Will Nadal pass Federer?

[El Dude]

I rather suspect the data is there already. As I've said in my preceding posts, you need to stop thinking about Roger and Rafa. What's going on with the lesser over 30s is what really matters. The sample is therefore much larger, and thus fewer additional years than you think are necessary

I addressed this above when I said that the entire top 50 has aged since around 2010. I did some research on this a few months ago. While ATP rankings didn't start until 1973, I looked at proposed rankings from before then all the way through 2016 and this is what I found:

The top 50 average age was 28 years old or higher from 1968-72, then dropped to the 26-27ish range for a few years. From 1975 (when it was 27.3) to 1986 (23.2) it became younger and younger, then rose again and was relatively stable, in the 24 to 25ish range until the mid-2000s, when it started creaping up. In 2010 it was at 26.1, the first time it was over 26 since the late 70s. It increased steeply for a few years, with the last three all being over 28; 2016 was 28.5, the highest since 1972 at least (which I estimate at 28.7).

One interesting thing to note is that the top 10 peaked at 29.6 in 2015, but dropped to 28.2 in 2016...but that's mainly because Roger and Ferrer dropped out.The current top 10 is back up to 28.8 and could get either older or younger before year's end, depending upon who edges in.

So yeah, it isn't just Fedal - the tour has aged. But the question is why and to what degree?

 

[El Dude]

@Federberg, after reading your last response I'm now convinced that the main point of difference is that you want to pigeon-hole me into something that I'm not doing or saying. All I can say is, stop it . I think you have this skewed view that I'm trying to map the entire future out in advance. Speculation is fun, no?

But to be clear: When I reference players from the past I am not saying that players today will follow the same trajectory. I am referencing the past to give context and try to understand what has come before, because it also helps us get a better sense of whether--and to what degree--the current context is truly different.

Thus the example of the age of Slam winners. I find it very meaningful that before 2017, only one Slam was won by any player at each age from 32 to 37, while 7 Slams were won by 31-year olds. That indicates that, historically speaking, there's a big drop-off from age 31 to 32. Again, seven 31-year olds won Slams; only one 32-year old has, one 33-year old, etc.

So if we want to know whether elites (Slam winners) are maintaining their primes later, we need to see more players win Slams at age 32 and older. So far in this bold new era we only have Roger. If Stan, Rafa, Novak, and Andy win Slams at age 32 and older, we'll have greater certainty. If they don't, well, that pushes us more towards the "Roger is an outlier" perspective.

And yes, I know you are talking about more than just the elites, but they're an important part of this discussion.

 

[Federberg]

I addressed this above when I said that the entire top 50 has aged since around 2010. I did some research on this a few months ago. While ATP rankings didn't start until 1973, I looked at proposed rankings from before then all the way through 2016 and this is what I found:

The top 50 average age was 28 years old or higher from 1968-72, then dropped to the 26-27ish range for a few years. From 1975 (when it was 27.3) to 1986 (23.2) it became younger and younger, then rose again and was relatively stable, in the 24 to 25ish range until the mid-2000s, when it started creaping up. In 2010 it was at 26.1, the first time it was over 26 since the late 70s. It increased steeply for a few years, with the last three all being over 28; 2016 was 28.5, the highest since 1972 at least (which I estimate at 28.7).

One interesting thing to note is that the top 10 peaked at 29.6 in 2015, but dropped to 28.2 in 2016...but that's mainly because Roger and Ferrer dropped out.The current top 10 is back up to 28.8 and could get either older or younger before year's end, depending upon who edges in.

So yeah, it isn't just Fedal - the tour has aged. But the question is why and to what degree?

That's the exact question. But I think your analysis while interesting is not refined enough. The older you are the more likely you are to retire because of injury. I don't think you normalise for anti-survivor bias. One of the ways to try to do this would be to analyse how the older folks are doing relative to their younger days. This is why I have a suspicion that the likes of Roger and Rafa may well be under-performing relative to the current norm. That would be truly fascinating to determine

 

[Federberg]

@Federberg, after reading your last response I'm now convinced that the main point of difference is that you want to pigeon-hole me into something that I'm not doing or saying. All I can say is, stop it . I think you have this skewed view that I'm trying to map the entire future out in advance. Speculation is fun, no?

