REVISED: Updated Open era achievment ranking system using current ATP weighting

[helterskelter]

I like what timnz is doing here but your questions are good ones. Why stop at finals, or at 500s? But once you start adding tournaments and including more than just reaching a final then you really are just replicating the ATP system. In that case might as well just use that.

You *should* just replicate the ATP system. If they did indeed aggregate tallies over years, the best thing to do would be use their tally. But they don't do it, so the goal should be to do it for them.

Although, as I said in my reply to timnz just now, there isn't going to be an unobjectionable system.

 

[timnz]

You *should* just replicate the ATP system. If they did indeed aggregate tallies over years, the best thing to do would be use their tally. But they don't do it, so the goal should be to do it for them.

Although, as I said in my reply to timnz just now, there isn't going to be an unobjectionable system.

I am comparing Open era greats. 250's were a lot easier to win for Greats in the 70's and 80's - hence, I have removed them to even the playing field. I did initially not include anything below 1000 points, but became aware that that was unfair to the older players who had many tournament wins in the 500 to 1000 range (and that could have even been rated as equivalent to 1000's). Hence, I decide the best compromise was to include 500's and up but no lower.

 

[timnz]

That makes sense, but I just don't think you can make a very close comparison between the 70s/80s and today whatever you do. It's also clear that comparing Masters 1000s and even Slams isn't comparing like with like. In the 70s and (early) 80s, they clearly took the Masters more seriously than the Australian Open. So, should you reverse points and make the Masters worth a Slam? I don't know.

I just don't think that everything is quantifiable. I know it's a widespread goal in our culture to make everything quantifiable, but I think it's a goal that must remain an unrealized dream.

I have been refining this ranking system on these boards for the last 5 years or so (there were earlier threads). It was only recently that I included events lower than 1000. This is my best attempt to get some sort of parity. The hardest part was 500 to 1000 events.

I would add that, even if you feel that it is impossible to compare players over those decades - the fact is people do. People are always talking about how Connors and Lendl won 8 slams and Djokovic has won 10. What I have tried to do, is that the conversation away from the 'just slams count' view - to try to show events that were important in the earlier period like the WCT finals. That is how I cover your question about the Australian Open by the way. Read the 1st thread in this series for an explanation on how that is done (Basically I count WCT finals and Grand Slam Cup achievements ONLY if the player didn't play all the slams that year).

For some example of why I don't include the 250 events. Connors has won 32, Nadal has won 9. Hence, it shows the relative ease of winning them in Connors era.

 

[helterskelter]

I have been refining this ranking system on these boards for the last 5 years or so (there were earlier threads). It was only recently that I included events lower than 1000. This is my best attempt to get some sort of parity. The hardest part was 500 to 1000 events.

I would add that, even if you feel that it is impossible to compare players over those decades - the fact is people do. People are always talking about how Connors and Lendl won 8 slams and Djokovic has won 10. What I have tried to do, is that the conversation away from the 'just slams count' view - to try to show events that were important in the earlier period like the WCT finals. That is how I cover your question about the Australian Open by the way. Read the 1st thread in this series for an explanation on how that is done (Basically I count WCT finals and Grand Slam Cup achievements ONLY if the player didn't play all the slams that year).

For some example of why I don't include the 250 events. Connors has won 32, Nadal has won 9. Hence, it shows the relative ease of winning them in Connors era.

I will read the first thread when I get a chance.

When I say that things can't be quantified, I mean only that whatever system you use will involve making a non-quantifiable decision about what to count and what not to count. I'm sure your reasoning is perfectly reasonable. I'm also sure that there are reasonable objections to it. That's just how it goes.

I know people compare across decades. I know that they will continue to do so. I know that those comparisons will forever be disputed. Trying to end the disputes is a laudable goal. But you won't end them. There will always be people who object and propose a different method of calculation.

 

[timnz]

I will read the first thread when I get a chance.

When I say that things can't be quantified, I mean only that whatever system you use will involve making a non-quantifiable decision about what to count and what not to count. I'm sure your reasoning is perfectly reasonable. I'm also sure that there are reasonable objections to it. That's just how it goes.

I know people compare across decades. I know that they will continue to do so. I know that those comparisons will forever be disputed. Trying to end the disputes is a laudable goal. But you won't end them. There will always be people who object and propose a different method of calculation.

For your convenience - here is the content of the first thread:

NOTE: After reflection in the last few months, I have decided to change the Open era achievement ranking system that I have been posted for a number of years. The reasons are principally around the Masters 1000 achievements and their equivalents. When I found out that Lendl only competed in 10 of 27 Masters 1000's in 1990-1992 (because they were not compulsory then), and that there were 8-9 other tournaments per year from 1990-1992 that were equivalent in points and prize money to the 'official' Masters 1000 - that made me wonder if our count of these events was fair.The problem that I have wrestled with mostly is this:

The earlier events didn't have the depth of top players that today's do (which is harder on more recent players) and, because of it not being compulsory, and there were other comparible events - (then it is unfair on the older players who didn't compete in what we deem Masters 1000 equivalents). So how do we work out equivalency given that there were completely different contexts for the playing of these events pre-2000 (particularly before 1993)?

Masters 1000 pre-1990 are difficult to agree on. There is no agreed 'Masters 1000' equivalent list. The only list that I have seen some agreement on, in these forums, is:

http://en.wikipedia.org/wiki/Tennis_Masters_Series_records_and_statistics

I have decided that we should therefore include all players 500 level event victories. Most of the 'defacto' Masters 1000's have been rated at the 500 level. Therefore in a ranking system, they don't give as many points, but at least they get represented somewhat. I have therefore changed the cut-off to 500 points and above, where previously I had 1000 points and above per event.

Details of the system:

Everything that in today's terms you can earn 500 points and above per event is counted. That is:

• Slam Victories (SV) 2000 ATP points
• Slam Runner-ups (SRU) 1200 ATP points
• Slam Semi-finals (SSF) 720 ATP points
• Season end final victories with no loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFNL) 1500 ATP points
• Season end final victories with one loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFOL) 1300 ATP points
• Season end final runner-ups with no loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFRUNL) 1000 ATP points
• Season end final runner-ups with one loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFRUOL) 800 ATP points
• Season end final semi-finals with no loss before the semi-final (WTF, WCT Finals * & Grand Slam Cup *) (SEFSFNL) ATP 600 points
• Masters 1000 equivalent victories (we will call (Top 9)) ATP 1000 points
• Masters 1000 equivalent runner-ups (TOP9RU) ATP 600 points
• Olympic Gold Metal Singles (OSG) ATP 750 points
• 500 Series equivalents (500S) ATP 500 points

* Note: To even out the fact that pre-mid 1980's great players tended to play 3 slams a year vs 4 slams a year for current players, I will only include WCT Finals and Grand Slam Cup placings if the player didn't play all the slams in that year eg I include Lendl's 1982 WCT Finals win but I don't include his 1985 WCT Finals win, since in 1982 he didn't play all the slams but in 1985 he did. That way it is fair to modern players that the older players aren't getting an extra event to score points in (since modern players don't have the WCT Finals or Grand Slam Cup to count).

REMEMBER: There is no agreed weighting of events. In this forum I have tried to get an agreed weightings but opinions as to the weighting vary greatly. The best I can do is use the current ATP weightings. Everytime I post these rankings people disagree with the weightings, but what can I do? - there is no agreed standard beyond the ATP weightings. Also note that this table doesn't represent 'Greatness' which is a subjective term. It simply represents an objective list of the achievements of open era players weighted at current ATP weightings.

For ease I have reduced the weighting points down by a factor of 1000 eg Slams are worth 2 instead of their ATP 2000.

Scale is: (SV x 2) + (SEFNL x 1.5) + (SEFOL x 1.3) + (SEFRUNL x 1) + (SRU x 1.2) + (TOP9 x 1) + (TOP9RU x 0.60) + (SEFRUOL x 0.80) + (OSG x 0.75) + (SSF x 0.72) + (SEFSFNL x 0.60) + (500S x 0.50)

Federer = (17 x 2) + (5 x 1.5) + (1 x 1.3) + (2 x 1) + (10 x 1.2) + (24 x 1) + (18 x 0.60) + (1 x 0.80) + (0 x 0.75) + (11 x 0.72) + (1 x 0.60) + (16 x 0.50) = 108.92

Lendl = (8 x 2) + ((5 + 2 - 1) x 1.5)) + (0 x 1.3) + (2 x 1) + (11 x 1.2) + (22 x 1) + (11 x 0.60) + (2 x 0.80) + (0 x 0.75) + (9 x 0.72) + ((5 -2) x 0.60) + (42 x 0.50) = 99.48

Connors = (8 x 2) + (2 x 1.5) + (1 x 1.3) + (1 x 1) + (7 x 1.2) + (17 x 1) + (9 x 0.60) + (0 x 0.80) + (0 x 0.75) + (16 x 0.72) + (4 x 0.60) + (49 x 0.50) = 90.52

Nadal = (14 x 2) + (0 x 1.5) + (0 x 1.3) + (2 x 1) + (6 x 1.2) + (27 x 1) + (14 x 0.60) + (0 x 0.80) + (1 x 0.75) + (3 x 0.72) + (0 x 0.60) + (16 x 0.50) = 83.51

