bandabou

Sep 15th, 2004, 10:06 AM

Landing On Top

With the announcement that Amelie Mauresmo is the WTA's new #1, there has been quite a bit of highly predictable dissatisfaction. It was a bit funny to sit down on Saturday morning and suddenly think, "OK, so who should be #1?" -- and not have a clue to the answer. Mauresmo's claim is certainly not overwhelming -- but the credentials of the other contenders (Lindsay Davenport, Justine Henin-Hardenne, Anastasia Myskina) aren't instantly overwhelming, either.

We of course can't make the WTA adopt a better ranking system -- we've been writing about this for years, and all they do is keep coming up with proposals for less accurate ranking systems. But we can at least try to reach an answer that satisfies us -- and maybe you.

So what we'll do is take the Top Four -- Mauresmo, Myskina, Davenport, and Henin-Hardenne -- and compare them under a variety of other measures.

Understand that this is based on results right now, this week. Lindsay Davenport still has a huge lead in the Race, and is the likely year-end #1. But we're looking at the current moment.

To start with, the rankings as the WTA lists them.

1 Mauresmo........4527

2 Myskina.........4155

3 Davenport.......4057

4 Henin-Hardenne..4004

Now let's try an obvious alternative: Won/Lost in the past twelve months. This is a measure, obviously, of who wins when she plays.

Rank..Player..........Wins..Losses.....%

1.....Davenport........56......8....87.9

2.....Henin-Hardenne...47......8....85.5

3.....Mauresmo.........57.....15....79.2

4.....Myskina..........53.....17....75.7

Obviously Davenport wins this measure, though Henin-Hardenne is close enough that it wouldn't be too bothersome if the Belgian were considered the #1. A minor caution is required: Davenport has played two Tier III events in the past twelve months (Strasbourg and Cincinnati), posting a 6-1 record. Myskina played Sopot, going 2-0 and withdrawing to play the Olympics, er, rest an injury. Tier III events really don't involve the same level of difficulty as stronger events, though, meaning that wins there mean less. Happily, it's easy to show that eliminating those events wouldn't affect the standings; Davenport would still have the best percentage and Myskina the worst.

For those interested in Slam won/lost, we have Mauresmo at technically 17-3.(85%, but of course she withdrew from the Australian Open), Henin at 11-2.(84.6%), Myskina 14-3.(82.4%), Davenport at 17-4.(81.0%). Thus Davenport, our winningest player overall, has the worst record at Slams! (Which goes far to explain the present situation, when you think about it.)

If we look at the players each has beaten, the following list shows each player's wins over then-Top Ten players over the past year, plus wins over players who are now Top Ten, plus the Williams Sisters. The number to the left of the player's name represents her ranking at the time she was beaten.

Davenport:

3. Mauresmo (Amelia Island)

5. Myskina (San Diego)

8. Dementieva (Sydney)

9. Zvonareva (Cincinnati)

[12. V. Williams (U. S. Open)]

[13. V. Williams (Los Angeles)]

[15. V. Williams (Stanford)]

[16. S. Williams (Los Angeles)]

Henin-Hardenne:

2. Clijsters (Australian Open)

2. Mauresmo (Olympics)

3. Myskina (Olympics)

4. Mauresmo (Sydney)

4. Davenport (Indian Wells)

5. Capriati (Los Angeles Championships)

5. Davenport (Australian Open)

5. Myskina (Indian Wells)

8. Dementieva (Filderstadt)

8. Myskina (Los Angeles Championships)

10. Rubin (Sydney)

[14. Kuznetsova (Indian Wells)]

[29. Kuznetsova (Dubai)]

[33. Kuznetsova (Australian Open)]

Mauresmo:

1. Henin-Hardenne (Amelia Island)

2. Henin-Hardenne (Los Angeles Championships)

8. Dementieva (Moscow)

8. Capriati (Berlin)

7. Myskina (Sydney)

9. Myskina (Philadelphia)

9. Dementieva (Los Angeles Championships)

9. Capriati (Rome)

10. Kuznetova (Olympics)

[14. Kuznetsova (Berlin)]

