The v2 Forum
Would you like to react to this message? Create an account in a few clicks or log in to continue.

Discussion about W/L ratio

+14
socal1976
summerblues
dummy_half
biugo
Born Slippy
kingraf
LuvSports!
Johnyjeep
JuliusHMarx
HM Murdock
DirectView2
temporary21
laverfan
It Must Be Love
18 posters

Page 3 of 6 Previous  1, 2, 3, 4, 5, 6  Next

Go down

Discussion about W/L ratio  - Page 3 Empty Discussion about W/L ratio

Post by It Must Be Love Tue 16 Sep 2014, 2:10 pm

First topic message reminder :

Discussion about W/L ratios, Nadal's spread of tournaments in terms of surface, music, and whether GOAT debates can be concluded objectively.


Edited extract from one of the posts about subjectivity in GOAT debates:

summerblues wrote:In principle, you cannot ever avoid this type of subjectivity - even falzy's square with four sides ultimately requires agreement on subjective definitions.  But that does not mean that the statement itself does not reflect objective reality.
Good you bring up a square has 4 sides.
For me the debate is not whether the square having 4 sides is similar to a player being the best at tennis. It's just obviously not true. I think the real question now is why is resolving the question as to who is best at tennis not comparable to how many sides a square has.


First let me start with this statement:
Even if Player A has slightly better statistics in every department to Player B, it does not mean necessarily he is better, although it is likely he is.
Now most reasonable people will agree with that, but in case some don't, I'll clarify further.
We can take any statistic, and show what the statement just said. I've talked at length about slams and the natural fluctuations in difficulty to win them in the past; this time let me choose another statistic... let's say finishing year end number 1.
Now let's say hypothetically in 2015: Federer retires in January, Djokovic takes a year break to look after his kids, Murray starts trying to rebuild Hadrian's wall, and Nadal, Cilic, Del Potro, Wawrinka all have injury problems. Ferrer turns out to be the guy heading to number 1 in the rankings. Now let's say you are called 'Player A'.
For Player A, it would be require you to be a better tennis player to be number 1 in 2007/2010 (where Federer and Nadal had stunning years respectively) than in this dystopian 2015 (let's keep ceteris paribus in terms of technology and say you simply have access to the best technology at the time).

For those now who think they have a valid counter point by saying 'Yes IMBL, but how do you prove that you have to have a better tennis level in 2007 to be year end number 1 than this dystopian 2015??'- I can assure you- you don't actually have a valid counter point. Think about it, my statement said it is 'not necessary' a player with better stats is a better player- so therefore the burden of proof is on you- I just have to show that what I'm saying is a possibility for that statement to be correct.

So when this statement is acknowledged, that is when the problem to get a definition really emerges:
Even if Player A has slightly better statistics in every department to Player B, it does not mean necessarily he is better.

If the statement is true (which is demonstrably is), then it can mean that any definition that just looks at statistics (can be all, can be some) simply cannot be reliably accurate.

So therefore any definition would have to include factors as well as the statistics, and this simply cannot be objective in reality.
So the definition of 'better player' will remain subjective- and this is why GOAT debates cannot be objectively resolved like 'the square has four sides'.


Last edited by It Must Be Love on Fri 19 Sep 2014, 6:12 pm; edited 2 times in total

It Must Be Love

Posts : 2691
Join date : 2013-08-14

Back to top Go down


Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by DirectView2 Tue 16 Sep 2014, 8:29 pm

kingraf wrote:What team? Your example made no sense. Simple. As. That.

It's embarrassing.

Do you get embarrased for an online sports discussion counter arguments?  picard

kingraf wrote:

I used "we" because I commented on this thread and felt we aren't making any headway as a collective.

If you commented on this thread, stick to "I" than "we" why are you judging for everybody?

Secondly "What headway as a collective" you were looking for from this thread initially and how i deviated the thread from it?

The objective hidden in the thread is "Nadal the Greatest", by some filtered stats, and when the stats are questioned you like to go nit picking on me personally, when many questioned the stats from the mods to various posters.

kingraf wrote:
I mean really. If Nadal loses ten games to Kasporov... Well okay then... If Daffy Duck wins a Grammy does that mean Djokovic is a great dancer? If Serena scores a goal playing yard soccer does that make Ronaldo a great Ballet dancer?

You answered the question yourself here and found it embarassing , it makes no sense to compare win/loss ratio overall when a player has excelled in one particular surface and not the same distinction on the other, to add more the particular player in question has missed at least 3 full seasons on weaker surface in comparison to no missed season on stronger surface.

kasparov - Nadal example was a perfect shot to explain things easier for you and it did work.


Last edited by DirectView2 on Tue 16 Sep 2014, 8:35 pm; edited 1 time in total

DirectView2

Posts : 589
Join date : 2014-06-16

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by DirectView2 Tue 16 Sep 2014, 8:34 pm

It Must Be Love wrote:For HM:

3/ Has Nadal played matches or tournaments disproportionately on clay compared to the ATP tour ? (let me answer that one for you: no.)

Dis agree, Nadal has skipped n number of mandatory tournaments and matches on weaker surface and by the myth of the stats never ever missed a clay season or FO Very Happy

So either discuss w/l stats individually representing every surface or safely skip the thread as it doesn't make sense otherwise.

DirectView2

Posts : 589
Join date : 2014-06-16

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by DirectView2 Tue 16 Sep 2014, 8:37 pm

@ It must be Love

Do you agree if Djokovic would have missed the entire clay season he would have thumping lead over Rafa in h2h stats?

DirectView2

Posts : 589
Join date : 2014-06-16

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by It Must Be Love Tue 16 Sep 2014, 8:38 pm

DirectView2 wrote:@ "It Must Be Love"

Surface distribution stats is different to tournament missed stats. thumbsup  Did you understand that or purposefully you want to stay way from that stats?

OK, let me try and make it even clearer:
Tournaments played + Tournaments not played = 100% of all tournaments within a career

If I have shown, as I have done, that Nadal's distribution of tournaments played is almost exactly matching with the ATP tour, and closer than Federer's; then by mathematical proof it must be that the surface distribution in tournaments he has not played in is also very close to the ATP tour's distribution, and closer than Federer's.

Does that make sense?

It Must Be Love

Posts : 2691
Join date : 2013-08-14

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by It Must Be Love Tue 16 Sep 2014, 8:39 pm

JJ, my reply you may have missed on other page:

It Must Be Love wrote:
Johnyjeep wrote:What exactly is your definition of "best"? For me - it is winning. Winning tournaments....
Tennis is as cut and dried as that. It's not there for entertainment purposes. That's just a by-product. It's a game which is entirely based on line calls. It is either in or out. It is not subjective.
Interesting post JJ, and one I totally disagree with you on, and I agree with HM (and I guess Julius too?).

Alright to counter what you're saying, rather than actually bringing players into this (after which we may go down a line of debate we've gone before), let me use this metaphor:

I'm a scientist trying to conduct an experiment. I'm seeing at which rate different plants grow. The first plant (Plant A) I put in conditions of 10°C and see how fast it grows. It doesn't grow very quickly.
The second plant (Plant B), a different plant, I put in conditions of 35°C. It grows very quickly. Then a third plant (Plant C), another different variety, I put in conditions of 150°C and it unsurprisingly  fails to grow at all.