Djokovic = (10 x 2) + (3 x 1.5) + (1 x 1.3) + (0 x 1) + (8 x 1.2) + (25 x 1) + (12 x 0.60) + (0 x 0.80) + (0 x 0.75) + (10 x 0.72) + (0 x 0.60) + (12 x 0.50) = 80.8

McEnroe = (7 x 2) + ((3 + (5 - 1)) x 1.5)) + (0 x 1.3) + ((1 + 3) x 1) + (4 x 1.2) + (19 x 1) + (7 x 0.60) + (0 x 0.80) + (0 x 0.75) + (8 x 0.72) + (0 x 0.60) + (23 x 0.50) = 73.76

Sampras = (14 x 2) + (0 + (2 - 1) x 1.5) + (5 x 1.3) + ((2 - 1) x 1) + (4 x 1.2) + (11 x 1) + (8 x 0.60) + (0 x 0.80) + (0 x 0.75) + (5 x 0.72) + ((4 - 1) x 0.60) + (12 x 0.50) = 69.0

Borg = (11 x 2) + ((1 + 1) x 1.5)) + (1 x 1.3) + ((1 + (3 - 1)) x 1) + (5 x 1.2) + (15 x 1) + (4 x 0.60) + (1 x 0.80) + (0 x 0.75) + (1 x 0.72) + (1 x 0.60) + (17 x 0.50) = 63.32

Agassi = (8 x 2) + (0 x 1.5) + (1 x 1.3) + ((3 - 1) x 1) + (7 x 1.2) + (17 x 1) + (5 x 0.60) + (1 x 0.80) + (1 x 0.75) + (11 x 0.72) + (1 x 0.60) + (6 x 0.50) = 61.77

Becker = (6 x 2) + ((1 + 1) x 1.5)) + (3 x 1.3) + ((4 + 1) x 1) + (4 x 1.2) + (13 x 1) + (8 x 0.60) + (0 x 0.80) + (0 x 0.75) + (8 x 0.72) + ((2 - 1) x 0.60) + (9 x 0.50)= 57.36

Edberg = (6 x 2) + (0 x 1.5) + (1 x 1.3) + ((2 - 1) x 1) + (5 x 1.2) + (8 x 1) + (12 x 0.60) + (0 x 0.80) + (0 x 0.75) + (8 x 0.72) + ((2 - 1) x 0.60) + (8 x 0.50) = 45.86

Wilander = (7 x 2) + (0 x 1.5) + (0 x 1.3) + (0 x 1) + (4 x 1.2) + (8 x 1) + (7 x 0.60) + (1 x 0.80) + (0 x 0.75) + (3 x 0.72) + (0 x 0.60) + (8 x 0.50****) = 37.96

**** I found it difficult to determine that Wilander's 500 level equivalents are. Depending upon approaches, I ended up with anything from 5 to 10. I have settled (for now) on 8.

 

[uscwang]

For your convenience - here is the content of the first thread:

NOTE: After reflection in the last few months, I have decided to change the Open era achievement ranking system that I have been posted for a number of years. The reasons are principally around the Masters 1000 achievements and their equivalents. When I found out that Lendl only competed in 10 of 27 Masters 1000's in 1990-1992 (because they were not compulsory then), and that there were 8-9 other tournaments per year from 1990-1992 that were equivalent in points and prize money to the 'official' Masters 1000 - that made me wonder if our count of these events was fair.The problem that I have wrestled with mostly is this:

The earlier events didn't have the depth of top players that today's do (which is harder on more recent players) and, because of it not being compulsory, and there were other comparible events - (then it is unfair on the older players who didn't compete in what we deem Masters 1000 equivalents). So how do we work out equivalency given that there were completely different contexts for the playing of these events pre-2000 (particularly before 1993)?

Masters 1000 pre-1990 are difficult to agree on. There is no agreed 'Masters 1000' equivalent list. The only list that I have seen some agreement on, in these forums, is:

http://en.wikipedia.org/wiki/Tennis_Masters_Series_records_and_statistics

I have decided that we should therefore include all players 500 level event victories. Most of the 'defacto' Masters 1000's have been rated at the 500 level. Therefore in a ranking system, they don't give as many points, but at least they get represented somewhat. I have therefore changed the cut-off to 500 points and above, where previously I had 1000 points and above per event.

Details of the system:

Everything that in today's terms you can earn 500 points and above per event is counted. That is:

• 
  
  
  • Slam Victories (SV) 2000 ATP points
  • Slam Runner-ups (SRU) 1200 ATP points
  • Slam Semi-finals (SSF) 720 ATP points
  • Season end final victories with no loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFNL) 1500 ATP points
  • Season end final victories with one loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFOL) 1300 ATP points
  • Season end final runner-ups with no loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFRUNL) 1000 ATP points
  • Season end final runner-ups with one loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFRUOL) 800 ATP points
  • Season end final semi-finals with no loss before the semi-final (WTF, WCT Finals * & Grand Slam Cup *) (SEFSFNL) ATP 600 points
  • Masters 1000 equivalent victories (we will call (Top 9)) ATP 1000 points
  • Masters 1000 equivalent runner-ups (TOP9RU) ATP 600 points
  • Olympic Gold Metal Singles (OSG) ATP 750 points
  • 500 Series equivalents (500S) ATP 500 points

* Note: To even out the fact that pre-mid 1980's great players tended to play 3 slams a year vs 4 slams a year for current players, I will only include WCT Finals and Grand Slam Cup placings if the player didn't play all the slams in that year eg I include Lendl's 1982 WCT Finals win but I don't include his 1985 WCT Finals win, since in 1982 he didn't play all the slams but in 1985 he did. That way it is fair to modern players that the older players aren't getting an extra event to score points in (since modern players don't have the WCT Finals or Grand Slam Cup to count).

REMEMBER: There is no agreed weighting of events. In this forum I have tried to get an agreed weightings but opinions as to the weighting vary greatly. The best I can do is use the current ATP weightings. Everytime I post these rankings people disagree with the weightings, but what can I do? - there is no agreed standard beyond the ATP weightings. Also note that this table doesn't represent 'Greatness' which is a subjective term. It simply represents an objective list of the achievements of open era players weighted at current ATP weightings.

For ease I have reduced the weighting points down by a factor of 1000 eg Slams are worth 2 instead of their ATP 2000.

Scale is: (SV x 2) + (SEFNL x 1.5) + (SEFOL x 1.3) + (SEFRUNL x 1) + (SRU x 1.2) + (TOP9 x 1) + (TOP9RU x 0.60) + (SEFRUOL x 0.80) + (OSG x 0.75) + (SSF x 0.72) + (SEFSFNL x 0.60) + (500S x 0.50)

Federer = (17 x 2) + (5 x 1.5) + (1 x 1.3) + (2 x 1) + (10 x 1.2) + (24 x 1) + (18 x 0.60) + (1 x 0.80) + (0 x 0.75) + (11 x 0.72) + (1 x 0.60) + (16 x 0.50) = 108.92

Lendl = (8 x 2) + ((5 + 2 - 1) x 1.5)) + (0 x 1.3) + (2 x 1) + (11 x 1.2) + (22 x 1) + (11 x 0.60) + (2 x 0.80) + (0 x 0.75) + (9 x 0.72) + ((5 -2) x 0.60) + (42 x 0.50) = 99.48

Connors = (8 x 2) + (2 x 1.5) + (1 x 1.3) + (1 x 1) + (7 x 1.2) + (17 x 1) + (9 x 0.60) + (0 x 0.80) + (0 x 0.75) + (16 x 0.72) + (4 x 0.60) + (49 x 0.50) = 90.52

Nadal = (14 x 2) + (0 x 1.5) + (0 x 1.3) + (2 x 1) + (6 x 1.2) + (27 x 1) + (14 x 0.60) + (0 x 0.80) + (1 x 0.75) + (3 x 0.72) + (0 x 0.60) + (16 x 0.50) = 83.51

Djokovic = (10 x 2) + (3 x 1.5) + (1 x 1.3) + (0 x 1) + (8 x 1.2) + (25 x 1) + (12 x 0.60) + (0 x 0.80) + (0 x 0.75) + (10 x 0.72) + (0 x 0.60) + (12 x 0.50) = 80.8

McEnroe = (7 x 2) + ((3 + (5 - 1)) x 1.5)) + (0 x 1.3) + ((1 + 3) x 1) + (4 x 1.2) + (19 x 1) + (7 x 0.60) + (0 x 0.80) + (0 x 0.75) + (8 x 0.72) + (0 x 0.60) + (23 x 0.50) = 73.76

Sampras = (14 x 2) + (0 + (2 - 1) x 1.5) + (5 x 1.3) + ((2 - 1) x 1) + (4 x 1.2) + (11 x 1) + (8 x 0.60) + (0 x 0.80) + (0 x 0.75) + (5 x 0.72) + ((4 - 1) x 0.60) + (12 x 0.50) = 69.0

Borg = (11 x 2) + ((1 + 1) x 1.5)) + (1 x 1.3) + ((1 + (3 - 1)) x 1) + (5 x 1.2) + (15 x 1) + (4 x 0.60) + (1 x 0.80) + (0 x 0.75) + (1 x 0.72) + (1 x 0.60) + (17 x 0.50) = 63.32