Myskina:

1. Clijsters (Leipzig)

2. Henin-Hardenne (Leipzig)

5. Capriati (Doha)

6. Capriati (Roland Garros)

7. Mauresmo (Moscow)

8. Sharapova (San Diego)

9. V. Williams (Roland Garros)

10. Rubin (Australian Open)

10. Dementieva (Roland Garros)

[11. Kuznetsova (Roland Garros)]

[20. Kuznetsova (Doha)]

[24. Sharapova (Indian Wells)]

[32. Sharapova (Australian Open)]

This presents a completely different picture. Davenport may be winning matches, but she isn't beating anyone in particular except Williamses in the throes of horrid slumps. We thought we would need some sort of formula for this, but it seems pretty clear that Henin-Hardenne -- the only one of the four to have beaten all three of her rivals -- is the best at beating top opponents, Myskina probably next, Mauresmo third, and Davenport dead last.

Now let's try another measure: Points per tournament. This, again, measures how effective a player is when she plays -- and includes quality points, measuring toughness of the opposition. That gives us this list:

Rank..Player........Points...Events...Pts/Event

1.....Henin-Hardenne..4004.....13........308.0

2.....Davenport.......4057.....15........270.5

3.....Mauresmo........4258.....18........236.6

4.....Myskina.........4194.....20........209.7

Once again, Henin-Hardenne comes out on top, and would still be #1 even if her total were divided by the old WTA minimum divisor of 14.

Let's try one more measure: "Winning the big ones." We'll examine all players who have won titles at the Tier II and higher levels. From that we'll create a ranking system of sorts, based solely on titles. We'll base this very loosely on the ATP Race system: We'll set a Slam as worth 200 points. That makes the Championships worth 150, and a Tier I worth 100. But we'll keep that wonderful invention, the Tier II, and make it worth 75 points. A Tier III we'll set equal to 35 points. But we'll also give bonuses and subtractions, somewhat arbitrarily: A Tier I or Tier II that is significantly stronger than average we will be worth 20 extra points; a weak Tier I or Tier II will lose 20 points. The list below will include all winners of Tier II or better events in the past year. Events are listed in the order played.

Bovina won:

New Haven (55/Tier II-)

Total: 55 points

Clijsters won:

Filderstadt (95/Tier II+)

Luxembourg (35/Tier III)

Los Angeles Championships (150/Championships)

Paris (55/Tier II-)

Antwerp (55/Tier II-)

Total: 390 points

Davenport won:

Pan Pacific (80/Tier I-)

Amelia Island (75/Tier II)

Stanford (55/Tier II-)

Los Angeles (75/Tier II)

San Diego (120/Tier I+)

Cincinnati (35/Tier III)

Total: 440 points

Dementieva won:

Shanghai (55/Tier II-)

Total: 55 points

Henin-Hardenne won:

Zurich (100/Tier I)

Sydney (95/Tier II+)

Australian Open (200/Slam)

Dubai (75/Tier II)

Indian Wells (120/Tier I+)

Olympics (80/Olympics)

Total: 670 points

Kuznetsova won:

Eastbourne (55/Tier II-)

U. S. Open (200 points)

Total: 255 points

Mauresmo won:

Philadelphia (55/Tier II-)

Berlin (80/Tier I-)

Rome (100/Tier I)

Canadian Open (80/Tier I-)

Total: 315 points

Myskina won:

Leipzig (95/Tier II+)

Moscow (80/Tier I-)

Doha (75/Tier II)

Roland Garros (200/Slam)

Total: 450 points

Sharapova won:

Japan Open (35/Tier III)

Quebec City (35/Tier III)

Birmingham (35/Tier III)

Wimbledon (200/Slam)

Total: 305 points

Sugiyama won:

Linz (55/Tier II-)

Gold Coast (35/Tier III)

Total: 90 points

Serena Williams won:

Miami (100/Tier I)

Total: 100 points

Venus Williams won:

Charleston (80/Tier I-)

Warsaw (55/Tier II-)

Total: 135 points

Not much doubt about how results stand up in this classification. Henin-Hardenne is #1, Davenport and Myskina effectively tied for #2 (we'll call it a tie, since we might have awarded slightly different bonus points for tournaments if we did this again), and Mauresmo clearly last; she is, in fact, behind Kim Clijsters also.