So I head back to the science committee, with a graph and spreadsheet in my hands and shout, 'Look, I've proved my case. Plant A grows at a higher rate than Plant C, but the fastest growing plant is Plant B for sure- look at my statistics !

So, have we actually proved that Plant B is the fastest growing ? Well we may have shown it's highly likely, but to make it clear that it is likely- it's also important to observe the variables first. You see, the problem with this experiment, was that the variables were changing.
It's only if we can accurately determine how those variables affected the statistics, that we can see whether we have 'proved' (or even less than that.. 'shown it's likely') that Plant B is the fastest growing plant.

^That is for context, all my quotes from now on will be me quoting JJ:

Woohoo science and sport - my two favourite subjects!!  Now you are talking my language. You have however, rather unwittingly I suspect, helped prove my point.
I don't think so...
But let's see

To put things simply, whenever an experiment is designed to prove or disprove any stated fact/null hypothesis at any level of learning, variables are kept to an absolute minimum.
 
Yes, that's correct

So when I state that  "the best player at a major (or any tournament) is the player who is holding the trophy at the end of the tournament" and then I design an experiment in which to do that (play the tournament - funnily enough exactly how it is done now), I think I am on fairly solid ground.  Variables are kept to an absolute minimum. In fact they couldn't be reduced any further.
Oh dear. You have missed out the obvious point.
I am not arguing that over a certain two weeks, the winner of a Grand Slam is the best player over those two weeks. However what I am arguing is that the variables would be different for each Grand Slam. It may for some reason or another (let's say either tennis is naturally in a transition period, or the top 300 players are on strike due to low pay), be that a slam is easier/ harder to win than another slam. Certainly it's almost impossible that the variables remain constant to the point that all the slams are of equal difficulty to win, there would naturally be fluctuations.
The issue with your experiment is that it assumes just because 'one player is best within those 2 weeks'- the level that one player had to play at is equivalent to the level another 'player who is best within those 2 weeks' will have to play at. It couldn't be more obviously flawed.

And thus, you cannot claim to be able to objectively say one all-time-great is better than another, with the pretence you are not being subjective.

It Must Be Love

Posts : 2691
Join date : 2013-08-14

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by JuliusHMarx Tue 16 Sep 2014, 8:44 pm

It Must Be Love wrote:And thus, you cannot claim to be able to objectively say one all-time-great is better than another, with the pretence you are not being subjective.

Excellent. So I can say that Henman is better than Federer, in my subjective opinion, and there is no way you can objectively disprove it. You can only counter it with your own subjective arguments, which I will reject as being subjective. Even better is that it renders any further argument as being pointless and unnecessary, as a) neither belief can be proven and b) there's is no need to argue or debate further*. Thank goodness for that.

* Unless a) there is an internal need or desire to change the other person's point of view to one's own, which would probably say more about the person who wants to continue the debate than the debate itself or b) you want to do it for just for fun to pass a bit of time, in which case the outcome is irrelevant anyway.

JuliusHMarx
julius
julius

Posts : 22358
Join date : 2011-07-01
Location : Paisley Park

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by DirectView2 Tue 16 Sep 2014, 8:49 pm

Ok since you fail to understand or purposefully refused to look into a stat, how about this, can you answer the question below?

1]Do you agree if Djokovic would have missed the entire clay season he would have thumping lead over Rafa in h2h stats?

2]Has Sampras retired by the end of 1999 his win/loss stat would have been way better at the expense of 2 slams, which one would you take it?

DirectView2

Posts : 589
Join date : 2014-06-16

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by kingraf Tue 16 Sep 2014, 8:53 pm

Okay I'll bite. But only because you're making it easy.

If Nadal faces Kasporov at chess, he's going to lose. There is simply not chance on earth for a different result. That doesn't apply to Nadal on hard court. It's completely inaccurate to assume Nadal's W/L decreases if he misses hard court swings. He went 22-0 on American Hard courts last year. He's 20-1 in New York since 2010.He's made the final of every hard court slam he has entered since USO 2011. I could easily argue nadal missing large chunks at places he'd improve his W/L.It's a stupid argument.
kingraf
kingraf
raf
raf

Posts : 16599
Join date : 2012-06-06
Age : 29
Location : To you I am there. To me I am here.... is it possible that I'm everywhere?

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by DirectView2 Tue 16 Sep 2014, 9:04 pm

kingraf wrote:Okay I'll bite. But only because you're making it easy.

If Nadal faces Kasporov at chess, he's going to lose. There is simply not chance on earth for a different result. That doesn't apply to Nadal on hard court. It's completely inaccurate to assume Nadal's W/L decreases if he misses hard court swings. He went 22-0 on American Hard courts last year. He's 20-1 in New York since 2010.He's made the final of every hard court slam he has entered since USO 2011. I could easily argue nadal missing large chunks at places he'd improve his W/L.It's a stupid argument.

Nice, since you want objective argument, I will bite back on you easy.

You are just seeing one dimensional not from all angles,

1]Nadal went 22-0 on hard courts once but how many times did he do that on Clay?

2]Nadal's losses more matches in hardcourt than in Clay, like the stat indicated, he losses close to 41% on indoors compared to 6% on clay.

So if Nadal plays 50 matches on clay and 50 matches on indoor chance are he will between 81-82% of matches but if Nadal skips those 50 matches and say plays just 5 of them in hard courts and 50 matches in clay then his win/loss stats would be 50/55 which could be 91%

In contrast Djokovic plays almost 100% on his week surface as well as strong surface, so Nadal always got Djokovic on his strong surface but Djokovic not always got Nadal on his strong surface.

This is an exact argument of how Federer never got Rafa on his best surface [except for 3 Wimbledon] at his prime where as Rafa always got Federer on his best surface at his prime.


DirectView2

Posts : 589
Join date : 2014-06-16

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by DirectView2 Tue 16 Sep 2014, 9:06 pm

@ King Raf, since you yourself agreed your argument was stupid, I will be glad if you don't post such things to deviate from the actual thread.

DirectView2

Posts : 589
Join date : 2014-06-16

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by kingraf Tue 16 Sep 2014, 9:34 pm

1 - uh... Once? 2010.
2 - sure he's got a weak Masters record. But he has two WTF final appearances and a Masters title indoor. How many players with a 60% win rate on hard court can boast that? For comparison, Murray has a better clay win percentage than Nadal has indoor. Problem is, Murray doesn't even have a clay court final, while Nadal has five finals at Elite level (2 WTF + 3 Ms). Nadal's Indoor record isn't in line with his indoor ability.
It's the same with Sampras on clay. From 1990-1998 Pete Sampras, he of one Masters on clay his whole career, had the fourth best W/L record on clay between 1990-1998. You telling me that's an accurate presentation?

I honestly don't understand your argument. Nadal doesn't play only 5 matches on hard or whatever, and he still hasn't played >50% of his season on clay, so that's irrelevant. Clay doesn't make up an unfair percentage of his season, and there is no justification in trying to pretend that clay isn't a surface on the world tour. It is. That simple. And I'm not gonna get into a Fed vs Nadal discussion. We've like just had a thread locked.
kingraf
kingraf
raf
raf

Posts : 16599
Join date : 2012-06-06
Age : 29
Location : To you I am there. To me I am here.... is it possible that I'm everywhere?