But to be clear: When I reference players from the past I am not saying that players today will follow the same trajectory. I am referencing the past to give context and try to understand what has come before, because it also helps us get a better sense of whether--and to what degree--the current context is truly different.

Thus the example of the age of Slam winners. I find it very meaningful that before 2017, only one Slam was won by any player at each age from 32 to 37, while 7 Slams were won by 31-year olds. That indicates that, historically speaking, there's a big drop-off from age 31 to 32. Again, seven 31-year olds won Slams; only one 32-year old has, one 33-year old, etc.

So if we want to know whether elites (Slam winners) are maintaining their primes later, we need to see more players win Slams at age 32 and older. So far in this bold new era we only have Roger. If Stan, Rafa, Novak, and Andy win Slams at age 32 and older, we'll have greater certainty. If they don't, well, that pushes us more towards the "Roger is an outlier" perspective.

And yes, I know you are talking about more than just the elites, but they're an important part of this discussion.

Lol! I can only apologise mate. But every time you reference guys from decades ago to try to either explain or project I strongly believe you are on the wrong track. I have to call it like I see it

 

[El Dude]

That's the exact question. But I think your analysis while interesting is not refined enough. The older you are the more likely you are to retire because of injury. I don't think you normalise for anti-survivor bias. One of the ways to try to do this would be to analyse how the older folks are doing relative to their younger days. This is why I have a suspicion that the likes of Roger and Rafa may well be under-performing relative to the current norm. That would be truly fascinating to determine

That would be hard to determine, except in a relatively superficial manner (e.g. win%). Part of the problem with that is the difference between Roger's currently selective approach vs. his "play everything" approach of his youth. We can look at his 94 win% this year, but that's playing a light schedule and without playing on his weakest surface. Then there's the even more difficult problem of trying to compare the quality of the field, etc...

But I don't know why you'd want to exclude injury from this. Injury has always been a part of the game, and always will be (probably!), and can't be separated from aging and performance levels.

Look, I don't disagree that a lot of players are maintaining prime levels into their 30s. In fact, it is pretty irrefutable. But it is not so clear-cut as I think you are saying it is, and there are numerous factors at play (and I realize I'm veering from your original point about projecting into the future, but I think I addressed that and have moved on to more interesting things - namely, your assertion of a paradigm shift).

For example, right now there are 6 players age 22 or younger in the top 50, while in 2007 there were 11. In the top 20 it is even more extreme: 2 now vs. 6 in 2007. On the other hand, in 2007 there was only 1 32+ year old in the top 50, while there are a 11 now.

From the above (and other, more detailed analyses) we can say that 1) There are more older players and 2) fewer younger players in the top 50 than there were ten years ago. But what we cannot say is A) To what degree the two impact each other and either way, B) why.

In other words, I'm not disagreeing that players are maintaining prime form later, I'm questioning to what degree and why. As I said up-thread, my sense is that it is a combination of factors (my five factors above) that form a perfect storm. Some of these factors might lead to lasting change (e.g. improved conditioning and tech), so that there is truly a new paradigm, but some of these factors might be temporary (e.g. the talent void of the 1989-95ish group).

Lol! I can only apologise mate. But every time you reference guys from decades ago to try to either explain or project I strongly believe you are on the wrong track. I have to call it like I see it

And sometimes how we see it is not how things actually are. That's not what I'm doing, as I just explained (unless you think I'm being disingenuous). I'm mainly just providing context, both because I find tennis history interesting, and because it gives us something to play off of. And I'm also saying that this whole issue is more complex and less clear-cut than finding any singular causative factor for why more players are maintaining into their 30s.

 

[Federberg]

No no! I'm not trying to exclude injuries. I'm saying that the data is skewed because lots of older players retire from the game. Anyway I'm off to bed!

 

[El Dude]

I'm not sure how that skews the data, considering it has been a consistent factor seems the dawn of time. But regardless, good night .

 

[mrzz]

Wow! It was nice to come home after some late night tennis and follow that dialogue. If I understand @Federberg ´s point about injuries, I guess what he means is that what would be career ending injuries in the past now are treatable, or better, are prevented (image diagnosis, etc.). So this surely explains part of the story.