Agassi = (8 x 2) + (0 x 1.5) + (1 x 1.3) + ((3 - 1) x 1) + (7 x 1.2) + (17 x 1) + (5 x 0.60) + (1 x 0.80) + (1 x 0.75) + (11 x 0.72) + (1 x 0.60) + (6 x 0.50) = 61.77

Becker = (6 x 2) + ((1 + 1) x 1.5)) + (3 x 1.3) + ((4 + 1) x 1) + (4 x 1.2) + (13 x 1) + (8 x 0.60) + (0 x 0.80) + (0 x 0.75) + (8 x 0.72) + ((2 - 1) x 0.60) + (9 x 0.50)= 57.36

Edberg = (6 x 2) + (0 x 1.5) + (1 x 1.3) + ((2 - 1) x 1) + (5 x 1.2) + (8 x 1) + (12 x 0.60) + (0 x 0.80) + (0 x 0.75) + (8 x 0.72) + ((2 - 1) x 0.60) + (8 x 0.50) = 45.86

Wilander = (7 x 2) + (0 x 1.5) + (0 x 1.3) + (0 x 1) + (4 x 1.2) + (8 x 1) + (7 x 0.60) + (1 x 0.80) + (0 x 0.75) + (3 x 0.72) + (0 x 0.60) + (8 x 0.50****) = 37.96

**** I found it difficult to determine that Wilander's 500 level equivalents are. Depending upon approaches, I ended up with anything from 5 to 10. I have settled (for now) on 8.

I'm thinking maybe you can make it into a table, which could facilitate comparison.

 

[timnz]

I'm thinking maybe you can make it into a table, which could facilitate comparison.

I don't know how to do tables in this forum. Also not sure where you aren't seeing the comparison - there are 12 players listed there. Perhaps you mean it would be clearer ?

 

[timnz]

Federer moving up.

Question - at what level would Federer's total be unassailable for Nadal and Djokovic?

Scale is: (SV x 2) + (SEFNL x 1.5) + (SEFOL x 1.3) + (SEFRUNL x 1) + (SRU x 1.2) + (TOP9 x 1) + (TOP9RU x 0.60) + (SEFRUOL x 0.80) + (OSG x 0.75) + (SSF x 0.72) + (SEFSFNL x 0.60) + (500S x 0.50)

Federer = (17 x 2) + (5 x 1.5) + (1 x 1.3) + (2 x 1) + (10 x 1.2) + (24 x 1) + (18 x 0.60) + (1 x 0.80) + (0 x 0.75) + (11 x 0.72) + (1 x 0.60) + (17 x 0.50) = 109.42

Nadal = (14 x 2) + (0 x 1.5) + (0 x 1.3) + (2 x 1) + (6 x 1.2) + (27 x 1) + (14 x 0.60) + (0 x 0.80) + (1 x 0.75) + (3 x 0.72) + (0 x 0.60) + (16 x 0.50) = 83.51

Djokovic = (10 x 2) + (3 x 1.5) + (1 x 1.3) + (0 x 1) + (8 x 1.2) + (25 x 1) + (12 x 0.60) + (0 x 0.80) + (0 x 0.75) + (10 x 0.72) + (0 x 0.60) + (12 x 0.50) = 80.8

 

[maruzo]

Great effort OP. Looks like a pretty spot on system for tallying up player accomplishments. I didn't realize Nadal is that far behind Federer.

I assume time will tell, OP.

 

[uscwang]

Federer moving up.

Question - at what level would Federer's total be unassailable for Nadal and Djokovic?

Novak is 28.6 pts behind Fed in your calculation. He has earned 13.9 pts so far this year, with 2.5 pts left.
So if Fed retires today, it would take 2~3 years for Novak to catch up with him. Don't think Fed can retire with peace of mind anytime soon in this regard.

 

[timnz]

Novak is 28.6 pts behind Fed in your calculation. He has earned 13.9 pts so far this year, with 2.5 pts left.
So if Fed retires today, it would take 2~3 years for Novak to catch up with him. Don't think Fed can retire with peace of mind anytime soon in this regard.

I remember when I started to do these rankings 4 or so years ago (in earlier threads) that Djokovic was in the Becker/Edberg range. He has certainly jumped up massively. Now he is above McEnroe, Agassi etc.

 

[PeterHo]

How meaningful is this total?

The obvious problem with this system, is players get rewarded for ultra LONG careers, steady long term performance (getting into finals/semis) instead of measuring peak to peak performance.

The longer a player plays, and the more tournies they enter, the higher their score.

Borg is very low down the list, even below Mcneroe, for that matter.

It would be interesting to count players peak years, take for example, compare points for all players from their best 5 years. Or take the lower common denominator, taking the no. Of years from the player with the shortest career.

In essense, this list represent players with most and biggest titles won/played deep. It is an enhancement to the record that Connors currently hold~most titles won.

 

[xFedal]

Why are atp 500 counted? why is Novak so below lendl and conners?

 

[timnz]

How meaningful is this total?

The obvious problem with this system, is players get rewarded for ultra LONG careers, steady long term performance (getting into finals/semis) instead of measuring peak to peak performance.

The longer a player plays, and the more tournies they enter, the higher their score.

Borg is very low down the list, even below Mcneroe, for that matter.

It would be interesting to count players peak years, take for example, compare points for all players from their best 5 years. Or take the lower common denominator, taking the no. Of years from the player with the shortest career.

In essense, this list represent players with most and biggest titles won/played deep. It is an enhancement to the record that Connors currently hold~most titles won.

Yes you could have a peak period comparison - but how do determine how long that is? Some players have a great 2 years, others a great 5 years. Depending on what window you chose you would end up with entirely different results. Nothing wrong with trying it though.
My system is what it says it is ie a listing of overall player achievement.

 

[timnz]

Why are atp 500 counted? why is Novak so below lendl and conners?

See the first post in this thread as to explanation of why 500s are included. As to why Lendl and Connors are ahead - see the entries against team. Again see first post in this thread.

See

http://tt.tennis-warehouse.com/inde...ng-system-using-current-atp-weighting.539099/

 

[timnz]

The gap between Nadal and Djokovic has become small.

Thoughts?

For ease I have reduced the weighting points down by a factor of 1000 eg Slams are worth 2 instead of their ATP 2000.

Scale is: (SV x 2) + (SEFNL x 1.5) + (SEFOL x 1.3) + (SEFRUNL x 1) + (SRU x 1.2) + (TOP9 x 1) + (TOP9RU x 0.60) + (SEFRUOL x 0.80) + (OSG x 0.75) + (SSF x 0.72) + (SEFSFNL x 0.60) + (500S x 0.50)

Nadal = (14 x 2) + (0 x 1.5) + (0 x 1.3) + (2 x 1) + (6 x 1.2) + (27 x 1) + (14 x 0.60) + (0 x 0.80) + (1 x 0.75) + (3 x 0.72) + (0 x 0.60) + (16 x 0.50) = 83.51

Djokovic = (10 x 2) + (3 x 1.5) + (1 x 1.3) + (0 x 1) + (8 x 1.2) + (26 x 1) + (12 x 0.60) + (0 x 0.80) + (0 x 0.75) + (10 x 0.72) + (0 x 0.60) + (12 x 0.50) = 81.8

• Slam Victories (SV) 2000 ATP points
• Slam Runner-ups (SRU) 1200 ATP points
• Slam Semi-finals (SSF) 720 ATP points
• Season end final victories with no loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFNL) 1500 ATP points
• Season end final victories with one loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFOL) 1300 ATP points
• Season end final runner-ups with no loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFRUNL) 1000 ATP points
• Season end final runner-ups with one loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFRUOL) 800 ATP points
• Season end final semi-finals with no loss before the semi-final (WTF, WCT Finals * & Grand Slam Cup *) (SEFSFNL) ATP 600 points
• Masters 1000 equivalent victories (we will call (Top 9)) ATP 1000 points
• Masters 1000 equivalent runner-ups (TOP9RU) ATP 600 points
• Olympic Gold Metal Singles (OSG) ATP 750 points
• 500 Series equivalents (500S) ATP 500 points

 

[Djokovic2011]

^^^ So if Novak wins the WTF undefeated, he'll go ahead of Nadal?

 

[timnz]

^^^ So if Novak wins the WTF undefeated, he'll go ahead of Nadal?

No he won' but it will be close. If Nadal doesn't score over 500 points and Djokovic goes unbeaten Nadal will still lead by 83.52 to 83.3. However that is assuming Nadal won't win any significant points at the wtf. I predict Nadal will make at least the semi finals

 

[Djokovic2011]

No he won' but it will be close. If Nadal doesn't score over 500 points and Djokovic goes Nadal will still lead by 83.52 to 83.3. However that is assuming Nadal won't win any significant points at the wtf. I predict Nadal will make at least the semi finals

Oh well, I can wait a few more months.

 

[SpicyCurry1990]

No he won' but it will be close. If Nadal doesn't score over 500 points and Djokovic goes Nadal will still lead by 83.52 to 83.3. However that is assuming Nadal won't win any significant points at the wtf. I predict Nadal will make at least the semi finals

He can still easily make the SF and not score points in your system.
The most likely scenario for him is to go 2-1 in groups with a loss to Fed/Djok and then lose the SF to the other. That would be 400 points.

