Top 20 Greatest Players of All Time (Yet another take)

[mrzz]

Reading through the posts I was thinking basically what Federberg has put it in other words just above, that once you state things like "player A should no be ahead of player B", you introduce a subjective bias on your system. Then you start tweaking it to achieve what you "want". The extra point for the runner up in the year end rankings is a good example. Why one point? What is the right proportion between #1 and #2?

The simple criteria, as the one you started with, is good simply because it avoids much (but not all) of these kind os discussions.

 

[El Dude]

^I think you're straying into the realms of subjectivity there Dude, even though you're trying to do the exact opposite. In my opinion I think you're on really shaky ground if you just dismiss Hewitt out of hand. Hewitt was at his best in an era that was transitioning towards greater use of the technology players today take forgranted.. particular stringing. To dismiss Hewitt would be like downgrading McEnroe because he wasn't able to make the transition from wood to newer materials while retaining his dominance. You should let the results of the data speak for itself, not dismiss it because it doesn't fit what you expect. Surely that's the greatest value you can extract out of it? Again.. at the risk of getting trollish response to my comment here (certainly not from you).. this is just my opinion.

I would also add that while Hewitt clearly wasn't effective coping with the bigger more athletic players that came shortly after him, the fact he was able to do what he did commands tremendous respect at least from me. I think we would all agree that the players today are far more professional, athletic, powerful than players just a few decades back. Sports technology has come a long way. I mean.. let's be real.. as talented as Laver clearly was.. deep down we all know he probably wouldn't translate quite as effectively into this era. Times move on, things improve. So all you can do IS compare players - without prejudice - in the era's they played in. The moment you dismiss an era.. the analysis loses it's value.. in my opinion

I hear you, but again the point isn't to dismiss Hewitt or any of the data, but to point out that focusing on weeks at #1 without looking at other data leads to skewed results. I mean, does anyone think Hewitt was a greater player than Becker or Wilander relative to their times? Or Connors/Lendl better than Nadal?

Reading through the posts I was thinking basically what Federberg has put it in other words just above, that once you state things like "player A should no be ahead of player B", you introduce a subjective bias on your system. Then you start tweaking it to achieve what you "want". The extra point for the runner up in the year end rankings is a good example. Why one point? What is the right proportion between #1 and #2?

The simple criteria, as the one you started with, is good simply because it avoids much (but not all) of these kind os discussions.

Again, I hear what you are saying but disagree with regards to runner's up. I'm including that because we have the data. The simplicity of the original system is there mainly because that's the only data we have across all eras: #1s and Slam totals. If we had #2-10 rankings, non-Slam wins, other titles weighted by importance, weeks at #1, etc, then I'd include all that.

I give it one point because I think that works with what I already established: 2 points for a shared #1 ranking and 3 points for a lone #1 ranking.

I'm not including runner's up to skew the data, or to "achieve what I want," but to be as inclusive as possible. I might even add lost Slam finals because I think we might have the data for that. Something like 1 point for appearing in a Slam final.

Again, the more data the better. But in terms of all-time rankings I want to stop at where the data ends for everyone. That might be:

#1 Rankings
Solo +3, Shared +2, Runner Up +1

Majors
Amateur/Pro Slams +2, Open Era GS +3, Finalist +1

 

[Federberg]

But that's the point. It shouldn't be about our biases and what we think. We should let the results of the analysis speak for itself. If anything what this might suggest is that the weightings of what is used for cross-era comparisons might have to be adjusted. We can't simply dismiss it because it might tell us something we don't agree with. Was Hewitt relatively more dominant than some of the guys you mention? Perhaps. But he won fewer slams, fewer tournaments, and the analysis will also reflect that. But to take a variable out because it might make him look better than someone we all think is better? There are always outliers. We can deal with that after a pure search for the truth..

 

[El Dude]

I don't think you hear what I am saying, Federberg. I am not saying that we should take weeks at #1 out, but that it shouldn't be privileged over a host of other data. My point was that ONLY looking at weeks at #1 (or any single statistic) gives us a skewed, one-dimensional perspective. I'm not saying that we dismiss Hewitt's 80 weeks at #1, but that we look at other things too.