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by Johnyjeep Tue 16 Sep 2014, 9:35 pm

It Must Be Love wrote:JJ, my reply you may have missed on other page:

It Must Be Love wrote:
Johnyjeep wrote:What exactly is your definition of "best"? For me - it is winning. Winning tournaments....
Tennis is as cut and dried as that. It's not there for entertainment purposes. That's just a by-product. It's a game which is entirely based on line calls. It is either in or out. It is not subjective.
Interesting post JJ, and one I totally disagree with you on, and I agree with HM (and I guess Julius too?).

Alright to counter what you're saying, rather than actually bringing players into this (after which we may go down a line of debate we've gone before), let me use this metaphor:

I'm a scientist trying to conduct an experiment. I'm seeing at which rate different plants grow. The first plant (Plant A) I put in conditions of 10°C and see how fast it grows. It doesn't grow very quickly.
The second plant (Plant B), a different plant, I put in conditions of 35°C. It grows very quickly. Then a third plant (Plant C), another different variety, I put in conditions of 150°C and it unsurprisingly  fails to grow at all.

So I head back to the science committee, with a graph and spreadsheet in my hands and shout, 'Look, I've proved my case. Plant A grows at a higher rate than Plant C, but the fastest growing plant is Plant B for sure- look at my statistics !

So, have we actually proved that Plant B is the fastest growing ? Well we may have shown it's highly likely, but to make it clear that it is likely- it's also important to observe the variables first. You see, the problem with this experiment, was that the variables were changing.
It's only if we can accurately determine how those variables affected the statistics, that we can see whether we have 'proved' (or even less than that.. 'shown it's likely') that Plant B is the fastest growing plant.

^That is for context, all my quotes from now on will be me quoting JJ:

Woohoo science and sport - my two favourite subjects!!  Now you are talking my language. You have however, rather unwittingly I suspect, helped prove my point.
I don't think so...
But let's see

To put things simply, whenever an experiment is designed to prove or disprove any stated fact/null hypothesis at any level of learning, variables are kept to an absolute minimum.
 
Yes, that's correct

So when I state that  "the best player at a major (or any tournament) is the player who is holding the trophy at the end of the tournament" and then I design an experiment in which to do that (play the tournament - funnily enough exactly how it is done now), I think I am on fairly solid ground.  Variables are kept to an absolute minimum. In fact they couldn't be reduced any further.
Oh dear. You have missed out the obvious point.
I am not arguing that over a certain two weeks, the winner of a Grand Slam is the best player over those two weeks. However what I am arguing is that the variables would be different for each Grand Slam. It may for some reason or another (let's say either tennis is naturally in a transition period, or the top 300 players are on strike due to low pay), be that a slam is easier/ harder to win than another slam. Certainly it's almost impossible that the variables remain constant to the point that all the slams are of equal difficulty to win, there would naturally be fluctuations.
The issue with your experiment is that it assumes just because 'one player is best within those 2 weeks'- the level that one player had to play at is equivalent to the level another 'player who is best within those 2 weeks' will have to play at. It couldn't be more obviously flawed.

And thus, you cannot claim to be able to objectively say one all-time-great is better than another, with the pretence you are not being subjective.

So we both agree that the best player wins any singular tournament. But a player who wins 20 singular tournaments is not better than someone who wins 5? Because you wish to insert some new variables that are entirely subjective?

You want to throw career w/l into the mix. Fair enough. I don't. I'm confident that when it comes to judging who's the best over any given period - a players career w/l ratio will be someway down the pecking order.

I mean I couldn't tell you who had the best career w/l before Nadal. Or what it was? Anyone? Sampras has won 14 GS. His w/l ratio?

Plus I get the impression you don't know much about doing research and rejecting null-hypothesis? Because my statement made no mention of quantifying the quality of grand slam wins against each other. I was in no way trying to do that. I was trying to highlight why success means you are the best. In this case - the easiest way to demonstrate this was to make a statement about who is the best player at a tournament. To try and do one for "who is the best ever" would be far more difficult. And I personally certainly wouldn't include a players w/l ratio ahead of far far more important metrics.

Which one of Nadal's French Open's were the best? Which one of Sampras's Wimbledon's was best. What about Borg FO and Wimbledon doubles? Which was the best. Who cares? No one. I've never seen anyone try and ascertain that. Why? Because it doesn't matter.


Last edited by Johnyjeep on Tue 16 Sep 2014, 10:08 pm; edited 2 times in total

Johnyjeep

Posts : 565
Join date : 2012-09-18

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by Johnyjeep Tue 16 Sep 2014, 9:37 pm

JuliusHMarx wrote:
It Must Be Love wrote:And thus, you cannot claim to be able to objectively say one all-time-great is better than another, with the pretence you are not being subjective.

Excellent. So I can say that Henman is better than Federer, in my subjective opinion, and there is no way you can objectively disprove it. You can only counter it with your own subjective arguments, which I will reject as being subjective. Even better is that it renders any further argument as being pointless and unnecessary, as a) neither belief can be proven and b) there's is no need to argue or debate further*. Thank goodness for that.

* Unless a) there is an internal need or desire to change the other person's point of view to one's own, which would probably say more about the person who wants to continue the debate than the debate itself or b) you want to do it for just for fun to pass a bit of time, in which case the outcome is irrelevant anyway.

thumbsup

Johnyjeep

Posts : 565
Join date : 2012-09-18

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by Johnyjeep Tue 16 Sep 2014, 9:40 pm

It Must Be Love wrote:
DirectView2 wrote:@ "It Must Be Love"

Surface distribution stats is different to tournament missed stats. thumbsup  Did you understand that or purposefully you want to stay way from that stats?

OK, let me try and make it even clearer:
Tournaments played + Tournaments not played = 100% of all tournaments within a career

If I have shown, as I have done, that Nadal's distribution of tournaments played is almost exactly matching with the ATP tour, and closer than Federer's; then by mathematical proof it must be that the surface distribution in tournaments he has not played in is also very close to the ATP tour's distribution, and closer than Federer's.

Does that make sense?

This just isn't even close to being correct.

Johnyjeep

Posts : 565
Join date : 2012-09-18

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by It Must Be Love Tue 16 Sep 2014, 10:12 pm

Julius wrote:Excellent. So I can say that Henman is better than Federer, in my subjective opinion, and there is no way you can objectively disprove it. You can only counter it with your own subjective arguments, which I will reject as being subjective.
If I said that in my opinion I subjectively disagreed with you and felt Federer was better than Henman, you can't 'reject' my opinion as 'subjective' as my opinion is subjective anyway.
Also I don't think debates where one winner is 'proved' is pointless- most debates can't have proof either way.

So what is your opinion on this debate then; do you agree with JJ that these 'which player is better' arguments can be 'won' objectively, or do you agree with me that both sides are always subjective ?
This leads onto another thing, when I said to you via PM something was bizarre; it was that I felt you would never ever ever agree with me on any topic publicly, even if you did actually agree with me.
For example if I had been arguing JJ's position that GOAT debates can be objectively concluded, I reckon you would still be disagreeing with me.

It Must Be Love

Posts : 2691
Join date : 2013-08-14

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by It Must Be Love Tue 16 Sep 2014, 10:16 pm

Johnyjeep wrote:

This just isn't even close to being correct.
It is, think about this carefully.
Nadal's proportion of tournaments played in terms of surface distribution is almost identical to the surface distribution on the ATP tour overall.
So therefore if you take the flipside and just consider the tournaments he hasn't played (all the rest), that would also be proportionate to the ATP tour overall.
For example Nadal has played 65% of his tournaments away from clay, and the tournaments overall away from clay is 66%- so even if you consider the flipside his tournament/ surface distribution is still proportionate.

