 5 at the time of the match (or maybe the ranking 6 months later--probably more accurate) and who had records on clay that were at least comparable to their records on other surfaces.

You keep saying how Sampras played in a much more competitive generation...well, that's your opinion. One could argue that unlike Federer, Sampras did not have any player who was dangerous on all the big stages throughout his career (for that matter, Sampras wasn't all that dangerous himself on one of the four biggest stages). Agassi comes closest, but even he was up and down, there was nobody with Nadal's consistency.

As to the rest, Becker and Edberg were past their prime when Sampras hit his...Krajicek was closer to being a true rival through the prime of his career than they. The reality is that Sampras played as many major finals against Cedric Pioline as he did against Becker and Edberg combined. And I still don't see why having a poor record against a very good, not great player like Krajicek is better than having a poor record against an all-time great like Nadal. If Federer were 2-6 against Roddick (as Samp was against RK at the same age), but had a winning record against Nadal, would that make him better?

I never wanted to discount his losses to Hewitt, Roddick and Safin. I don't know where that's coming from. I acknowledged them and stated how close the H2H records were while admitting it was still a losing record. The obvious point there was that Pete, while past his prime, was still very close with players in theirs although having a losing record...mostly seperated by just one single loss. Unlike the Fed is with Murray. And Murray and the Fed are only about six years apart, not a decade apart.


In other words, you're discounting them, while still including records of players who were just as passed their prime as Sampras as proof of Sampras' superiority.

Wow...so don't use actual results at all. Play "let's pretend." Because every match Pete played against a clay court specialist of his time, win or lose, (with the one exception you deemed OK, Kavelnikov) was when the player wasn't at his highest ranking (not even necessarily past his prime) NONE of those ACTUAL RESULTS should count when trying to prove Pete wasn't as horrible as everyone thought on clay. This wasn't even a direct comparison to Federer, it was just to show Pete had better results than people might have thought he did on the surface...but no...all those players were past their prime, or slumping, not at their highest ranking, had a bad day, were injured, etc....so they just don't count when proving Pete didn't suck so horribly on the red stuff.

Damn...there are a lot of W/L records in all the recordbooks everyone needs to rethink, according to you, when comparing players...and not even players of different generations! All those results I had shown that Pete had on clay, you want to disqualify, even though most of those players (except Gomez) WERE in his generation.

Fine--use Juan Carlos Ferrero as an example then. Same generation as Federer, but still by no means as significant a win in 2008 as over Davydenko or Djokovic. To be honest, I don't want to use Head:Head records from individual players at all. If I want to judge clay court ability against the best, I want to use the aggregate number of all the matches against the players who were at the top of their games. To do this properly, you need to filter out players who are not playing well for whatever reason. Age isn't the only thing involved. In the years that he beat them, the guys you mentioned were not great clay courters (aside from Kafelnikov)--they were losing lots of matches to mediocre players.

And I'm not saying you have to rethink the WL records themselves, I just wouldn't attach all that much significance to them. In most cases, the balance of power between players shifts over time, so what's useful to look at it is not the final aggregate number, but the balance at different stages of their careers.

I guess the reason I shouldn't be surprised with this level of disingenuousness from you goes back to when I first started posting. While I could certainly receive a few bonehead awards myself (Seles 3-1 over Graf in slams I called "ownership," comes to mind, I'll own that one, and any others) I showed, with actual results, (heaven forbid) how the Williams sisters dominated Hingis from '99 on. (and that's ironically generous, since Hingis was still in her prime and the sisters were still reaching theirs at that point...and according to you, it's only when players are at the top of their games that their results should count)

Your reply? Literally, and I quote: "As to your stats, five of Hingis' losses to Venus and Serena came after the 2001 Australian open when her foot problems began, the year she began to complain of foot problems. So no, they did not dominate her from '99 on."

Seriously, you thought it was OK to just not count all of Hingis's 2001 and 2002 results, because she complained of foot problems.

No, I thought that Hingis 2001 and 2002 results couldn't be used to prove The Williamses dominated her in 1999 and 2000--which is what you were doing. If you think there's not enough data in 1999 and 2000 to get a firm handle on the rivalry, '97 and '98 are objectively just as relevant as 2001 and 2002. By cutting out '97 and '98 you were cherry picking yourself.

But what added to the level of silliness was just a few posts above that laughable one, you showed how Seles was 3-12 vs Hingis at the time. (Yup.... all of a sudden H2H actual results were OK to use) Funny, Monica complained of foot problems and other ailments since practically day one of her comeback after the stabbings, so according to the logic above, none of those 15 matches counted, and "no Hingis did not dominate Seles."

I've never made any attempt to suggest that Hingis record against Seles is meaningful in terms of how she would have done against Seles in her prime. But in fact you were the one dismissing data again, since I was attempting to use Seles as an example of the fact that Hingis could, in fact, beat players who hit the ball hard. So again--it was your selective disqualification of results involved, and you dismissed that because oh, well, Seles feet hurt. Hingis also had against a good record against Pierce.


Selective disqualification of actual results, and inclusion of others, and retroactive physical excuses wiping out undesired match results.