Also, there's always going to be a subjective element in such rankings, unless we take a single data point and rank by that. But as soon as we take multiple criteria than we start adding a subjective element in terms of how to assign value, how to weigh those different factors.

 

[Federberg]

Can't argue with that. Perhaps we start by looking at the key variables then rank them? Without prejudice and before we see what the results are? Best we can do..


Sent from my iPhone using Tapatalk

 

[El Dude]

Sounds good! I'd be happy with a co-conspirator in this madness .

In no particular order, and off the top of my head, I think these factors or variables are most important:

Majors - wins
Majors - other results (especially second week: QF, SF, F)
Titles - all
Titles - Masters (or equivalent)
Titles - World Tour Finals (or equivalent)
Winning percentage
Rankings - year-end #1
Rankings - weeks at #1
Rankings - other (top 5/10)
Miscellaneous Accomplishments - Calendar Slam, Three Slams in a year, streaks, etc

As I said, there's a difference between pre- and post-Open Era (or, more properly, the ATP era, 1973-present) in terms of availability of data. We know, for instance, that Rod Laver won 200 titles--51 more than #2, Jimmy Connors--and we seem to know what those tournaments were, but it is very difficult to weigh them relative to each other like we can ATP events like 250, 500, and Masters.

Even the record on a relatively recent player like Jimmy Connors is confusing. Connors has 109 titles according to one source, 149 according to another. I assume that the 109 is ATP and the extra 40 are other tournaments - but that's a sizeable gap and its hard to account for them.

Down the rabbit hole we go...

 

[Federberg]

I think those look good. Apart from Rankings - other (top 5/10). Purely because we're dealing with the best of the best here. Unless you're thinking number of years in the top 10 or something like that. I agree with you about the difficulties in assessing number of tournament wins for someone before the open era. I have a suspicion that a lot of those matches were the champion of the prior year sits waiting in the final scenarios. Re: Connors, perhaps 109 is ATP and the others are WTC?

How about number of years winning 5 or more titles?

 

[Fiero425]

I think those look good. Apart from Rankings - other (top 5/10). Purely because we're dealing with the best of the best here. Unless you're thinking number of years in the top 10 or something like that. I agree with you about the difficulties in assessing number of tournament wins for someone before the open era. I have a suspicion that a lot of those matches were the champion of the prior year sits waiting in the final scenarios. Re: Connors, perhaps 109 is ATP and the others are WTC?

How about number of years winning 5 or more titles?

You have to break it down; even after the OPEN era began! There have been so many additions over the years, it wouldn't necessarily be fair for early pioneers of the Open era! With exhibitions and a lot more money being available for them, players actually signed contracts they knew would practically take them out of the running for some tournaments; WCT, WTT, and other events not part of the ITF! Some say that's why Connors wasn't able to complete his own calendar year GS in '74!

 

[mrzz]

I give it one point because I think that works with what I already established: 2 points for a shared #1 ranking and 3 points for a lone #1 ranking.

Don't get me wrong with the "what you want" part. I am pretty sure that what you want is simply the most fair system possible. But that's the point, if you have a lot of "degrees of freedom" to tweak the system, you'll try to optimize it's results. Whatever you think that leads to "unfair" results might be changed (with the best possible intention), and that's precisely the problem.

I am not saying that I am against a complex system (on the contrary, I think is pretty interesting). I am just stating that a complex system is much more open to debate, because you can judge each parameter by the result it produces, basically just changing where lays the subjectivity of making such a list (even if surely reducing it). That's why I think the simple systems have virtues of their own.

 

[El Dude]

How about number of years winning 5 or more titles?

The problem with this sort of thing is that it ends up with an arbitrary qualifier. Why not 4 or 6 titles? 5 is a nice round number, but I'd like to minimize arbitrary qualifications as much as possible - unless there's a good reason for the cut-off (e.g. runner-up for #1 ranking is someone who might deserve an argument to be #1, like Novak in 2013 or Roger in 2012; or the finalist in a Slam is someone who almost won and should be acknowledged).