It Must Be Love

Posts : 2691
Join date : 2013-08-14

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by It Must Be Love Tue 16 Sep 2014, 10:30 pm

JohnyJeep wrote:So we both agree that the best player wins any singular tournament. But a player who wins 20 singular tournaments is not better than someone who wins 5? Because you wish to insert some new variables that are entirely subjective?
No, I am not inserting new variables; you're ignoring what is obvious- a 'singular tournament' cannot be used as a consistent unit (despite the deserved winner winning each particular event)- it is not a case of 20x>5x (which as long as x is positive you can be sure of, ie objective), but 20x>5y (which you cannot be sure of, ie subjective).
Let me use a very extreme example to illustrate my point:
-The top 200 all go on strike for the next 5 years, and all players under 21 also go on strike
-The player currently ranked 258 wins 15 Grand Slams in the next 5 years, and does well in the variables you yourself think important for the 'GOAT' debate.
-Surely you can then see that 'objectively' claiming he's better than anyone based on his Grand Slam success (and other variables) is ridiculous.
-My point is not that this extreme example is likely to happen, but that naturally- these fluctuations in difficulty of winning slams is always likely to exist- so it very difficult for my objectively to claim victory for any particular player.

JJ wrote:What about Borg FO and Wimbledon doubles? Which was the best. Who cares? No one.
But an argument such as this only becomes relevant when someone like you tries to claim that they can 'objectively' see who is a better player- based on statistics. Most people say 'ok this is my subjective opinion, and this is my evidence as to why I believe it'. So my reply to you is not needed to be used much by 'anyone'.

It Must Be Love

Posts : 2691
Join date : 2013-08-14

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by Johnyjeep Tue 16 Sep 2014, 10:31 pm

IMBL I do admit I probably posted a bit hastily on that. I'll have a think. But not now as I'm knackered!! Just when I read it, it sounded completely wrong.

Johnyjeep

Posts : 565
Join date : 2012-09-18

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by It Must Be Love Tue 16 Sep 2014, 10:34 pm

Johnyjeep wrote:IMBL I do admit I probably posted a bit hastily on that. I'll have a think. But not now as I'm knackered!! Just when I read it, it sounded completely wrong.
You're referring to the point on Nadal's tournament surface distribution matching the ATP tour in general meaning the flipside is also correct, yes ?

Don't worry, we all get tired, also I know I can rely on you for strong but well-tempered debates, which I like thumbsup

It Must Be Love

Posts : 2691
Join date : 2013-08-14

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by JuliusHMarx Tue 16 Sep 2014, 10:37 pm

It Must Be Love wrote:
Julius wrote:Excellent. So I can say that Henman is better than Federer, in my subjective opinion, and there is no way you can objectively disprove it. You can only counter it with your own subjective arguments, which I will reject as being subjective.
If I said that in my opinion I subjectively  disagreed with you and felt Federer was better than Henman, you can't 'reject' my opinion as 'subjective' as my opinion is subjective anyway.
Also I don't think debates where one winner is 'proved' is pointless- most debates can't have proof either way.

So what is your opinion on this debate then; do you agree with JJ that these 'which player is better' arguments can be 'won' objectively, or do you agree with me that both sides are always subjective ?
This leads onto another thing, when I said to you via PM something was bizarre; it was that I felt you would never ever ever agree with me on any topic publicly, even if you did actually agree with me.
For example if I had been arguing JJ's position that GOAT debates can be objectively concluded, I reckon you would still be disagreeing with me.

You're making assumptions about me that aren't true. But as I cannot prove that, and it doesn't really bother me, I won't spend time trying.
By 'reject' I mean 'disagree' - I see no purpose to getting bogged down in semantics. "I will consider your subjective option wrong and thus disregard it" - is that better?
Why do I have to agree with either you or JJ? Perhaps I partially agree with both of you. Perhaps black and white are not the only options.
There is an overwhelming weight of statistical evidence that Federer is better than, say, Matteo Viola . That is close enough to an objective truth as to make it, for all intents and purposes 'objective'.
The lines between objectivity and subjectivity thus become blurred, or grey. The more evenly balanced the weight of evidence the more subjective it becomes. The point of the argument thus becomes, for some posters, to change the subjective opinions of other posters. To me that makes the argument ultimately pointless (in the sense that that is not my goal). I find myself interested in why there is this desire to defend one's champion and to attempt bring others into the fold, as it were. I have my theories, but they are not really tennis related, so there is no need to go into detail on this forum.

JuliusHMarx
julius
julius

Posts : 22358
Join date : 2011-07-01
Location : Paisley Park

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by It Must Be Love Tue 16 Sep 2014, 10:41 pm

Julius wrote:The lines between objectivity and subjectivity thus become blurred, or grey. The more evenly balanced the weight of evidence the more subjective it becomes.
This is undeniably a fair point.
However in 'GOAT' debates; generally there is no huge gap like there is between Federer and Viola, so subjectivity would play bigger role I feel.

It Must Be Love

Posts : 2691
Join date : 2013-08-14

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by Johnyjeep Tue 16 Sep 2014, 10:48 pm

This has to be the most complicated way of saying "player x is better than player y because I said so" I've ever seen.

Of course there is no stated fixed indicator to measure against across tournaments. We all know this. Who wins the biggest tournaments most times has always been seen as pretty good indicator. W/L hasn't. That I'm afraid is not subjective.

Tbh..the one extreme example you gave there very very much reminds me of a poster who engaged me a PM conversation once. And now I think about it, I'm fairly confident you are one and the same.

Don't take this the wrong way..but I don't wish to carry on this thread of conversation as a consequence.

Johnyjeep

Posts : 565
Join date : 2012-09-18

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by JuliusHMarx Tue 16 Sep 2014, 10:52 pm

It Must Be Love wrote:
Julius wrote:The lines between objectivity and subjectivity thus become blurred, or grey. The more evenly balanced the weight of evidence the more subjective it becomes.
This is undeniably a fair point.
However in 'GOAT' debates; generally there is no huge gap like there is between Federer and Viola, so subjectivity would play bigger role I feel.  

Yes, that is more or less what I said. But that point, which we agree on, is moot here, because this thread is in no way related to a GOAT debate is it?

JuliusHMarx
julius
julius

Posts : 22358
Join date : 2011-07-01
Location : Paisley Park

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by JuliusHMarx Tue 16 Sep 2014, 10:54 pm

Johnyjeep wrote:This has to be the most complicated way of saying "player x is better than player y because I said so" I've ever seen.

Perhaps because if it's complicated, it might be hoped to have more chance of convincing others to modify their subjectivity to match one's own.

JuliusHMarx
julius
julius

Posts : 22358
Join date : 2011-07-01
Location : Paisley Park

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by It Must Be Love Tue 16 Sep 2014, 10:56 pm

JohnyJeep wrote:Who wins the biggest tournaments most times has always been seen as pretty good indicator.
It always has been, and it always will be, and I agree with that.
However this is different to what you were claiming earlier, which is that these claims can be made without being subjective, which they can't.

Tbh..the one extreme example you gave there very very much reminds me of a poster who engaged me a PM conversation once. And now I think about it, I'm fairly confident you are one and the same.
I wrote a long article saying who I am- what more am I meant to do? I sign off as Amritia, I did on my first post back a week or so ago.