Of course there is a chance Djok/Fed wind up in the same group and Nadal in the other though (this goes away if Murray elects to skip the event for Davis Cup as then Fed moves up to 2).

 

[Krish872007]

He can still easily make the SF and not score points in your system.
The most likely scenario for him is to go 2-1 in groups with a loss to Fed/Djok and then lose the SF to the other. That would be 400 points.

Of course there is a chance Djok/Fed wind up in the same group and Nadal in the other though (this goes away if Murray elects to skip the event for Davis Cup as then Fed moves up to 2).

Hey man, how've you been? Long time no see on these boards.

 

[mika1979]

No he won' but it will be close. If Nadal doesn't score over 500 points and Djokovic goes unbeaten Nadal will still lead by 83.52 to 83.3. However that is assuming Nadal won't win any significant points at the wtf. I predict Nadal will make at least the semi finals

I love your work bro. The time and effort is truly appreciated. Just do yourself a favour and forget about replying to every crazy person on here. This study seems pretty objective and you can't argue with your efforts to tweak things. Just dont worry about changing things up too much to suit the needs of the fan bases. I think that even though they dont understand it, people critical of these stats, want these numbers based on title wins only. Maybe a gs final is not worth as much as what atp points suggest. To me i think that what you have is fine as clearly there isnt much between rafa and novak but i think that some of the people have a problem with this

 

[timnz]

I love your work bro. The time and effort is truly appreciated. Just do yourself a favour and forget about replying to every crazy person on here. This study seems pretty objective and you can't argue with your efforts to tweak things. Just dont worry about changing things up too much to suit the needs of the fan bases. I think that even though they dont understand it, people critical of these stats, want these numbers based on title wins only. Maybe a gs final is not worth as much as what atp points suggest. To me i think that what you have is fine as clearly there isnt much between rafa and novak but i think that some of the people have a problem with this

You get it! My point is to just objectively write down what each player has achieved at 500 points or more. There isn't much to argue about - it is what it is. I haven't missed out any event where they could have won 500 points or more. The controversy comes in the weighting. But, as I have said, many times - there is no consensus agreement on what acheivements should weigh. (Believe me I have tried on these boards). Hence, the only thing I can do is use the ATP Weighting. That is why the title of this thread is 'Updated Open era achievment ranking system using current ATP weighting' - and it is undisputable that that is what it is. Anybody else can create another system called 'Updated Open era achievment ranking system using XYZ system of weighting'. However, good luck in getting agreement on XYZ weighting.

Thanks for your comments.

 

[xFedal]

You get it! My point is to just objectively write down what each player has achieved at 500 points or more. There isn't much to argue about - it is what it is. I haven't missed out any event where they could have won 500 points or more. The controversy comes in the weighting. But, as I have said, many times - there is no consensus agreement on what acheivements should weigh. (Believe me I have tried on these boards). Hence, the only thing I can do is use the ATP Weighting. That is why the title of this thread is 'Updated Open era achievment ranking system using current ATP weighting' - and it is undisputable that that is what it is. Anybody else can create another system called 'Updated Open era achievment ranking system using XYZ system of weighting'. However, good luck in getting agreement on XYZ weighting.

Thanks for your comments.

Is Djokovic most impressive for his age? 81.8?

 

[TMF]

timnz,
I thought you were going to remove the ATP points earned from Olympic because they are not going to award any point in 2016. This does make sense since the Olympic is never tie-in with the ATP and ITF for over 40 straight years when the ATP was established in 1972 to begin with. Since the open era, only 7 Olympic tennis events were held, which means many professional tennis players never had a chance to compete. And now that the Olympic has decided not to award any points, I think it's even more skewed to include the Olympic if we are going to have such method to determine the player's career achievement and where they ranked on the list.

Other sports like the NBA don't take Olympic performance/results or any stats to compare the player's achievement and where they rank them. It's purely coming from the regular season and post season. Career stats/numbers are not effected by Olympic either.

 

[timnz]

timnz,
I thought you were going to remove the ATP points earned from Olympic because they are not going to award any point in 2016. This does make sense since the Olympic is never tie-in with the ATP and ITF for over 40 straight years when the ATP was established in 1972 to begin with. Since the open era, only 7 Olympic tennis events were held, which means many professional tennis players never had a chance to compete. And now that the Olympic has decided not to award any points, I think it's even more skewed to include the Olympic if we are going to have such method to determine the player's career achievement and where they ranked on the list.

Other sports like the NBA don't take Olympic performance/results or any stats to compare the player's achievement and where they rank them. It's purely coming from the regular season and post season. Career stats/numbers are not effected by Olympic either.

Hi TMF - I am still reflecting on what I will do, but probably, come next August once the Olympics are finished, I will still list it but have it like (1 x 0) - so people can see it is there but it adds no points. Anyhow, like I say, it doesn't happen until next August, so the 750 points stays until then.

Appreciate your post

 

[xFedal]

Hi TMF - I am still reflecting on what I will do, but probably, come next August once the Olympics are finished, I will still list it but have it like (1 x 0) - so people can see it is there but it adds no points. Anyhow, like I say, it doesn't happen until next August, so the 750 points stays until then.

Appreciate your post

Hey Timnz is Novak points most impressive for his age? 81.8?

 

[timnz]

Hey Timnz is Novak points most impressive for his age? 81.8?

It is impressive, but at 28.5 years old probably Federer, Lendl and perhaps Connors were higher.

 

[timnz]

Movers after the 2015 WTF (changes highlighted in Red):-

Everything that in today's terms you can earn 500 points and above per event is counted. That is:

For ease I have reduced the weighting points down by a factor of 1000 eg Slams are worth 2 instead of their ATP 2000.

Scale is: (SV x 2) + (SEFNL x 1.5) + (SEFOL x 1.3) + (SEFRUNL x 1) + (SRU x 1.2) + (TOP9 x 1) + (TOP9RU x 0.60) + (SEFRUOL x 0.80) + (OSG x 0.75) + (SSF x 0.72) + (SEFSFNL x 0.60) + (500S x 0.50)

Federer = (17 x 2) + (5 x 1.5) + (1 x 1.3) + (3 x 1) + (10 x 1.2) + (24 x 1) + (18 x 0.60) + (1 x 0.80) + (0 x 0.75) + (11 x 0.72) + (1 x 0.60) + (17 x 0.50) = 110.42

Nadal = (14 x 2) + (0 x 1.5) + (0 x 1.3) + (2 x 1) + (6 x 1.2) + (27 x 1) + (14 x 0.60) + (0 x 0.80) + (1 x 0.75) + (3 x 0.72) + (1 x 0.60) + (16 x 0.50) = 84.11

Djokovic = (10 x 2) + (3 x 1.5) + (2 x 1.3) + (0 x 1) + (8 x 1.2) + (26 x 1) + (12 x 0.60) + (0 x 0.80) + (0 x 0.75) + (10 x 0.72) + (0 x 0.60) + (12 x 0.50) = 83.1

• Slam Victories (SV) 2000 ATP points
• Slam Runner-ups (SRU) 1200 ATP points
• Slam Semi-finals (SSF) 720 ATP points
• Season end final victories with no loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFNL) 1500 ATP points
• Season end final victories with one loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFOL) 1300 ATP points
• Season end final runner-ups with no loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFRUNL) 1000 ATP points
• Season end final runner-ups with one loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFRUOL) 800 ATP points
• Season end final semi-finals with no loss before the semi-final (WTF, WCT Finals * & Grand Slam Cup *) (SEFSFNL) ATP 600 points
• Masters 1000 equivalent victories (we will call (Top 9)) ATP 1000 points
• Masters 1000 equivalent runner-ups (TOP9RU) ATP 600 points
• Olympic Gold Metal Singles (OSG) ATP 750 points
• 500 Series equivalents (500S) ATP 500 points
• 
  REMEMBER: There is no agreed weighting of events. In this forum I have tried to get an agreed weightings but opinions as to the weighting vary greatly. The best I can do is use the current ATP weightings. Everytime I post these rankings people disagree with the weightings, but what can I do? - there is no agreed standard beyond the ATP weightings. Also note that this table doesn't represent 'Greatness' which is a subjective term. It simply represents an objective list of the achievements of open era players weighted at current ATP weightings.

 

[xFedal]

Movers after the 2015 WTF (changes highlighted in Red):-

Everything that in today's terms you can earn 500 points and above per event is counted. That is:

For ease I have reduced the weighting points down by a factor of 1000 eg Slams are worth 2 instead of their ATP 2000.