Two general comments: First, while I do agree that now we have different a situation, so we cannot trust too much on past data, on the other hand we cannot simply assume that all past data is irrelevant. It is, after all, the only data we have. Knowledge of the data is always useful, naive conclusions based on it (I am not saying that you did that, El Dude) are the problem. For example, in other posts and blogs El Dude have already shown that you should expect a pattern of an aging top 10 (which simply means that we have basically the same groups of players aging together) followed by a sharp decline given by a new crop replacing the old one. This pattern will probably continue -- with different boundaries maybe.

Second, other thing that might be important (a candidate for a sixth factor to the proposed list) is that the modern game may demand more focus and dedication from the players. So, maybe only when a guy reaches around thirty he may be able to focus more on his training, and also by this time he accumulated enough resources that he can chose better his schedule, pay for a larger/better team, and etc. In other words, in the 80´s one guy maybe could win out of sheer talent -- now this guy would need to work on his talent a bit more.

 

[Moxie]

I'm bringing over here El Dude's definition of "prime," etc.
Prime = span of good years (or "plateau"). For a great player, this might be the span in which they win big titles; for a second tier type, it might be the time in which they were ranked in the top 10.
Peak = best years ("mountains").
A game = highest level of play at any point in their career, but generally only possible in prime.

I know you guys have been arguing if all bets are off the table, or if it's worth putting it in the context of history (and I see both points from @ElDude and @Federberg.) And I wasn't sure where to put this question: but how do we determine Federer's (or Nadal's) "prime," anymore? Particularly Federer's. By the usual definition, it was done long ago. And he had, up until recently, a fairly graceful and predictable arc. Nadal was the one with the more up-and-down career. But, at 35, with two Majors in the bag, and the odds-on favorite for the USO, how is this not a "prime" year for Roger? And for Rafa, at least potentially? This is where I'm kinda with Federberg, in that the charts have just been broken. Though I do think it's fair to look at, for example, all the players of RF's generation that have retired, and the benefit they've received of the talentless/spineless generation behind them. We DO have to question why. And I agree that one of the answers is that they're both so #!@$ talented, that they've rewritten the script. But the other question is why not others, and why Stan Wawrinka, at this age, and etc. with the 30-somethings. There is a seismic shift happening, age-wise, in tennis. Will it all go back to normal when the Big 4 retire, or is this the way it works now?

 

[El Dude]

Good stuff, @mrzz. The injury thing makes more sense to me now - good point, if that what you mean, @Federberg. If not, good point to mrzz . Either way, I hadn't considered that.

Your first point: Well said. Not much to add, except I fully agree - and your example is a good one. The basic trends will re-occur, but how and when will change.

Your second point: Another good point that I've only partially considered. As someone just over 40, I very much understand that what was easy--in terms of the body--when I was 25, is not so easy now. A 35 year old can perhaps be just as fit as a 25 year old, but it takes more work, and more care. I remember when I could stay up and drink until all hours and be fully recovered by lunch. Now on the rare occasions that I over-indulge, it takes me the whole day to feel myself again.

There's another level that I think you're getting at, which is more strategic and we can see it with Roger. He's adapted his game and is being more picky about what he plays, so in essence he's optimizing his chances of playing his best in any given tournament.

So again, I think Perfect Storm Theory is the most adequate to the task of explaining the current context - that it is a combination of various factors that is allowing players to extend their prime years deeper into their 30s.

 

[El Dude]

@Moxie, who knows - that's the question we don't have the answer to. But it doesn't have to be one or the other. Maybe once the extraordinary Big Four are gone, things swing back a bit younger again, but not like it was in 1980-2010ish, which if you look at tennis history as a whole is actually an anomaly in how young players were. In other words, from one perspective we're just now going back to the historical norms of the 1970s and before, when players often remained peak deep into their 30s and won titles into their 40s. As recently as 1972, Pancho Gonzales won a Masters equivalent at age 43!

As I was writing up those definitions, I was thinking about how hard it was to apply them to Fedal. But I would also say both are still in their primes and never left; just as a plateau has peaks and valleys, so too does a player's prime. The peaks are the highest point, so what is remarkable about Roger is that he's not only still in his prime, but he's having a peak year at 35-36. As I said elsewhere, I think this is now his best year since at least 2009, and possibly since 2007. He's actually won an ATP 500 more than 2009, although he skipped clay this year so we don't know how he'd compare to his clay season in 2009.