Very nice.
And I would have felt close to 100% certain that Hingis would have beaten Majoli at the '97 French, and almost 100% sure at the '91 French that Agassi would have beaten Courier (even with some clay results), yet those "givens" just didn't happen. Weird how record books are written.
So someone who won the FO isn't necessarily a quality clay court player, and especially not at the time when so many players focused solely on clay?? Um...OK.


If you want to use your Gaudio example, the only comparable player in my list of FO winners would have been Gomez, since he was 30 and 32 when Pete beat him, as Gaudio is 30 now. The rest of those players were well within Pete's time, including Muster...unless, of course, Rafa isn't within the Fed's time.
Again, look at all the records that would get kicked out because of this. Actual results don't count if an opponent isn't in the top 5 or 10 at the time of the match, or within 6 months of the match...and at a time when the clay court champions were almost focusing solely on clay, thus having lower rankings because of it, no less.

Wow.


Agassi was an all-court threat for many years, even if he went away around the Brooke Shields time. And you're conveniently forgetting how strong Jim Courier was at all the slams in '91, '92, and '93, let alone a boatload of clay-court specialists thrown into all those years.

But you're saying Courier was one of the top threats on all the big events for Sampras entire career. He wasn't. He wasn't much of a threat at all in 1995, Sampras won many of his slams without having to worry about Courier, Becker, or Edberg...to use them as proof of the highly competitive nature of Sampras' time is only appropriate to the mid-90's. After that, the competition doesn't look all that impressive.

A 6-4 record with Krajicek, it's a losing record, for sure, with only 10 matches as its total over many (seven) years, yes, but Krajicek won Wimbledon, he's won a slam. What does that say about Federer being 2-5 (2-6 counting China exhib) against Murray, when Murray hasn't even won a slam yet?

And yes, having a winning record against a much better player who has many more slams and titles where the number of matches played over a longer period of time is much more statistically and realistically significant than having a losing record player to someone you've played much less over that time span who possesses just one single slam.

Pete being (20-14) vs Agassi and (16-4) vs Courier (10-3 if you want to throw out from '95 on, and I just know you do since Courier wasn't in his prime anymore ) while being (4-6) vs Krajicek is absolutely better than the Fed being (6-13) vs Nadal and (2-5) with Murray and (16-2) vs Roddick. Yes-sir-ee-Bob, the Fed would be delighted to be hypothetically (13-6) vs Rafa and hypothetically (4-6, or 2-6, even) vs Andy Roddick, or Andy Murray, for that matter, seeing as how Rafa has 6 slams and Roddick has 1 and Murray has, um, 0.

No, I thought that Hingis 2001 and 2002 results couldn't be used to prove The Williamses dominated her in 1999 and 2000--which is what you were doing. If you think there's not enough data in 1999 and 2000 to get a firm handle on the rivalry, '97 and '98 are objectively just as relevant as 2001 and 2002. By cutting out '97 and '98 you were cherry picking yourself.

I've never made any attempt to suggest that Hingis record against Seles is meaningful in terms of how she would have done against Seles in her prime.

But in fact you were the one dismissing data again, since I was attempting to use Seles as an example of the fact that Hingis could, in fact, beat players who hit the ball hard.

So again--it was your selective disqualification of results involved, and you dismissed that because oh, well, Seles feet hurt.

Federer has a very different pattern than Sampras. Basically, players either never or almost never beat Federer or they beat him quite regularly. Despite having a couple of players he has trouble with--and they're being very good players, so he has to play them frequently--he still has been compiling much better W-L records over the past few years than Sampras did in his prime. In large part, this is because he has no weak surface and is more consistent, so he very very rarely loses to rank-and-file players. I happen to think that whether a player is vulnerable to losing to the likes of Galo Blanco is just as important as how he does against the two or three players at the top of the game. More so, in a way, because you always play far more matches against ordinary players than great ones, so your performance against the ordinary is less prone to statistical anomalies. I also feel more comfortable in assuming that rank-and-file players of different eras are similar in their abilities than trying to determine the ability of the small number of players at the top.

By the way, while players don't tend to play as exclusively on clay any more, there are still plenty of players with huge disparities in their W-L record on clay. In fact, Nadal's career record on clay is .917 vs. .748 on other surfaces, a difference of 0.169. That's a bigger difference than Bruguera (.688-.524=.164), and almost as big as Muster (.762-.583=.179). The difference between Rafa and these guys isn't that he is so much better than them on hard courts, it's that he's so much better than those guys all the time--which is why beating Muster (unless it was in 1995 or 1996) doesn't give me much faith in someone's ability to beat Rafa. Most of his career, he wasn't that special. I realize he won a ton of clay tournaments but many of them were minor tournaments like Umag. Other than the one year he one, his French Open record isn't much better than Sampras'.

Facing a tired and achey Nadal, Federer had a golden opportunity but massively CHOKED in the 5th set of their AO final.

Time will tell whether Fed has the mental chops. If Sampras could win a Slam at 31, after not winning even a Little Sisters of the Poor tournament for over 2 years, certainly Federer can win #14 and #15. But if he fails, it will reveal a lot about his mental strength, and his place amongst tennis's greatest will be adjusted accordingly.