Scale is: (SV x 2) + (SEFNL x 1.5) + (SEFOL x 1.3) + (SEFRUNL x 1) + (SRU x 1.2) + (TOP9 x 1) + (TOP9RU x 0.60) + (SEFRUOL x 0.80) + (OSG x 0.75) + (SSF x 0.72) + (SEFSFNL x 0.60) + (500S x 0.50)

Federer = (17 x 2) + (5 x 1.5) + (1 x 1.3) + (3 x 1) + (10 x 1.2) + (24 x 1) + (18 x 0.60) + (1 x 0.80) + (0 x 0.75) + (11 x 0.72) + (1 x 0.60) + (17 x 0.50) = 110.42

Nadal = (14 x 2) + (0 x 1.5) + (0 x 1.3) + (2 x 1) + (6 x 1.2) + (27 x 1) + (14 x 0.60) + (0 x 0.80) + (1 x 0.75) + (3 x 0.72) + (1 x 0.60) + (16 x 0.50) = 84.11

Djokovic = (10 x 2) + (3 x 1.5) + (2 x 1.3) + (0 x 1) + (8 x 1.2) + (26 x 1) + (12 x 0.60) + (0 x 0.80) + (0 x 0.75) + (10 x 0.72) + (0 x 0.60) + (12 x 0.50) = 83.1

• Slam Victories (SV) 2000 ATP points
• Slam Runner-ups (SRU) 1200 ATP points
• Slam Semi-finals (SSF) 720 ATP points
• Season end final victories with no loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFNL) 1500 ATP points
• Season end final victories with one loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFOL) 1300 ATP points
• Season end final runner-ups with no loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFRUNL) 1000 ATP points
• Season end final runner-ups with one loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFRUOL) 800 ATP points
• Season end final semi-finals with no loss before the semi-final (WTF, WCT Finals * & Grand Slam Cup *) (SEFSFNL) ATP 600 points
• Masters 1000 equivalent victories (we will call (Top 9)) ATP 1000 points
• Masters 1000 equivalent runner-ups (TOP9RU) ATP 600 points
• Olympic Gold Metal Singles (OSG) ATP 750 points
• 500 Series equivalents (500S) ATP 500 points
• 
  REMEMBER: There is no agreed weighting of events. In this forum I have tried to get an agreed weightings but opinions as to the weighting vary greatly. The best I can do is use the current ATP weightings. Everytime I post these rankings people disagree with the weightings, but what can I do? - there is no agreed standard beyond the ATP weightings. Also note that this table doesn't represent 'Greatness' which is a subjective term. It simply represents an objective list of the achievements of open era players weighted at current ATP weightings.

The day draws near when Nole overtakes Nadal, and then it will be Nole and the 'mighty' Federer. Boy I am going to enjoy this journey.

 

[timnz]

The day draws near when Nole overtakes Nadal, and then it will be Nole and the 'mighty' Federer. Boy I am going to enjoy this journey.

If it happens, it will certainly raise the debate about slams vs other titles. There are some people in this forum who believe that 1 slam ahead is worth more than any summation of other titles eg 11 slams + 0 other titles > 10 slams + 5 WTF + 30 Masters 1000's ie they only count slam wins - everything else isn't worth anything unless you are equal in slam count - then they are used as a tie breaker. I don't agree with this at all. I believe tennis is more than 8 weeks a year.

 

[CYGS]

The day draws near when Nole overtakes Nadal, and then it will be Nole and the 'mighty' Federer. Boy I am going to enjoy this journey.

I'm afraid he will twist the "ranking" again.

 

[Novak And The Net.]

LMAO at all the muppets taking this ranking system seriously.

 

[timnz]

LMAO at all the muppets taking this ranking system seriously.

Each to their own mate. Those with an interest in total points won at 500 level and above weighted at current ATP weightings can see them here.

 

[SpicyCurry1990]

Hey man, how've you been? Long time no see on these boards.

Been super busy lately, haven't had a chance to post much, but still been on occasionally just browsing. I still owe NaTF a big post in regards to our Hewitt/Murray comparison at some point ha. Been enjoying Novak's dominant run. The clunker from Tuesday was a bit annoying as it dropped his 5 year win rate to just below 90% in ElDanger's rankings, but hopefully he will get it back at AO and if there was ever a good match to lose its it the one where you aren't out ha.

 

[SpicyCurry1990]

Hey @timnz - Let me know what you think about this approach

I've often felt that your system undervalued slams too much, but I understand completely why you do it that way (no agreement on what they SHOULD be worth, hence take the objective 2000 afforded by the ATP). I like the idea you present here that slams don't mean everything, but I also endorsed the idea that huge slam gaps really need to be reflected more in the rankings somehow, so that one can't SOLELY makeup gaps without them.

So what I devised was simply using a multiplication factor. Take the point totals for each player * the # of majors won by that player. This scales up the total value of each player's achievements with each additional slam. Thus it makes it harder to overcome that deficit without more slams, but still leaves it possible, hence still provides value to other achievements. Plus it takes no subjective re-evaluation since we are using an already established objective number (outside of 1 case I'll mention below).

For ease I then reduced weighting by a factor of 10 to keep the values within the same degree of magnitude as they are now to get:

Federer 187.714
Nadal 117.754
Sampras 96.6
Borg1 88.648
Djokovic 83.1
Borg2* 82.316
Lendl 79.584
Connors 72.416
Borg3** 69.652
McEnroe 51.632
Agassi 49.416
Becker 34.416
Edberg 27.516
Wilander 26.572

The only cloudy case is what multiplier to use for Borg because of the disagreements about the value of his YEC as majors. Counting all 3 as majors you get borg1. Only counting his WTF as majors gives you borg2, and not counting any and going with 11 slams gives you value 3. As you can see this system allows for Lendl/Connors even with an 8 to 11 slam deficit to surpass Borg (if we went with that as his major count), but would punish the 8 to 14 deficit vs Sampras/Nadal.

Using this system, Djokovic just passed Connors and Borg3 at USO, Lendl at Shanghai, and Borg2 at WTF. He will pass Borg1 with slam #11 and Sampras with slam #12. At slam #13 he will be slightly below Nadal, but could easily make up the deficit through just 2 Masters.

 

[uscwang]

Hey @timnz - Let me know what you think about this approach

I've often felt that your system undervalued slams too much, but I understand completely why you do it that way (no agreement on what they SHOULD be worth, hence take the objective 2000 afforded by the ATP). I like the idea you present here that slams don't mean everything, but I also endorsed the idea that huge slam gaps really need to be reflected more in the rankings somehow, so that one can't SOLELY makeup gaps without them.

So what I devised was simply using a multiplication factor. Take the point totals for each player * the # of majors won by that player. This scales up the total value of each player's achievements with each additional slam. Thus it makes it harder to overcome that deficit without more slams, but still leaves it possible, hence still provides value to other achievements. Plus it takes no subjective re-evaluation since we are using an already established objective number (outside of 1 case I'll mention below).

For ease I then reduced weighting by a factor of 10 to keep the values within the same degree of magnitude as they are now to get:

Federer 187.714
Nadal 117.754
Sampras 96.6
Borg1 88.648
Djokovic 83.1
Borg2* 82.316
Lendl 79.584
Connors 72.416
Borg3** 69.652
McEnroe 51.632
Agassi 49.416
Becker 34.416
Edberg 27.516
Wilander 26.572

The only cloudy case is what multiplier to use for Borg because of the disagreements about the value of his YEC as majors. Counting all 3 as majors you get borg1. Only counting his WTF as majors gives you borg2, and not counting any and going with 11 slams gives you value 3. As you can see this system allows for Lendl/Connors even with an 8 to 11 slam deficit to surpass Borg (if we went with that as his major count), but would punish the 8 to 14 deficit vs Sampras/Nadal.

Using this system, Djokovic just passed Connors and Borg3 at USO, Lendl at Shanghai, and Borg2 at WTF. He will pass Borg1 with slam #11 and Sampras with slam #12. At slam #13 he will be slightly below Nadal, but could easily make up the deficit through just 2 Masters.

I think you are lost in number games. The composite scores should quantify and reflect people's general opinion about tennis players, not distort it. Does it sound right to you that Federer is twice as good as Sampras, or 6 times better than Edberg?

 

[Nathaniel_Near]

Hey @timnz - Let me know what you think about this approach

I've often felt that your system undervalued slams too much, but I understand completely why you do it that way (no agreement on what they SHOULD be worth, hence take the objective 2000 afforded by the ATP). I like the idea you present here that slams don't mean everything, but I also endorsed the idea that huge slam gaps really need to be reflected more in the rankings somehow, so that one can't SOLELY makeup gaps without them.

So what I devised was simply using a multiplication factor. Take the point totals for each player * the # of majors won by that player. This scales up the total value of each player's achievements with each additional slam. Thus it makes it harder to overcome that deficit without more slams, but still leaves it possible, hence still provides value to other achievements. Plus it takes no subjective re-evaluation since we are using an already established objective number (outside of 1 case I'll mention below).

For ease I then reduced weighting by a factor of 10 to keep the values within the same degree of magnitude as they are now to get:

Federer 187.714
Nadal 117.754
Sampras 96.6
Borg1 88.648
Djokovic 83.1
Borg2* 82.316
Lendl 79.584
Connors 72.416
Borg3** 69.652
McEnroe 51.632
Agassi 49.416
Becker 34.416
Edberg 27.516
Wilander 26.572

The only cloudy case is what multiplier to use for Borg because of the disagreements about the value of his YEC as majors. Counting all 3 as majors you get borg1. Only counting his WTF as majors gives you borg2, and not counting any and going with 11 slams gives you value 3. As you can see this system allows for Lendl/Connors even with an 8 to 11 slam deficit to surpass Borg (if we went with that as his major count), but would punish the 8 to 14 deficit vs Sampras/Nadal.

Using this system, Djokovic just passed Connors and Borg3 at USO, Lendl at Shanghai, and Borg2 at WTF. He will pass Borg1 with slam #11 and Sampras with slam #12. At slam #13 he will be slightly below Nadal, but could easily make up the deficit through just 2 Masters.