 

[Moxie]

I took Mrzz's point earlier that maturity surely features. Tennis isn't just a physical game, it's a mental dogfight. Maturity and experience help. A lot. Then there's being about 1000 times better than everyone else, and healthy. I love seeing Roger and Rafa back in the fight, and proving just how good they really are. There's time for the youngsters or otherwise comers to have their days in the sun. For now I'm enjoying the two best ever playing well. And don't worry...I'm *this* close to conceding that Roger is the best. But I will wait until it all plays out.

 

[Federberg]

If I understand @Federberg ´s point about injuries, I guess what he means is that what would be career ending injuries in the past now are treatable, or better, are prevented (image diagnosis, etc.). So this surely explains part of the story.

Thanks for that @mrzz. You've made it clear to me now that I was confusing @El Dude. That's not actually what I was saying. I think the focus here is on when I said this...

"But I think your analysis while interesting is not refined enough. The older you are the more likely you are to retire because of injury. I don't think you normalise for anti-survivor bias. One of the ways to try to do this would be to analyse how the older folks are doing relative to their younger days. This is why I have a suspicion that the likes of Roger and Rafa may well be under-performing relative to the current norm. That would be truly fascinating to determine"

Ok what I mean is that any analysis that looks at the population of over 30s players rankings through time as an aggregate will always be flawed because a lot of players getting into their 30s will retire from the game (it could be injury, family anything). Why do I say this? Well consider the group of players +/-2.5yrs older than Federer now, vs that same group 5yrs ago. Many of those 5yrs ago have fallen out of the game, for any number of reasons. So if you look at the average ranking of the group now versus 5yrs ago it would be wrong to assume that the variation in the ranking average has a specific meaning. Particularly as the group size might be significantly different. This is what I called anti-survivor bias, as in the average ranking cannot tell us how those who've left the game would have done and we will confer on the survivors attributes that are false because certain members have left the group. The point is, what we're really trying to find out, is... have the over-30s really improved as players?

So.. stick with me here... what I suggested would be a better way would be to eliminate this problem by looking at the individuals within the group and assessing their rankings change vs 5yrs ago individually. Only when you have that data, can you aggregate back up. The same problem occurs in financial analysis, so I was trying to account for that. If you don't the results often tell a completely different story. I fear.. I've probably confused everyone now... I tried!

 

[mrzz]

Thanks for that @mrzz. You've made it clear to me now that I was confusing @El Dude. That's not actually what I was saying. I think the focus here is on when I said this...

"But I think your analysis while interesting is not refined enough. The older you are the more likely you are to retire because of injury. I don't think you normalise for anti-survivor bias. One of the ways to try to do this would be to analyse how the older folks are doing relative to their younger days. This is why I have a suspicion that the likes of Roger and Rafa may well be under-performing relative to the current norm. That would be truly fascinating to determine"

Ok what I mean is that any analysis that looks at the population of over 30s players rankings through time as an aggregate will always be flawed because a lot of players getting into their 30s will retire from the game (it could be injury, family anything). Why do I say this? Well consider the group of players +/-2.5yrs older than Federer now, vs that same group 5yrs ago. Many of those 5yrs ago have fallen out of the game, for any number of reasons. So if you look at the average ranking of the group now versus 5yrs ago it would be wrong to assume that the variation in the ranking average has a specific meaning. Particularly as the group size might be significantly different. This is what I called anti-survivor bias, as in the average ranking cannot tell us how those who've left the game would have done and we will confer on the survivors attributes that are false because certain members have left the group. The point is, what we're really trying to find out, is... have the over-30s really improved as players?

So.. stick with me here... what I suggested would be a better way would be to eliminate this problem by looking at the individuals within the group and assessing their rankings change vs 5yrs ago individually. Only when you have that data, can you aggregate back up. The same problem occurs in financial analysis, so I was trying to account for that. If you don't the results often tell a completely different story. I fear.. I've probably confused everyone now... I tried!

I got your point, and it is a good one. I guess it is not hard to grasp -- in other words what you are saying is that if you want to understand the effects of age, you need to study a "static" group, and see how it evolves. It is a fair point, and "anti-survivor bias" is a fine definition for it, by the way. Of course, one important thing one should always account for is the "survival rate" (which I also borrowed from financial analysis, good analogy).

But this does not mean that studying the average age of the top 10 -- a "dynamic" group by definition -- is meaningless (I know you did not say that). In fact, both analysis nicely complement each other.