Let's sum this up. We have four distinct measures apart from the WTA's: Won/lost, top opponents beaten, big events won, and points per event (divisor). Our four players stack up as follows in these categories:

Davenport: #1 in won/lost, #2 in divisor, #2 in big ones

Henin-Hardenne: #1 in big ones, #1 in quality opponents beaten, #1 in divisor, #2 in won/lost

Mauresmo: #3 or worse in all categories

Myskina: #2 in quality opponents beaten, #2 in big ones

The overall picture is clear: Henin-Hardenne is #1 in three of four categories, and #2 in the other. Seems pretty obvious that, under a better ranking system, she would still be #1. Davenport is probably #2, Myskina #3, Mauresmo #4.

We feel much better. Sure, the WTA rankings are a mess, but at least we know who should be #1.

So why is Mauresmo #1? In simplest terms, because the WTA ranking system ignores losses. As long as they add points, instead of looking at per-tournament results, this is liable to keep happening. (Recall that, in the won/lost statistics, Mauresmo had the most wins, though she also had a lot of losses.)

If we know the WTA, this is going to induce them to once more mess with the ranking system, probably by increasing Slam points yet again. Of course, that won't work; the players they want to penalize (Davenport and Mauresmo) had the most Slam wins. The logical course would be to go back to the divisor, under which Henin-Hardenne remains #1. But what's logic in dealing with the WTA?

Bob Larson´s tennisnewsletter.

With the announcement that Amelie Mauresmo is the WTA's new #1, there has been quite a bit of highly predictable dissatisfaction. It was a bit funny to sit down on Saturday morning and suddenly think, "OK, so who should be #1?" -- and not have a clue to the answer. Mauresmo's claim is certainly not overwhelming -- but the credentials of the other contenders (Lindsay Davenport, Justine Henin-Hardenne, Anastasia Myskina) aren't instantly overwhelming, either.

We of course can't make the WTA adopt a better ranking system -- we've been writing about this for years, and all they do is keep coming up with proposals for less accurate ranking systems. But we can at least try to reach an answer that satisfies us -- and maybe you.

So what we'll do is take the Top Four -- Mauresmo, Myskina, Davenport, and Henin-Hardenne -- and compare them under a variety of other measures.

Understand that this is based on results right now, this week. Lindsay Davenport still has a huge lead in the Race, and is the likely year-end #1. But we're looking at the current moment.

To start with, the rankings as the WTA lists them.

1 Mauresmo........4527

2 Myskina.........4155

3 Davenport.......4057

4 Henin-Hardenne..4004

Now let's try an obvious alternative: Won/Lost in the past twelve months. This is a measure, obviously, of who wins when she plays.

Rank..Player..........Wins..Losses.....%

1.....Davenport........56......8....87.9

2.....Henin-Hardenne...47......8....85.5

3.....Mauresmo.........57.....15....79.2

4.....Myskina..........53.....17....75.7

Obviously Davenport wins this measure, though Henin-Hardenne is close enough that it wouldn't be too bothersome if the Belgian were considered the #1. A minor caution is required: Davenport has played two Tier III events in the past twelve months (Strasbourg and Cincinnati), posting a 6-1 record. Myskina played Sopot, going 2-0 and withdrawing to play the Olympics, er, rest an injury. Tier III events really don't involve the same level of difficulty as stronger events, though, meaning that wins there mean less. Happily, it's easy to show that eliminating those events wouldn't affect the standings; Davenport would still have the best percentage and Myskina the worst.