Last edited by It Must Be Love on Tue 16 Sep 2014, 11:10 pm; edited 2 times in total

It Must Be Love

Posts : 2691
Join date : 2013-08-14

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by It Must Be Love Tue 16 Sep 2014, 10:58 pm

Julius wrote:Yes, that is more or less what I said. But that point, which we agree on, is moot here, because this thread is in no way related to a GOAT debate is it?
Glad we agree, I've been saying that for the past few pages to JJ Wink
It is a debate about the GOAT debate- Inception-y !

It Must Be Love

Posts : 2691
Join date : 2013-08-14

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by JuliusHMarx Tue 16 Sep 2014, 11:21 pm

It Must Be Love wrote:It is a debate about the GOAT debate- Inception-y !

I hope not - that would be even worse.

JuliusHMarx
julius
julius

Posts : 22358
Join date : 2011-07-01
Location : Paisley Park

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by temporary21 Tue 16 Sep 2014, 11:28 pm

Does this happen everytime? W/L stats are clearly skewed by who they play and what tournies they play in. All the really great players have win rates into at least the 70% mark, playing the most competitive tournaments, its pretty hard to say whos better than the other though, they still all played different people and tournies to get their w/l rates.

I would enquire, politely if I may, that why a thread discussing how good Nadal is seemingly soo disliked, and keeps getting heated, when a thread about the greatness of Federer was perfectly OK. If someone wants to put that forward, thats their right , if its not your discussion of choice its best to leave rather than risk it getting sour...

temporary21

Posts : 5092
Join date : 2014-09-07

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by It Must Be Love Tue 16 Sep 2014, 11:37 pm

temporary21 wrote:I would enquire, politely if I may, that why a thread discussing how good Nadal is seemingly soo disliked, and keeps getting heated, when a thread about the greatness of Federer was perfectly OK.
The two threads are inherently different, so not sure if this comparison stands tbh.

It Must Be Love

Posts : 2691
Join date : 2013-08-14

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by temporary21 Tue 16 Sep 2014, 11:39 pm

It Must Be Love wrote:
temporary21 wrote:I would enquire, politely if I may, that why a thread discussing how good Nadal is seemingly soo disliked, and keeps getting heated, when a thread about the greatness of Federer was perfectly OK.
The two threads are inherently different, so not sure if this comparison stands tbh.

Ok, then lets just all cool our heads a little then, no need for things to heat up over such a small thing, maybe if we compared W/L ratios from previous gens, like Petes and Agassis

temporary21

Posts : 5092
Join date : 2014-09-07

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by It Must Be Love Tue 16 Sep 2014, 11:42 pm

temporary21 wrote:
It Must Be Love wrote:
temporary21 wrote:I would enquire, politely if I may, that why a thread discussing how good Nadal is seemingly soo disliked, and keeps getting heated, when a thread about the greatness of Federer was perfectly OK.
The two threads are inherently different, so not sure if this comparison stands tbh.

Ok, then lets just all cool our heads a little then, no need for things to heat up over such a small thing, maybe if we compared W/L ratios from previous gens, like Petes and Agassis
I am perfectly cool, and enjoy these debates Wink
A bit tired today, but will be having a look at some historic W/L ratios tomorrow.

It Must Be Love

Posts : 2691
Join date : 2013-08-14

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by temporary21 Tue 16 Sep 2014, 11:46 pm

Soo Pete apparently only had a career W/L of about 77% his highest being 83% on grass, unsurprisingly,

andre was close on 76%, hes more all rounded, but youd say Pete had the better of the careers.

Feds got 81% and Rafas at 83%, again way too close, and theyre quite obviously pretty hard to pick from. Theyre both astonishing stats compared to the Americans either way.

temporary21

Posts : 5092
Join date : 2014-09-07

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by biugo Wed 17 Sep 2014, 12:20 am

It Must Be Love wrote:
It Must Be Love wrote:
DirectView2 wrote:Nadal has more less picked and chosen which surface he wants to play
I read somewhere Nadal's matches played actually matches the ATP spread overall better than Federer's, but I need to double check that.
OK, good news: my memory was not failing me and that is indeed correct.

If you look at the tournaments for 2013 the total number of hard court tournaments is 37, clay court tournaments is 22, and grass court tournaments is 6.

Tournament Percentages:
Hard - 57%
Clay - 34%
Grass- 9%

Let's look at the total matches on each surface for Federer and Nadal after 2013 (not including carpet):

Federer:
Hard - 681 matches (64%)
Clay - 248 matches (23%)
Grass - 140 matches (13%)

Nadal:
Hard - 402 matches (52%)
Clay - 314 matches (40%)
Grass - 63 matches (8%)

Federer's plays 7% more on hard courts, 11% less on clay courts, and 4% more on grass courts compared to the layout of the tour. That's a total deviation of 22%.

Nadal plays 5% less on hard courts, 6% more on clay courts, and 1% less on grass courts compared to the layout of the tour. That's a total deviation of 12%.

Interesting, so the distribution of Nadal's career is actually closer to the actual layout of the tour. Federer deviates more on every single surface.

Hello IMBL, I will bring back some oil on this fire, but mostly becuase I'm late in the debate (and felt like reacting when I read this post) - however, I don't really understand the points DirectView2 was trying to make.

But anyway, I find these stats are skewed and thus it doesn't help pushing through your argument. Maybe the conclusion is correct, but if the demonstration is false you may lose your support. (however it's efficient tactics in politics, although annoying)

So, here's where it's wrong: boxing

1- you're comparing % of tournaments proposed to % of matches played.
I think it's quite obvious that apart from both stats being about tennis, they don't make much sense in comparison.
It would be fair if you mulitplied the number of each tournament with the respective number of match you could play at each tournament (7 for GS, 4 or 5 or 6? for the rest).
But it would be a bit annoying to do, so it would probably be simpler to check the % spread of tournaments entered by each player.
In way or the other, we would be comparing apples with apples or oranges with oranges.
And a man who would enter the "right" proportion of tournaments would not be seen as planning only to favour his favourite surface even if he won 100% on that surface - which would be fair. (a wonder could enter 24 tournaments, incl. 4 grass, and have 50% of matches played on grass)

but

2- using the total number of tournaments proposed per year can't describe potential of play.
Because several tournaments the same week. (for example, if one glory week had 32 grass tournaments, Grass would suddenly be the most common surface on tour, but a player could only play 4 tournaments on it)
>> if you look at tournaments per week, it gives 55% of weeks proposing a HC, 36% a CC and 9% Grass tournament.

but

3- Players from a certain rank have mandatory tournaments
So you've got 13 mandatory tournaments whose HC/CC/G spread are locked (I'll keep MC in, as it's a 1000, so it's at least more relevant than any 500)
that's 61% HC, 31% CC, 8% G which can't be escaped.
For the remaining minimum 5 tournaments to play, there is no court restriction (15 hc, 11cc, 3g).
So we can add them proportionally and have for the 18 tournaments: HC 59% (10,6 tournaments), CC 33% (5,9), GC 8% (1,5)


and

4- It is finally a skewed stat because it's an arbitary look at the ATP year. It could as fairly be argued that the top ATP season has 5 parts: HC season, CC season, Grass fortnight, HC season, Indoor Season (HC) and that top players are not the kind to play HC then CC then HC then CC etc... so from Doha to Miami they play HC, then CC until RG, then Grass until SW19, then HC til WTF.
This would give us: 68% HC, 23% CC, 9% Grass

As you see, you could easily back up your position on Nadal with stats that make more sense. But as easily, it could be disproved with other stats that make equal sense... It's very easy to be biased either way.