Accurate looking results. Of course, I realise that the points total is not the point here and that it's about the order. Using the ranking system as a means for "objective data" only works for a list of most achieved players according to ranking points, and even then it runs into problems — at least this is what it is proclaimed to be and not anything more.

At least this effort tries to function as a method for figuring out greatness rather than operating solely under the ranking system as some unsatisfactory aegis of objectivity (which it isn't), though it still suffers by using the same fundamental framework originally presented in this thread which frames what has often been a volatile era by only one standard (thus it is a lazy framework).

 

[SpicyCurry1990]

I think you are lost in number games. The composite scores should quantify and reflect people's general opinion about tennis players, not distort it. Does it sound right to you that Federer is twice as good as Sampras, or 6 times better than Edberg?

The baseline "objective" list based on pure points goes against this as well though, does it not? Does it sound right to you that Connors and Lendl are above Sampras? Or that they are so far above Borg that even if he had added 8 more slams (the career total won by Connors and Lendl), they would still be ahead? That's what the current framework displays.

This effort produces issues with relative comparative evaluation about "how much greater" at the top end (as you would expect with a multiplication factor), but I think it does a good job of reflecting general opinion about rank order.

In any case your question about X being Y times better than Z, I have answered in another topic for the methodology I use to rank players based on slam win ratios over slam win differences. Here I was merely attempting to inject discussion to an established premise.

Accurate looking results. Of course, I realise that the points total is not the point here and that it's about the order. Using the ranking system as a means for "objective data" only works for a list of most achieved players according to ranking points, and even then it runs into problems — at least this is what it is proclaimed to be and not anything more.

At least this effort tries to function as a method for figuring out greatness rather than operating solely under the ranking system as some unsatisfactory aegis of objectivity (which it isn't), though it still suffers by using the same fundamental framework originally presented in this thread which frames what has often been a volatile era by only one standard (thus it is a lazy framework).

My goal was to stay in line WITH the general framework presented here and to work from that to find enhancement as opposed to stating an entirely new framework (which I did to you in a different topic, which reminds me been busy with holidays I know there is a lot I need to reply to you there as well).

 

[Nathaniel_Near]

The baseline "objective" list based on pure points goes against this as well though, does it not? Does it sound right to you that Connors and Lendl are above Sampras? Or that they are so far above Borg that even if he had added 8 more slams (the career total won by Connors and Lendl), they would still be ahead? That's what the current framework displays.

This effort produces issues with relative comparative evaluation about "how much greater" at the top end (as you would expect with a multiplication factor), but I think it does a good job of reflecting general opinion about rank order.

In any case your question about X being Y times better than Z, I have answered in another topic for the methodology I use to rank players based on slam win ratios over slam win differences. Here I was merely attempting to inject discussion to an established premise.



My goal was to stay in line WITH the general framework presented here and to work from that to find enhancement as opposed to stating an entirely new framework (which I did to you in a different topic, which reminds me been busy with holidays I know there is a lot I need to reply to you there as well).

Ye I could tell your goal. I like your take.

 

[uscwang]

The baseline "objective" list based on pure points goes against this as well though, does it not? Does it sound right to you that Connors and Lendl are above Sampras? Or that they are so far above Borg that even if he had added 8 more slams (the career total won by Connors and Lendl), they would still be ahead? That's what the current framework displays.

This effort produces issues with relative comparative evaluation about "how much greater" at the top end (as you would expect with a multiplication factor), but I think it does a good job of reflecting general opinion about rank order.

In any case your question about X being Y times better than Z, I have answered in another topic for the methodology I use to rank players based on slam win ratios over slam win differences. Here I was merely attempting to inject discussion to an established premise.



My goal was to stay in line WITH the general framework presented here and to work from that to find enhancement as opposed to stating an entirely new framework (which I did to you in a different topic, which reminds me been busy with holidays I know there is a lot I need to reply to you there as well).

I'm not defending timnz' system. I have my own index for tennis greatness, which I felt reflect tennis common sense better than using GS titles or ATP pts alone.
http://tt.tennis-warehouse.com/index.php?threads/larrys-tennis-greatness-index.529698/
The bottom line is, any system that puts Fed's score at SEVEN times Edberg's is seriously flawed. It's not going to fly.

 

[timnz]

I'm not defending timnz' system. I have my own index for tennis greatness, which I felt reflect tennis common sense better than using GS titles or ATP pts alone.
http://tt.tennis-warehouse.com/index.php?threads/larrys-tennis-greatness-index.529698/
The bottom line is, any system that puts Fed's score at SEVEN times Edberg's is seriously flawed. It's not going to fly.

See, here's the thing. Having a system 'Reflecting common sense' means that you could (not saying you do) have a prescribed end goal in mind. And if that is the case then that casts doubt on the system itself. When people critique my system (which they, of course, are absolutely free to do) - they say - but that isn't right having Connors above Borg). What I want to ask is why? What is telling you that you think Borg should be above Connors? Doesn't that mean you have a system of evaluating players in your mind that you are appealing to? In a strange way I like that my system doesn't match at all my own personal subjective feeling as to the hierarchy of players. I tried to be objective because of what it may reveal. One thing that it has revealed, is that most people hugely underrate Lendl.

 

[djokerer]

NOTE: After reflection in the last few months, I have decided to change the Open era achievement ranking system that I have been posted for a number of years. The reasons are principally around the Masters 1000 achievements and their equivalents. When I found out that Lendl only competed in 10 of 27 Masters 1000's in 1990-1992 (because they were not compulsory then), and that there were 8-9 other tournaments per year from 1990-1992 that were equivalent in points and prize money to the 'official' Masters 1000 - that made me wonder if our count of these events was fair.

The problem that I have wrestled with mostly is this:

The earlier events didn't have the depth of top players that today's do (which is harder on more recent players) and, because of it not being compulsory, and there were other comparible events - (then it is unfair on the older players who didn't compete in what we deem Masters 1000 equivalents). So how do we work out equivalency given that there were completely different contexts for the playing of these events pre-2000 (particularly before 1993)?

Masters 1000 pre-1990 are difficult to agree on. There is no agreed 'Masters 1000' equivalent list. The only list that I have seen some agreement on, in these forums, is:

http://en.wikipedia.org/wiki/Tennis_Masters_Series_records_and_statistics

I have decided that we should therefore include all players 500 level event victories. Most of the 'defacto' Masters 1000's have been rated at the 500 level. Therefore in a ranking system, they don't give as many points, but at least they get represented somewhat. I have therefore changed the cut-off to 500 points and above, where previously I had 1000 points and above per event.

Details of the system:

Everything that in today's terms you can earn 500 points and above per event is counted. That is:

• Slam Victories (SV) 2000 ATP points
• Slam Runner-ups (SRU) 1200 ATP points
• Slam Semi-finals (SSF) 720 ATP points
• Season end final victories with no loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFNL) 1500 ATP points
• Season end final victories with one loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFOL) 1300 ATP points
• Season end final runner-ups with no loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFRUNL) 1000 ATP points
• Season end final runner-ups with one loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFRUOL) 800 ATP points
• Season end final semi-finals with no loss before the semi-final (WTF, WCT Finals * & Grand Slam Cup *) (SEFSFNL) ATP 600 points
• Masters 1000 equivalent victories (we will call (Top 9)) ATP 1000 points
• Masters 1000 equivalent runner-ups (TOP9RU) ATP 600 points
• Olympic Gold Metal Singles (OSG) ATP 750 points
• 500 Series equivalents (500S) ATP 500 points

* Note: To even out the fact that pre-mid 1980's great players tended to play 3 slams a year vs 4 slams a year for current players, I will only include WCT Finals and Grand Slam Cup placings if the player didn't play all the slams in that year eg I include Lendl's 1982 WCT Finals win but I don't include his 1985 WCT Finals win, since in 1982 he didn't play all the slams but in 1985 he did. That way it is fair to modern players that the older players aren't getting an extra event to score points in (since modern players don't have the WCT Finals or Grand Slam Cup to count).

REMEMBER: There is no agreed weighting of events. In this forum I have tried to get an agreed weightings but opinions as to the weighting vary greatly. The best I can do is use the current ATP weightings. Everytime I post these rankings people disagree with the weightings, but what can I do? - there is no agreed standard beyond the ATP weightings. Also note that this table doesn't represent 'Greatness' which is a subjective term. It simply represents an objective list of the achievements of open era players weighted at current ATP weightings.

For ease I have reduced the weighting points down by a factor of 1000 eg Slams are worth 2 instead of their ATP 2000.