For those interested in Slam won/lost, we have Mauresmo at technically 17-3.(85%, but of course she withdrew from the Australian Open), Henin at 11-2.(84.6%), Myskina 14-3.(82.4%), Davenport at 17-4.(81.0%). Thus Davenport, our winningest player overall, has the worst record at Slams! (Which goes far to explain the present situation, when you think about it.)

If we look at the players each has beaten, the following list shows each player's wins over then-Top Ten players over the past year, plus wins over players who are now Top Ten, plus the Williams Sisters. The number to the left of the player's name represents her ranking at the time she was beaten.

Davenport:

3. Mauresmo (Amelia Island)

5. Myskina (San Diego)

8. Dementieva (Sydney)

9. Zvonareva (Cincinnati)

[12. V. Williams (U. S. Open)]

[13. V. Williams (Los Angeles)]

[15. V. Williams (Stanford)]

[16. S. Williams (Los Angeles)]

Henin-Hardenne:

2. Clijsters (Australian Open)

2. Mauresmo (Olympics)

3. Myskina (Olympics)

4. Mauresmo (Sydney)

4. Davenport (Indian Wells)

5. Capriati (Los Angeles Championships)

5. Davenport (Australian Open)

5. Myskina (Indian Wells)

8. Dementieva (Filderstadt)

8. Myskina (Los Angeles Championships)

10. Rubin (Sydney)

[14. Kuznetsova (Indian Wells)]

[29. Kuznetsova (Dubai)]

[33. Kuznetsova (Australian Open)]

Mauresmo:

1. Henin-Hardenne (Amelia Island)

2. Henin-Hardenne (Los Angeles Championships)

8. Dementieva (Moscow)

8. Capriati (Berlin)

7. Myskina (Sydney)

9. Myskina (Philadelphia)

9. Dementieva (Los Angeles Championships)

9. Capriati (Rome)

10. Kuznetova (Olympics)

[14. Kuznetsova (Berlin)]

Myskina:

1. Clijsters (Leipzig)

2. Henin-Hardenne (Leipzig)

5. Capriati (Doha)

6. Capriati (Roland Garros)

7. Mauresmo (Moscow)

8. Sharapova (San Diego)

9. V. Williams (Roland Garros)

10. Rubin (Australian Open)

10. Dementieva (Roland Garros)

[11. Kuznetsova (Roland Garros)]

[20. Kuznetsova (Doha)]

[24. Sharapova (Indian Wells)]

[32. Sharapova (Australian Open)]

This presents a completely different picture. Davenport may be winning matches, but she isn't beating anyone in particular except Williamses in the throes of horrid slumps. We thought we would need some sort of formula for this, but it seems pretty clear that Henin-Hardenne -- the only one of the four to have beaten all three of her rivals -- is the best at beating top opponents, Myskina probably next, Mauresmo third, and Davenport dead last.

Now let's try another measure: Points per tournament. This, again, measures how effective a player is when she plays -- and includes quality points, measuring toughness of the opposition. That gives us this list:

Rank..Player........Points...Events...Pts/Event

1.....Henin-Hardenne..4004.....13........308.0

2.....Davenport.......4057.....15........270.5

3.....Mauresmo........4258.....18........236.6

4.....Myskina.........4194.....20........209.7

Once again, Henin-Hardenne comes out on top, and would still be #1 even if her total were divided by the old WTA minimum divisor of 14.

Let's try one more measure: "Winning the big ones." We'll examine all players who have won titles at the Tier II and higher levels. From that we'll create a ranking system of sorts, based solely on titles. We'll base this very loosely on the ATP Race system: We'll set a Slam as worth 200 points. That makes the Championships worth 150, and a Tier I worth 100. But we'll keep that wonderful invention, the Tier II, and make it worth 75 points. A Tier III we'll set equal to 35 points. But we'll also give bonuses and subtractions, somewhat arbitrarily: A Tier I or Tier II that is significantly stronger than average we will be worth 20 extra points; a weak Tier I or Tier II will lose 20 points. The list below will include all winners of Tier II or better events in the past year. Events are listed in the order played.