And just for the trouble (and since 2009, to avoid carpet considerations)

Nadal:
HC 54%
CC 37%
Grass 9%

Federer:
HC 66%
CC 24%
Grass 10%

Djokovic
HC 68%
CC 24%
Grass 8%

And it is not a surprise since Nadal has mostly missed HC tournaments.

If I follow my calculations, the average spread bracket for 18 tournaments should be
HC: 59% - 64%
CC: 28% - 33%
Grass: 8% - 8,5%

But avoiding fractionned tournaments (ie not 1,4294 grass tournie played) anything within these brackets would seem normal to me
HC: 55% - 67%
CC: 22% - 39%
Grass: 5,5% - 11%

For a last non related point, I can see an advantage for Clay court specialists though. With 7 tournaments in the HC seasons not overlapping with MS or GS (and specially the 3 tournaments after Wimby) there is an opportunity to get more points out of CC season - whereas there is no HC tournament during the CC season (to help those who can't perform on Clay) and the top players will usually not take part in these tournaments to keep preparing on the right surface..

biugo

Posts : 335
Join date : 2014-08-19

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by dummy_half Wed 17 Sep 2014, 12:41 am

One other reason Nadal has such a good career winning % is that he reached a high level of success early, at age 18 and in only his third full year as a pro. By comparison, all of Federer, Djokovic and Murray took longer to reach the top, so had played more games and with a poorer winning percentage before reaching what could be considered their prime years - Rafa won a slam in only his 5th GS event, compared with Federer's first being his 15th tournament.

There's no doubting Rafa is one of the all time greats overall and almost unarguably the clay court GOAT (a few of Borg's most ardent fans may try to make the case otherwise). 5 slam titles across the other three tournaments is a very good return for someone who was initially considered to be a specialist, but the overwhelming bulk of his career record is based on his performances on clay. It could be argued that this is both the great strength of Nadal's record and also the weakness - away from clay, which makes up about 1/3 of the season, he has rarely been the best player, and quite frequently not even the second best (2010 and the parts of 2013 when he was fit being the exceptions)

dummy_half

Posts : 6341
Join date : 2011-03-11
Age : 52
Location : East Hertfordshire

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by temporary21 Wed 17 Sep 2014, 12:48 am

True but then it all goes roundabout to the idea that the others guys clay cabinets are by extension very bare, and they are therefore, similarly not all rounded enough etc. Thing is without an inarguable objective measure of success, theres no fair way to tell. All you can do is put the stats in a blender and see what comes out.
Problem is, the more you put it through the grinder, the less valuable the assessment is, if it takes you 10 spins to find the stats in a way that suits your argument, thats 9 to 1 when it doesnt suit it, making the stat way less valuable, therefore theres nothing that can be done. Desnt mean people shouldnt spin, and shouldnt be allowed to spin, but it means theres no real need to take it seriously as it cant be seriously, and properly assessed

temporary21

Posts : 5092
Join date : 2014-09-07

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by It Must Be Love Wed 17 Sep 2014, 1:09 am

Biugo that is an excellent and well written counter-post, I will try and address your points in order:

Buigo wrote:1- you're comparing % of tournaments proposed to % of matches played
My reply to this point is very strong- if you notice I've done the calculations regarding tournaments ! In case you missed it, this is what I found:
If you think looking at the number of tournaments they have played on each surface as opposed to number of matches is more relevant, it strengthens my case.

Federer
Hard - 168 tournaments (64%)
Clay - 65 tournaments (25%)
Grass - 31 tournaments (12%)

Nadal
Hard - 101 tournaments (56%)
Clay - 63 tournaments (35%)
Grass - 16 tournaments (9%)

And what is the tour layout? Hard - 57%, Clay - 34%, Grass - 9%

Federer has about the same deviation from the tour and Nadal is practically dead on.


biugo wrote:2- using the total number of tournaments proposed per year can't describe potential of play.

I was trying to count the overlaps, but the schedule for every year kept changing (ie the overlaps were different each time), so it became nearly impossible to calculate.
If it's an consolation your figures: 'if you look at tournaments per week, it gives 55% of weeks proposing a HC, 36% a CC and 9% Grass tournament' are basically identical to the one I used- so not much that would make a difference anyway.

biugo wrote:3/ (Mandatory tournaments): So we can add them proportionally and have for the 18 tournaments: HC 59% (10,6 tournaments), CC 33% (5,9), GC 8% (1,5)
There's no huge difference between the overall spread and what you've said with the figures on mandatory tournaments here-like 2-3%; Nadal is still extremely close to both the overall spread and mandatory spread.

biugo wrote:4/ But avoiding fractionned tournaments (ie not 1,4294 grass tournie played) anything within these brackets would seem normal to me
HC: 55% - 67%
CC: 22% - 39%
Grass: 5,5% - 11%
So according to this (remember not counting overall but the specific time period you chose), Nadal is in the expected frame for CC and grass, and only 1% off in HC. Djokovic is also 1% of HC, but too high; while Federer is 1% below the max limit for HC- it appears that neither of these 3 are making huge efforts to avoid any particular surface.


Last edited by It Must Be Love on Wed 17 Sep 2014, 1:20 am; edited 1 time in total

It Must Be Love

Posts : 2691
Join date : 2013-08-14

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by It Must Be Love Wed 17 Sep 2014, 1:15 am

Must say, very good and well researched post biugo thumbsup

There is another post I would like to raise, and I think it's still relevant now, and it's what I was saying to BS earlier, with some more info:

Born Slippy showed that in the past few years Nadal's W/L record on outdoor hard courts had been 85%, which is actually above his current W/L record. Thus, if he had played more tournaments on outdoor hard (some of which he's missed due to injury) and continued with his normal average in the past few years- then his W/L ratio could have actually been higher.
Of course we don't know this for sure, other factors could come into play, but the fact is it is entirely possible injuries during his prime have actually made his W/L ratio lower (even though they've occurred away from clay).

It Must Be Love

Posts : 2691
Join date : 2013-08-14

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by biugo Wed 17 Sep 2014, 3:01 am

Thanks for your reply IMBL thumbsup and indeed I missed the post where you made the calculation again with the nb of tournaments. When I started the post I was not trying to prove you wrong, but to go through a sound process - more like a pet peeve Smile

I also don't think any top player are trying to avoid particular surface - being top players, they don't need to. However, given the choice, they'll favour the surface where they are even better - and Nadal is obviously better on Clay while Fed and Djoko are obviously better on HC (I don't count grass, as it's almost non existent on the tour unfortunately)
And to defend the wide % bracket of "normal" behaviour, it is not uncommon to see top 10 try their luck at Clay tournaments between AO and IW...

Anyway, it's not the thread topic really.
So on topic, my feeling is that injury time off benefits Rafa in the W/L. It's all speculation of course. But I think having incomplete seasons fits him well, and is positive for his longevity after all. In other words, he breaks down quickly, so the W/L doesn't have much time to fade... And he heals well (not like Andy, ie) so W/L doesn't suffer either when he's back.