Scale is: (SV x 2) + (SEFNL x 1.5) + (SEFOL x 1.3) + (SEFRUNL x 1) + (SRU x 1.2) + (TOP9 x 1) + (TOP9RU x 0.60) + (SEFRUOL x 0.80) + (OSG x 0.75) + (SSF x 0.72) + (SEFSFNL x 0.60) + (500S x 0.50)

Federer = (17 x 2) + (5 x 1.5) + (1 x 1.3) + (3 x 1) + (10 x 1.2) + (24 x 1) + (18 x 0.60) + (1 x 0.80) + (0 x 0.75) + (11 x 0.72) + (1 x 0.60) + (17 x 0.50) = 110.42

Lendl = (8 x 2) + ((5 + 2 - 1) x 1.5)) + (0 x 1.3) + (2 x 1) + (11 x 1.2) + (22 x 1) + (11 x 0.60) + (2 x 0.80) + (0 x 0.75) + (9 x 0.72) + ((5 -2) x 0.60) + (42 x 0.50) = 99.48

Connors = (8 x 2) + (2 x 1.5) + (1 x 1.3) + (1 x 1) + (7 x 1.2) + (17 x 1) + (9 x 0.60) + (0 x 0.80) + (0 x 0.75) + (16 x 0.72) + (4 x 0.60) + (49 x 0.50) = 90.52

Nadal = (14 x 2) + (0 x 1.5) + (0 x 1.3) + (2 x 1) + (6 x 1.2) + (27 x 1) + (14 x 0.60) + (0 x 0.80) + (1 x 0.75) + (3 x 0.72) + (1 x 0.60) + (16 x 0.50) = 84.11

Djokovic = (10 x 2) + (3 x 1.5) + (2 x 1.3) + (0 x 1) + (8 x 1.2) + (26 x 1) + (12 x 0.60) + (0 x 0.80) + (0 x 0.75) + (10 x 0.72) + (0 x 0.60) + (12 x 0.50) = 83.1

McEnroe = (7 x 2) + ((3 + (5 - 1)) x 1.5)) + (0 x 1.3) + ((1 + 3) x 1) + (4 x 1.2) + (19 x 1) + (7 x 0.60) + (0 x 0.80) + (0 x 0.75) + (8 x 0.72) + (0 x 0.60) + (23 x 0.50) = 73.76

Sampras = (14 x 2) + (0 + (2 - 1) x 1.5) + (5 x 1.3) + ((2 - 1) x 1) + (4 x 1.2) + (11 x 1) + (8 x 0.60) + (0 x 0.80) + (0 x 0.75) + (5 x 0.72) + ((4 - 1) x 0.60) + (12 x 0.50) = 69.0

Borg = (11 x 2) + ((1 + 1) x 1.5)) + (1 x 1.3) + ((1 + (3 - 1)) x 1) + (5 x 1.2) + (15 x 1) + (4 x 0.60) + (1 x 0.80) + (0 x 0.75) + (1 x 0.72) + (1 x 0.60) + (17 x 0.50) = 63.32

Agassi = (8 x 2) + (0 x 1.5) + (1 x 1.3) + ((3 - 1) x 1) + (7 x 1.2) + (17 x 1) + (5 x 0.60) + (1 x 0.80) + (1 x 0.75) + (11 x 0.72) + (1 x 0.60) + (6 x 0.50) = 61.77

Becker = (6 x 2) + ((1 + 1) x 1.5)) + (3 x 1.3) + ((4 + 1) x 1) + (4 x 1.2) + (13 x 1) + (8 x 0.60) + (0 x 0.80) + (0 x 0.75) + (8 x 0.72) + ((2 - 1) x 0.60) + (9 x 0.50)= 57.36

Edberg = (6 x 2) + (0 x 1.5) + (1 x 1.3) + ((2 - 1) x 1) + (5 x 1.2) + (8 x 1) + (12 x 0.60) + (0 x 0.80) + (0 x 0.75) + (8 x 0.72) + ((2 - 1) x 0.60) + (8 x 0.50) = 45.86

Wilander = (7 x 2) + (0 x 1.5) + (0 x 1.3) + (0 x 1) + (4 x 1.2) + (8 x 1) + (7 x 0.60) + (1 x 0.80) + (0 x 0.75) + (3 x 0.72) + (0 x 0.60) + (8 x 0.50****) = 37.96

**** I found it difficult to determine that Wilander's 500 level equivalents are. Depending upon approaches, I ended up with anything from 5 to 10. I have settled (for now) on 8.

When you decide to include particular tournament for a particular player for a particular time, your metric just crossed over from being objective to subjective.
You ruined a nice thing.

 

[timnz]

Hey @timnz - Let me know what you think about this approach

I've often felt that your system undervalued slams too much, but I understand completely why you do it that way (no agreement on what they SHOULD be worth, hence take the objective 2000 afforded by the ATP). I like the idea you present here that slams don't mean everything, but I also endorsed the idea that huge slam gaps really need to be reflected more in the rankings somehow, so that one can't SOLELY makeup gaps without them.

So what I devised was simply using a multiplication factor. Take the point totals for each player * the # of majors won by that player. This scales up the total value of each player's achievements with each additional slam. Thus it makes it harder to overcome that deficit without more slams, but still leaves it possible, hence still provides value to other achievements. Plus it takes no subjective re-evaluation since we are using an already established objective number (outside of 1 case I'll mention below).

For ease I then reduced weighting by a factor of 10 to keep the values within the same degree of magnitude as they are now to get:

Federer 187.714
Nadal 117.754
Sampras 96.6
Borg1 88.648
Djokovic 83.1
Borg2* 82.316
Lendl 79.584
Connors 72.416
Borg3** 69.652
McEnroe 51.632
Agassi 49.416
Becker 34.416
Edberg 27.516
Wilander 26.572

The only cloudy case is what multiplier to use for Borg because of the disagreements about the value of his YEC as majors. Counting all 3 as majors you get borg1. Only counting his WTF as majors gives you borg2, and not counting any and going with 11 slams gives you value 3. As you can see this system allows for Lendl/Connors even with an 8 to 11 slam deficit to surpass Borg (if we went with that as his major count), but would punish the 8 to 14 deficit vs Sampras/Nadal.

Using this system, Djokovic just passed Connors and Borg3 at USO, Lendl at Shanghai, and Borg2 at WTF. He will pass Borg1 with slam #11 and Sampras with slam #12. At slam #13 he will be slightly below Nadal, but could easily make up the deficit through just 2 Masters.

I really appreciate your comments and ideas!

I myself believe that Slams should be weighted heavier. However, as I have said, that is just my opinion. So the best I can do is use the ATP weightings and say what my ranking system is. It is simply every achievement added up at 500 points and higher on the current ATP scale. Perhaps it doesn't match 'intuitive ranking' . But it isn't a measure of
'Greatness' simply 'achievement'. Now regarding your formulation . Dont' you feel you perhaps have gone to the other end of the spectrum (in terms of weighting slams). Would what you have formulated result in slams being weighed around 10 times a Masters 1000? That surely is far too much.

 

[Lord Anomander]

I really appreciate your comments and ideas!

I myself believe that Slams should be weighted heavier. However, as I have said, that is just my opinion. So the best I can do is use the ATP weightings and say what my ranking system is. It is simply every achievement added up at 500 points and higher on the current ATP scale. Perhaps it doesn't match 'intuitive ranking' . But it isn't a measure of
'Greatness' simply 'achievement'. Now regarding your formulation . Dont' you feel you perhaps have gone to the other end of the spectrum (in terms of weighting slams). Would what you have formulated result in slams being weighed around 10 times a Masters 1000? That surely is far too much.

One could argue that a slam is approximately four times the value of a masters (instead of just twice as it is with the current ratings). This is due to the fact that a slam is only held four times a year, whereas masters are played approx. twice as often (9 times). I think slams are heavily overrated as it already is (without weighing it 10 times a masters 1000). At least in common sense all titles besides grand slam titles are only used to define who was greater if they are equal in slams.

A useful criteria for weighting could be to multiply each result with the number of appearances as it refers to the winning percentage. Of course somebody who played 10 slams and won 8 is apparently better than somebody who played 50 and won 15. But I'm actually fine with this rating/ranking system of yours as it is.

 

[timnz]

When you decide to include particular tournament for a particular player for a particular time, your metric just crossed over from being objective to subjective.
You ruined a nice thing.

What particular tournament. What particular player? I counted every players 500 series totals. Not sure what you are referring to.

 

[djokerer]

What particular tournament. What particular player? I counted every players 500 series totals. Not sure what you are referring to.

Refer your Op the example of Lendl wct 1982 vs 1985

 

[timnz]

Refer your Op the example of Lendl wct 1982 vs 1985

That has been in my system for years (there was a large thread that predates this one). It isn't to do with Lendl, it is to do with a practice of the 1970s and early 1980s players of typically playing on 3 slams a year. If I am to have a system which compares players for the total open era period...it is not fair to the older players who only get 3 out of 4 tries at the top event in the system. To balance the score card I include results of the WCT finals/Grand slam cup ONLY IF a player didn't play 4 slams in THAT PARTICULAR year. The WCT finals was rated as an important event back then - and particularly in the 1970s as a defacto major. So this tweek didn't impact Lendl only, it impacted Connors, Borg, McEnroe, Becker, Sampras and Edberg. It was an acknowledgement of the reality of what was constituted as a major back then. Lednl WCT FINALS 1985 and McEnroe's 1983 and Borg's 1974 runner-up did NOT get included because in those years those players played all the slams in that year.

This offset has been around in my system for at least 3 or 4 years.

 

[theroguedog]

NOTE: After reflection in the last few months, I have decided to change the Open era achievement ranking system that I have been posted for a number of years. The reasons are principally around the Masters 1000 achievements and their equivalents. When I found out that Lendl only competed in 10 of 27 Masters 1000's in 1990-1992 (because they were not compulsory then), and that there were 8-9 other tournaments per year from 1990-1992 that were equivalent in points and prize money to the 'official' Masters 1000 - that made me wonder if our count of these events was fair.