Don't get me wrong with the "what you want" part. I am pretty sure that what you want is simply the most fair system possible. But that's the point, if you have a lot of "degrees of freedom" to tweak the system, you'll try to optimize it's results. Whatever you think that leads to "unfair" results might be changed (with the best possible intention), and that's precisely the problem.

I am not saying that I am against a complex system (on the contrary, I think is pretty interesting). I am just stating that a complex system is much more open to debate, because you can judge each parameter by the result it produces, basically just changing where lays the subjectivity of making such a list (even if surely reducing it). That's why I think the simple systems have virtues of their own.

I agree that simple systems have a certain virtue to them, and I think the one devised for all-time players is simple enough, even with runner's up and finalists. The point system is arbitrary but it is also intuitive enough to make it worth using, I think.

 

[JesuslookslikeBorg]

Connors played a lot of small tourneys in the 1970s arranged by his promoter/manager so they don't really count..

he played in some weird circuit in the 1970s, tennis had 3 circuits..everyone needs to be aware of this. atp tour, wct tour, and whatever Connors one was called.

 

[Fiero425]

Connors played a lot of small tourneys in the 1970s arranged by his promoter/manager so they don't really count..

he played in some weird circuit in the 1970s, tennis had 3 circuits..everyone needs to be aware of this. atp tour, wct tour, and whatever Connors one was called.

Isn't that what I said? uzzled :angel:

 

[Federberg]

Does this mean we should focus more on Lendls 94?

 

[Fiero425]

Does this mean we should focus more on Lendls 94?

That's a big number that hasn't been approached in a while; even with the dominance of the BIG 3! I can't see it happening anytime soon! Even the women are having trouble since they're so "deep!" Martina's 167 titles will be safe record for all time IMO; for men and women! :dodgy:

 

[El Dude]

Roger's got 82. Let's say he gets 4 next year, 3 in 2016 and 2017 and then 2 in 2018 - that's 94 total. I think he can do it.

 

[Fiero425]

Roger's got 82. Let's say he gets 4 next year, 3 in 2016 and 2017 and then 2 in 2018 - that's 94 total. I think he can do it.

Please GAWD no! I couldn't stand to see the GOAT decline like that! I was thinking the best thing to do was leave after winning his last Wimbledon in '12! He'll pick up a couple Masters here and there, but I can't see him doing much else except maybe embarrassing himself! Today is an embarrassment! Killing himself to make a final, then not play it at all is just the worst; esp. under the circumstances with DC soon! :nono :angel: :dodgy:

 

[Federberg]

Roger's got 82. Let's say he gets 4 next year, 3 in 2016 and 2017 and then 2 in 2018 - that's 94 total. I think he can do it.

Please GAWD no! I couldn't stand to see the GOAT decline like that! I was thinking the best thing to do was leave after winning his last Wimbledon in '12! He'll pick up a couple Masters here and there, but I can't see him doing much else except maybe embarrassing himself! Today is an embarrassment! Killing himself to make a final, then not play it at all is just the worst; esp. under the circumstances with DC soon! :nono :angel: :dodgy:

Lol! You're so pessimistic! The way I see it.. if Stan and Marin can win slams Roger can as well. It defies logic to think otherwise. So I'll keep hoping there are one or two left. As for getting to 94 it's doable for sure, but he needs to have another strong year next year. If he won 5 or 6 it would suddenly seem not just gettable but inevitable

 

[Obsi]

Don't forget that before the 1990's, slams were not the most important factor in GOAT discussion.

 

[Federberg]

Don't forget that before the 1990's, slams were not the most important factor in GOAT discussion.

This is a point I've made repeatedly. Although I would refine your statement by saying only Wimbledon and US Open had the weight all slams do now. It's very unfair to judge historic players based on the aspirations of current players. But what else can we do if we want to at least attempt to compare. That's why I place a higher premium on tournament wins and weeks at number 1. These are timeless aspirations

 

[brokenshoelace]

Don't forget that before the 1990's, slams were not the most important factor in GOAT discussion.

They were. There may not have been the defining factor the way they are now, but they certainly were the most important factor. Otherwise, what was?