And for the more factual part, I'm with Dummy Half on the fact the Nadal performed very early in his career, which helped keep a high W/L.

Nadal has the most amazing W/L ratio but should have done better on crushing the field on a given year (although he did it repeatedly on the Clay season).
I regards Connors as the most impressive W/L of all time:

A: 1253 W / 1531 played - 81,8%
to realize how impressive it is: to achieve this, Nadal should from now on add the career statistics of Djoko (80,8% in 726 matches) to be on par with Jimmy... Alright he's not played the best tournaments all the time, but his GS match win says he was always sharp.

B: year 1974 - W/L 93/4, 100% win in slams... If only he could have played RG... Maybe he'd have stopped Borg?

However the injuries certainly didn't help Nadal with his legacy (#1 spot comes to mind, rather than tournament wins). A good stat for him: his W/L against top 10 players.

And by the way, IMBL, you could check the W/L or H2H ratios weighted with the surface % we talked about - to counter the argument "H2H are so positive because top players cared to showed up in Clay tournaments while Nadal was kicked by 2nd tier players on hard". I think Nadal would still be ahead.

biugo

Posts : 335
Join date : 2014-08-19

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by summerblues Wed 17 Sep 2014, 3:53 am

Very very nice analysis biugo.  I saw this thread earlier today and was going to point out to Amri that his choice of using average across all ATP tournaments was not really the appropriate metric for looking at top players' surface spreads, but I see you already covered it so beautifully.

summerblues

Posts : 4551
Join date : 2012-03-07

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by summerblues Wed 17 Sep 2014, 4:15 am

Amri, regarding W/L ratios:

I do not want to go into a discussion of how useful a measure W/L is, but want to point out one place where your presentation of them introduces an age-bias so to speak (in that you effectively give more weight to success at younger age relative to success at older age).

You say:

It Must Be Love wrote:
IMBL wrote:here is the breakdown:

Nadal:
Accumulatively by aged 22: 81.4%
" By age 23: 81.6% (400/490)
" By age 24: 82.5% (471/571)
" By age 25: 82.5 % (540/655)
" By age 26: 82.8% (582/ 703)
" By age 27: 83.7% (653/780)

Federer's maximum ever has been 82%

Suppose you have two players, who year-by-year have the following W/L ratios (starting at age 20)

Player A: 70%, 90%, 90%, 90%, 90%, 90%, 70%, 70%, 70%, 70%
Player B: 70%, 70%, 70%, 70%, 90%, 90%, 90%, 90%, 90%, 70%

For simplicity's sake, let's assume that each of them plays the same number of matches each year.  Then, your presentation will give:

Player A:
by 20: 70%
by 21: 80%
by 22: 83%
by 23: 85%
by 24: 86%
by 25: 87%
by 26: 84%
by 27: 83%
by 28: 81%
by 29: 80%

Player B:
by 20: 70%
by 21: 70%
by 22: 70%
by 23: 70%
by 24: 74%
by 25: 77%
by 26: 79%
by 27: 80%
by 28: 81%
by 29: 80%

At first blush, player A may look better as they have higher numbers in most of the table (you could e.g., say, player's B maximum has ever been 81% but player's A maximum exceeded that significantly), even though in reality their numbers are equally good - both have 5 years at 90% and 5 years at 70%.

You can really only compare these numbers across the entire career - but even there you run against trouble as not all players have equally long careers.

Another alternative would be to not look at the numbers chronologically, but compare player A's best year vs player B's best year, then player A's best two years vs player B's best two years, and so on.  That would eliminate the "bias" where more weight is effectively given to performances at younger age and would show that player A and player B had, in fact, had identical performances.

summerblues

Posts : 4551
Join date : 2012-03-07

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by kingraf Wed 17 Sep 2014, 4:40 am

A few points from me
1 - let's at least be fair. It's one thing to criticise Nadal's ability to stay #1 (I mean, what's nearly three years at the top), but a complete other to say he also goes large periods where he isn't even the second best. He has 250ish weeks at #2... That's almost double more than the next. In fact only Federer, Connors and Sampras have spent more time guaranteed the 1v2 final.

2 - While its possible being injured so often has prolonged Rafa, it's conjecture to say it's also helped His W/L. His record on Hard court slams since 2009 read, W, SF, Qf, W, SF, F, F, W, F. That's a 53-8 record, which accounts for 7.5% of his all time wins but only 5.8% of his all time losses. So mathematical odds are playing in the three hard court slams he missed would have improved his W/L.

The other thing is how many tournaments did the other greats play on clay in comparison to the spread? I wouldn't be suprised if it wasn't enough to make a dent on their career W/Ls.
kingraf
kingraf
raf
raf

Posts : 16599
Join date : 2012-06-06
Age : 29
Location : To you I am there. To me I am here.... is it possible that I'm everywhere?

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by socal1976 Wed 17 Sep 2014, 6:00 am

I can see how IMBL is working here it is not just the H2H record that favors Nadal over Federer but Nadal's H2H against the rest of the field as well. While Nadal's H2h against fed gets a lot of attention this pertinent statistic which maybe just as telling rarely gets discussed. I think both are pertinent measures but in no way dispositive of who is the better player. It is certainly part of the analysis. Murdoch and others make a good point in that Nadal's win percentage unless he retires early, which is a possibility ala Borg.

socal1976

Posts : 14212
Join date : 2011-03-18
Location : southern california

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by It Must Be Love Wed 17 Sep 2014, 11:12 am

biugo wrote:Thanks for your reply IMBL thumbsup and indeed I missed the post where you made the calculation again with the nb of tournaments. When I started the post I was not trying to prove you wrong, but to go through a sound process - more like a pet peeve Smile
No, you were absolutely right in asking for the tournament calculations too OK

I also don't think any top player are trying to avoid particular surface - being top players, they don't need to. However, given the choice, they'll favour the surface where they are even better - and Nadal is obviously better on Clay while Fed and Djoko are obviously better on HC (I don't count grass, as it's almost non existent on the tour unfortunately)
True, also due to mandatory tournaments it's not players even if they wanted could play 90% of tennis on a certain surface or something.

biugo wrote:
And by the way, IMBL, you could check the W/L or H2H ratios weighted with the surface % we talked about - to counter the argument "H2H are so positive because top players cared to showed up in Clay tournaments while Nadal was kicked by 2nd tier players on hard". I think Nadal would still be ahead.
Don't think people on this forum would be too happy if I started bringing out H2H stats Wink

It Must Be Love

Posts : 2691
Join date : 2013-08-14

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by It Must Be Love Wed 17 Sep 2014, 11:15 am

summerblues wrote:Very very nice analysis biugo.  I saw this thread earlier today and was going to point out to Amri that his choice of using average across all ATP tournaments was not really the appropriate metric for looking at top players' surface spreads, but I see you already covered it so beautifully.
True, true; but before we get too over-excited I did respond fully to the first point (i.e. tournaments instead of matches), and even his spread of mandatories he did was not that much different at all from the spread on the whole ATP tour- so as Nadal is bang on with the ATP tour on the whole, he's still very close to the mandatory spread as well.
If there was a huge difference between the ATP tour on the whole and mandatory tournaments (which there isn't), I would then argue the mandatories should better match the ATP tour.

summerblues wrote:Amri, regarding W/L ratios:

I do not want to go into a discussion of how useful a measure W/L is, but want to point out one place where your presentation of them introduces an age-bias so to speak (in that you effectively give more weight to success at younger age relative to success at older age).
Yes, you raised this point last time, and it was a valid one.
However, I must note that I've not actually paid any year by year comparisons here- so the way or organising the W/L record (which I've done year by year) doesn't actually have that much of an impact at all. If I was going a comparison, between Federer and Nadal (God forbid), then either way even if we don't organise it and just take the lump figure with both aged 28 Nadal is ahead anyway.
This time I've shown the improvement with age to make another point actually, about how it's improved with time- so on the whole missing time in his prime could hurt the overall average. Anyway more on that in a bit.