Bovina won:

New Haven (55/Tier II-)

Total: 55 points

Clijsters won:

Filderstadt (95/Tier II+)

Luxembourg (35/Tier III)

Los Angeles Championships (150/Championships)

Paris (55/Tier II-)

Antwerp (55/Tier II-)

Total: 390 points

Davenport won:

Pan Pacific (80/Tier I-)

Amelia Island (75/Tier II)

Stanford (55/Tier II-)

Los Angeles (75/Tier II)

San Diego (120/Tier I+)

Cincinnati (35/Tier III)

Total: 440 points

Dementieva won:

Shanghai (55/Tier II-)

Total: 55 points

Henin-Hardenne won:

Zurich (100/Tier I)

Sydney (95/Tier II+)

Australian Open (200/Slam)

Dubai (75/Tier II)

Indian Wells (120/Tier I+)

Olympics (80/Olympics)

Total: 670 points

Kuznetsova won:

Eastbourne (55/Tier II-)

U. S. Open (200 points)

Total: 255 points

Mauresmo won:

Philadelphia (55/Tier II-)

Berlin (80/Tier I-)

Rome (100/Tier I)

Canadian Open (80/Tier I-)

Total: 315 points

Myskina won:

Leipzig (95/Tier II+)

Moscow (80/Tier I-)

Doha (75/Tier II)

Roland Garros (200/Slam)

Total: 450 points

Sharapova won:

Japan Open (35/Tier III)

Quebec City (35/Tier III)

Birmingham (35/Tier III)

Wimbledon (200/Slam)

Total: 305 points

Sugiyama won:

Linz (55/Tier II-)

Gold Coast (35/Tier III)

Total: 90 points

Serena Williams won:

Miami (100/Tier I)

Total: 100 points

Venus Williams won:

Charleston (80/Tier I-)

Warsaw (55/Tier II-)

Total: 135 points

Not much doubt about how results stand up in this classification. Henin-Hardenne is #1, Davenport and Myskina effectively tied for #2 (we'll call it a tie, since we might have awarded slightly different bonus points for tournaments if we did this again), and Mauresmo clearly last; she is, in fact, behind Kim Clijsters also.

Let's sum this up. We have four distinct measures apart from the WTA's: Won/lost, top opponents beaten, big events won, and points per event (divisor). Our four players stack up as follows in these categories:

Davenport: #1 in won/lost, #2 in divisor, #2 in big ones

Henin-Hardenne: #1 in big ones, #1 in quality opponents beaten, #1 in divisor, #2 in won/lost

Mauresmo: #3 or worse in all categories

Myskina: #2 in quality opponents beaten, #2 in big ones

The overall picture is clear: Henin-Hardenne is #1 in three of four categories, and #2 in the other. Seems pretty obvious that, under a better ranking system, she would still be #1. Davenport is probably #2, Myskina #3, Mauresmo #4.

We feel much better. Sure, the WTA rankings are a mess, but at least we know who should be #1.

So why is Mauresmo #1? In simplest terms, because the WTA ranking system ignores losses. As long as they add points, instead of looking at per-tournament results, this is liable to keep happening. (Recall that, in the won/lost statistics, Mauresmo had the most wins, though she also had a lot of losses.)

If we know the WTA, this is going to induce them to once more mess with the ranking system, probably by increasing Slam points yet again. Of course, that won't work; the players they want to penalize (Davenport and Mauresmo) had the most Slam wins. The logical course would be to go back to the divisor, under which Henin-Hardenne remains #1. But what's logic in dealing with the WTA?

Bob Larson´s tennisnewsletter.