 

[Federberg]

I got your point, and it is a good one. I guess it is not hard to grasp -- in other words what you are saying is that if you want to understand the effects of age, you need to study a "static" group, and see how it evolves. It is a fair point, and "anti-survivor bias" is a fine definition for it, by the way. Of course, one important thing one should always account for is the "survival rate" (which I also borrowed from financial analysis, good analogy).

But this does not mean that studying the average age of the top 10 -- a "dynamic" group by definition -- is meaningless (I know you did not say that). In fact, both analysis nicely complement each other.

Oh yes absolutely it's not meaningless, but we need to be aware of the flaws in the output The problem of course is that the sample can diminish so drastically that it has almost no value

 

[El Dude]

Ok, good idea about studying a "static" group....maybe I'll work on something when I get a chance. A couple problems come to mind, though.

First of all, it doesn't include any data about how many of a given generation retired, so in a sense it creates a "survivor bias" by only looking at the players who kept playing, which by definition are players with greater longevity.

Secondly, the more doable (that is, simple) such a study is, the less accurate or meaningful in terms of drawing conclusions. For instance, I could see looking at the ten highest ranked 30-somethings, and then compare their rankings at age 25. Maybe do that in five year increments: 2017, 2012, 2007, etc, comparing each group to their age 25 ranking.

A lot could slip through such an approach, but I think it would give us meaningful data.


Sent from my iPhone using Tapatalk

 

[GameSetAndMath]

Time to bump this thread. After all the four slams of 2017 are over, the difference in GS count between the two is exactly the same as it was at this time last year (when most thought Fedal are done). What do you folks think now? Rafa will be playing for at least five more years perhaps. But, he will be hot and cold as usual and won't be able to maintain this level throughout. Roger perhaps has two or three more years. But, he will be playing less and less to protect his body.

IMO, a diff of 3 is not enough with five more years of Rafa on tour. A diff of 5 would be safe. A diff of 4 would be semi-safe.

 

[rafanoy1992]

Unless Federer retires tomorrow or any day of the year, I think Federer will still be the all-time leader slam titles.

I know Nadal is five years younger than Federer but for Nadal to win 4 more slams at age 31 and up will be incredible hard to accomplish.

Also, you gotta remember Djokovic, Murray and Wawrinka still have something to say about Federer and Nadal.

One more thing: Federer is still playing so he could still win 1 more slam to career.

 

[El Dude]

I would be surprised if Roger won more than one more Slam, and if Rafa won more than two. First of all, Roger is 36 and partially won two Slams this year because Novak was out. If Novak comes back strong, Roger wins one Slam next year at most, and maybe none.

I don't buy this notion that Rafa will play five more years. Is that based upon Roger's age? Rafa plays 2-3 more years, or until his body really starts breaking down. When was the last time he strung together two high level AND healthy seasons in a row? 2010-11? Hard to imaguine 2018 will be just as good as 2017, and let's be honest: like Roger, Rafa's season looks better than it actually was because Novak was out and he had some really easy paths to titles (according to Tennis Abstract, the US Open was the easiest path to a Slam title in 30 years!).

And let's not forget that there are some good young players getting better and better each year. If the Big Four continue to rule, I think 2018 is their last year, and unlike 2017, it won't be just Fedal. In 2019, Roger turns 38 and Rafa 33.

So my guess is that Roger wins 0-2 more, and finishes with 19-21. Rafa wins 0-4 more, and finishes with 16-20. The only way Rafa passes Roger is if A) Roger hits his low projection and B) Rafa hits his high projection. Pretty unlikely.

 

[GameSetAndMath]

I don't buy this notion that Rafa will play five more years. Is that based upon Roger's age? Rafa plays 2-3 more years, or until his body really starts breaking down.

It has long been said that Rafa's body will break down and he will quit. But, he shows no signs of that. Of course, he will not be able to maintain this year's level for the next five years. Rafa will always be having the pattern of "on again and off again". But, I don't see why he can't play for 5 more years (and two good years of that five). Also, one must keep in mind that Rafa capable of winning RG even in his bad years.

In summary, unless Fed adds one more GS, I think situation is kind of dicy (although Nadal may not be able to pass Fed, the margin is too small to sleep peacefully without counting sheep). That is the relation between goat and sheep.