It Must Be Love

Posts : 2691
Join date : 2013-08-14

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by JuliusHMarx Wed 17 Sep 2014, 11:32 am

It Must Be Love wrote:If I was going a comparison, between Federer and Nadal (God forbid), then either way even if we don't organise it and just take the lump figure with both aged 28 Nadal is ahead anyway.

If you were doing one? Do you mean if you were doing the one that you then did Wink
Luckily, by avoiding comparing Rafa and Fed, we're avoiding the GOAT debate that IMBL wanted to avoid.

JuliusHMarx
julius
julius

Posts : 22358
Join date : 2011-07-01
Location : Paisley Park

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by It Must Be Love Wed 17 Sep 2014, 11:37 am

JuliusHMarx wrote:
It Must Be Love wrote:If I was going a comparison, between Federer and Nadal (God forbid), then either way even if we don't organise it and just take the lump figure with both aged 28 Nadal is ahead anyway.

If you were doing one? Do you mean if you were doing the one that you then did Wink
Luckily, by avoiding comparing Rafa and Fed, we're avoiding the GOAT debate that IMBL wanted to avoid.
The God forbid was put sarcastically; I can make any comparison I want and we can debate it maturely without constant snide comments.
In this instance SB was actually right that I had done a study before on another forum on this, and the way it was arranged meant the player who initially has the advantage keeps that advantage, which could be seen as unfair.
I'm not sure if your comments are meant to come across as snide, but they do.

It Must Be Love

Posts : 2691
Join date : 2013-08-14

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by socal1976 Wed 17 Sep 2014, 11:51 am

Some very interesting posts on this thread. And a pretty civil goat debate as far as that is concerned. I think if Nadal does get to 17 he will be the consensus GOAT. Given his dominance of Federer since his teenage years and fed being the next most likely GOAT candidate then I doubt many except ardent fed fans would disagree. For me I already would place Nadal as the GOAT. But can admit that the case is not clear cut. Yet in time if Fed doesn't add anymore slams and Nadal keeps adding his case gets further strengthened. Weeks at #1 is a good stat but really it doesn't outweigh slam count in any way. I mean Djokovic will probably surpass Nadal in total weeks at #1 but if Nadal has twice as many slams no one, not even I will rate Novak as better.

socal1976

Posts : 14212
Join date : 2011-03-18
Location : southern california

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by JuliusHMarx Wed 17 Sep 2014, 12:04 pm

It's not snide - at least no more than the intention of the thread as I see it.
As JJ said 'This has to be the most complicated way of saying "player x is better than player y because I said so" I've ever seen.'

Here's my theory. The argument you wish to present is that choosing one player as better than another is subjective (unless the stats are wildly different enough to make it effectively objective). Well, let's agree to that then. Ah, but then stats between some players (not mentioning Fed and Rafa if possible) are very close - close enough that saying one player is better than the other is primarily subjective. Hmmm, OK-ish, sounds reasonable, I'll go along with that. Aha - time for the trump card - anyone who says Fed is better than Rafa can't prove it, they're only being subjective. My saying Rafa is better than Fed is equally as justified as anyone saying Fed is better than Rafa. And by agreeing to my previous arguments, all the Fed fans have painted themselves into a logical corner. It's true that I can't 'prove' Rafa is better, but I've tricked the Fed fans into logically agreeing that they can't 'prove' Fed is better. Job done. And all without even bringing Federer into the thread at all or having any sort of GOAT debate (well, not much).

That's honestly how I see it. Have I got that about right? Was that pretty much the intention all along ? I could be wrong, but that is my reading of the situation. Of course, it's all subjective Smile

JuliusHMarx
julius
julius

Posts : 22358
Join date : 2011-07-01
Location : Paisley Park

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by It Must Be Love Wed 17 Sep 2014, 12:33 pm

Julius wrote:The argument you wish to present is that choosing one player as better than another is subjective (unless the stats are wildly different enough to make it effectively objective). Well, let's agree to that then. Ah, but then stats between some players (not mentioning Fed and Rafa if possible) are very close - close enough that saying one player is better than the other is primarily subjective. Hmmm, OK-ish, sounds reasonable, I'll go along with that.
Yes I agree with that, I'm not sure why you're saying 'OK-ish' when you yourself seemed to agree with me earlier:
Julius wrote:The lines between objectivity and subjectivity thus become blurred, or grey. The more evenly balanced the weight of evidence the more subjective it becomes.
So of course there is always a subjective element to these discussions, as you yourself say.

Julius wrote:Aha - time for the trump card - anyone who says Fed is better than Rafa can't prove it, they're only being subjective.
No, no; this is not a trump card -it's obvious. Just like the same would apply to me if I were to argue Borg, Laver, Nadal etc. are the best.

Julius wrote:It's true that I can't 'prove' Rafa is better, but I've tricked the Fed fans into logically agreeing that they can't 'prove' Fed is better.
No, because the word 'tricked' seems to indicate that I am trying to spread misinformation- when in reality what I'm saying is obvious and you yourself wrote a post basically saying what I was saying.

Honestly Julius, I'm not trying to 'trick' anyone; the reason I argue the case I do is because I genuinely believe I am right in what I say.

Julius wrote:Was that pretty much the intention all along ?
See this is where your suspicions get ridiculous, I didn't mention anything about this particular debate (objectivity vs subjectivity) in the first few posts of this thread; HM first raised it, I said I fully agreed; and then JJ wrote a counter-point. I don't have a crystal ball to know this was going to happen.

Julius wrote:It's not snide - at least no more than the intention
OK, I see now, I thought you were just continuously being snide for no reason; when in reality you had come up with an inaccurate theory.


Also, on a last note, to make it clear; what I'm arguing applies to me too. If I say for example that I've proved Nadal is better than Borg/Laver etc.; because I seee him as statistically superior- a valid counter point could be that due to the different variables my claim is subjective- and I should look into the variables as well as the stats to form a more well-backed opinion.

It Must Be Love

Posts : 2691
Join date : 2013-08-14

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by temporary21 Wed 17 Sep 2014, 12:40 pm

Theres no proof of any intention soo lets keep it on topic.

Of course the F/N's stats are gonna be compared but thats on topic from the thread, lets not let this derail the thread


soo... back on track! Did anyone get more stuff on the W/L ratios of past big greats?


Last edited by temporary21 on Wed 17 Sep 2014, 1:30 pm; edited 1 time in total (Reason for editing : Edited by the user)

temporary21

Posts : 5092
Join date : 2014-09-07

Back to top Go down

Discussion about W/L ratio  - Page 3 Empty Re: Discussion about W/L ratio

Post by Sponsored content


Sponsored content


Back to top Go down

Page 3 of 6 Previous  1, 2, 3, 4, 5, 6  Next

Back to top

- Similar topics

 
Permissions in this forum:
You cannot reply to topics in this forum