The problem that I have wrestled with mostly is this:

The earlier events didn't have the depth of top players that today's do (which is harder on more recent players) and, because of it not being compulsory, and there were other comparible events - (then it is unfair on the older players who didn't compete in what we deem Masters 1000 equivalents). So how do we work out equivalency given that there were completely different contexts for the playing of these events pre-2000 (particularly before 1993)?

Masters 1000 pre-1990 are difficult to agree on. There is no agreed 'Masters 1000' equivalent list. The only list that I have seen some agreement on, in these forums, is:

http://en.wikipedia.org/wiki/Tennis_Masters_Series_records_and_statistics

I have decided that we should therefore include all players 500 level event victories. Most of the 'defacto' Masters 1000's have been rated at the 500 level. Therefore in a ranking system, they don't give as many points, but at least they get represented somewhat. I have therefore changed the cut-off to 500 points and above, where previously I had 1000 points and above per event.

Details of the system:

Everything that in today's terms you can earn 500 points and above per event is counted. That is:

• Slam Victories (SV) 2000 ATP points
• Slam Runner-ups (SRU) 1200 ATP points
• Slam Semi-finals (SSF) 720 ATP points
• Season end final victories with no loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFNL) 1500 ATP points
• Season end final victories with one loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFOL) 1300 ATP points
• Season end final runner-ups with no loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFRUNL) 1000 ATP points
• Season end final runner-ups with one loss before the final (WTF, WCT Finals * & Grand Slam Cup *) (SEFRUOL) 800 ATP points
• Season end final semi-finals with no loss before the semi-final (WTF, WCT Finals * & Grand Slam Cup *) (SEFSFNL) ATP 600 points
• Masters 1000 equivalent victories (we will call (Top 9)) ATP 1000 points
• Masters 1000 equivalent runner-ups (TOP9RU) ATP 600 points
• Olympic Gold Metal Singles (OSG) ATP 750 points
• 500 Series equivalents (500S) ATP 500 points

* Note: To even out the fact that pre-mid 1980's great players tended to play 3 slams a year vs 4 slams a year for current players, I will only include WCT Finals and Grand Slam Cup placings if the player didn't play all the slams in that year eg I include Lendl's 1982 WCT Finals win but I don't include his 1985 WCT Finals win, since in 1982 he didn't play all the slams but in 1985 he did. That way it is fair to modern players that the older players aren't getting an extra event to score points in (since modern players don't have the WCT Finals or Grand Slam Cup to count).

REMEMBER: There is no agreed weighting of events. In this forum I have tried to get an agreed weightings but opinions as to the weighting vary greatly. The best I can do is use the current ATP weightings. Everytime I post these rankings people disagree with the weightings, but what can I do? - there is no agreed standard beyond the ATP weightings. Also note that this table doesn't represent 'Greatness' which is a subjective term. It simply represents an objective list of the achievements of open era players weighted at current ATP weightings.

For ease I have reduced the weighting points down by a factor of 1000 eg Slams are worth 2 instead of their ATP 2000.

Scale is: (SV x 2) + (SEFNL x 1.5) + (SEFOL x 1.3) + (SEFRUNL x 1) + (SRU x 1.2) + (TOP9 x 1) + (TOP9RU x 0.60) + (SEFRUOL x 0.80) + (OSG x 0.75) + (SSF x 0.72) + (SEFSFNL x 0.60) + (500S x 0.50)

Federer = (17 x 2) + (5 x 1.5) + (1 x 1.3) + (3 x 1) + (10 x 1.2) + (24 x 1) + (18 x 0.60) + (1 x 0.80) + (0 x 0.75) + (11 x 0.72) + (1 x 0.60) + (17 x 0.50) = 110.42

Lendl = (8 x 2) + ((5 + 2 - 1) x 1.5)) + (0 x 1.3) + (2 x 1) + (11 x 1.2) + (22 x 1) + (11 x 0.60) + (2 x 0.80) + (0 x 0.75) + (9 x 0.72) + ((5 -2) x 0.60) + (42 x 0.50) = 99.48

Connors = (8 x 2) + (2 x 1.5) + (1 x 1.3) + (1 x 1) + (7 x 1.2) + (17 x 1) + (9 x 0.60) + (0 x 0.80) + (0 x 0.75) + (16 x 0.72) + (4 x 0.60) + (49 x 0.50) = 90.52

Nadal = (14 x 2) + (0 x 1.5) + (0 x 1.3) + (2 x 1) + (6 x 1.2) + (27 x 1) + (14 x 0.60) + (0 x 0.80) + (1 x 0.75) + (3 x 0.72) + (1 x 0.60) + (16 x 0.50) = 84.11

Djokovic = (10 x 2) + (3 x 1.5) + (2 x 1.3) + (0 x 1) + (8 x 1.2) + (26 x 1) + (12 x 0.60) + (0 x 0.80) + (0 x 0.75) + (10 x 0.72) + (0 x 0.60) + (12 x 0.50) = 83.1

McEnroe = (7 x 2) + ((3 + (5 - 1)) x 1.5)) + (0 x 1.3) + ((1 + 3) x 1) + (4 x 1.2) + (19 x 1) + (7 x 0.60) + (0 x 0.80) + (0 x 0.75) + (8 x 0.72) + (0 x 0.60) + (23 x 0.50) = 73.76

Sampras = (14 x 2) + (0 + (2 - 1) x 1.5) + (5 x 1.3) + ((2 - 1) x 1) + (4 x 1.2) + (11 x 1) + (8 x 0.60) + (0 x 0.80) + (0 x 0.75) + (5 x 0.72) + ((4 - 1) x 0.60) + (12 x 0.50) = 69.0

Borg = (11 x 2) + ((1 + 1) x 1.5)) + (1 x 1.3) + ((1 + (3 - 1)) x 1) + (5 x 1.2) + (15 x 1) + (4 x 0.60) + (1 x 0.80) + (0 x 0.75) + (1 x 0.72) + (1 x 0.60) + (17 x 0.50) = 63.32

Agassi = (8 x 2) + (0 x 1.5) + (1 x 1.3) + ((3 - 1) x 1) + (7 x 1.2) + (17 x 1) + (5 x 0.60) + (1 x 0.80) + (1 x 0.75) + (11 x 0.72) + (1 x 0.60) + (6 x 0.50) = 61.77

Becker = (6 x 2) + ((1 + 1) x 1.5)) + (3 x 1.3) + ((4 + 1) x 1) + (4 x 1.2) + (13 x 1) + (8 x 0.60) + (0 x 0.80) + (0 x 0.75) + (8 x 0.72) + ((2 - 1) x 0.60) + (9 x 0.50)= 57.36

Edberg = (6 x 2) + (0 x 1.5) + (1 x 1.3) + ((2 - 1) x 1) + (5 x 1.2) + (8 x 1) + (12 x 0.60) + (0 x 0.80) + (0 x 0.75) + (8 x 0.72) + ((2 - 1) x 0.60) + (8 x 0.50) = 45.86

Wilander = (7 x 2) + (0 x 1.5) + (0 x 1.3) + (0 x 1) + (4 x 1.2) + (8 x 1) + (7 x 0.60) + (1 x 0.80) + (0 x 0.75) + (3 x 0.72) + (0 x 0.60) + (8 x 0.50****) = 37.96

**** I found it difficult to determine that Wilander's 500 level equivalents are. Depending upon approaches, I ended up with anything from 5 to 10. I have settled (for now) on 8.

If you haven't seen this info it's worth a look... a different slant on a number of different ranking categories....

http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0017249

Who Is the Best Player Ever? A Complex Network Analysis of the History of Professional Tennis
Filippo Radicchi

Published: February 9, 2011 DOI: 10.1371/journal.pone.0017249

Table 1. Top 30 players in the history of tennis.
doi:10.1371/journal.pone.0017249.t001

I should have included this on the first post...

The study, published in the scientific journal PLoS ONE, discards traditional methods, such as weeks at No. 1 and Grand Slam titles. Instead, it analyzes all matches played in men's tennis since the beginning of the open era, in 1968, and awards each player a "prestige score" based on matches, especially victories, against quality opponents.

"What's really important is not to win many matches, but to win matches against other good players," says Filippo Radicchi, a statistical physicist at Northwestern University who authored the study.

Connors won 178 quality matches, more than any other player. By Radicchi's measure, Federer has 39 quality victories. Nadal has just 21. However, Radicchi notes that his method favors retired players, because the historical stature of current pros has yet to be determined. "I'll run this algorithm again in 10 years and see if the ranking is still the same," he says.

 

[timnz]

If you haven't seen this info it's worth a look... a different slant on a number of different ranking categories....

http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0017249

Who Is the Best Player Ever? A Complex Network Analysis of the History of Professional Tennis
Filippo Radicchi

Published: February 9, 2011DOI: 10.1371/journal.pone.0017249

Table 1. Top 30 players in the history of tennis.
doi:10.1371/journal.pone.0017249.t001

Interesting. I have to confess to not grasping their complex mathematical algorithims. I guess if a system states its assumptions then people are clear what they are getting. So all power to them. I get grief over my system. I'd imagine that those who have a system with Vilas, as great a player as he was, having a superior career to Agassi, Edberg, Federer, Sampras - may have some detractors.