ATP & WTA rankings

[ashkor87]

Nice work!

[ashkor87]

Trouble is, some of these are higher ranked players like Swiatek beating someone in the first round of a slam . Not sure it tells me much.

[ashkor87]

I have no idea and no access to data to do this but it would be worth repeating with, say, players ranked within 10 places of each other, and leaving out the top 3 or 5, who will always beat anybody...

[atlpam]

It would be interesting to break into groups of 20. Stats where the higher ranked player is in the top 20, ranked 21-40, 41-60, etc.

[Suliso]

12-14% advantage for the higher ranked player is almost nothing, but I suspect it's skewed by matches between players in the 40-100 range. Difference there is really small, probably as little as 4-5%.

[ponchi101]

Trouble is, some of these are higher ranked players like Swiatek beating someone in the first round of a slam . Not sure it tells me much.

I have no idea and no access to data to do this but it would be worth repeating with, say, players ranked within 10 places of each other, and leaving out the top 3 or 5, who will always beat anybody...

You like to pick the example that suits your hypothesis. Here, choosing Swiatek to win a match means, as you say, very little. Picking a specific player to beat another is what we have been discussing: you need a lot of data. Let's remember (you and I are old enough) Graf-McNeil at Wimbledon (I forget the year). How much money would you have placed on McNeil? Me, zero. But McNeil won, which is the reason we remember that match as an epic upset.
What this means is simple: RANKINGS are a reliable piece of information. Of course the lower ranked player will beat the higher ranked player with certain frequency; otherwise, the rankings would be frozen perennially. But the higher ranked player will beat him more times than not. Not #3 beating #5; those are specific players. IN GENERAL.
See it this way. for 2022, there were 2,600 matches in the ATP (approx.). Assume you are given $100 for every match, to bet on the winner, and you have no info other than the ranking. Well, you can get all sophisticated and stuff, but it is obvious that the strategy is to bet on the higher ranked player, ALWAYS. Do that, and you would end up with a nice (using my numbers) $74K. Just by betting on that data.

Now, your second question, which is translated to this: how finely do rankings matter? By that I mean (and you mean): the closer the ranking of the two players, the least important the ranking is for the result. A match between #2 and #3 is a 50/50 match, IF you go only by the rankings. So, as you asked:
Exclude the top 3 ranked players, and look only at matches between players within 10 ranking spots. For 2022 (ATP):
173 Higher ranked player
203 Lower ranked player

So, by these constraints, yes, ranking does not determine the outcome. That is 46% for the HIGHER ranked player. But I suspect that we have a distortion there: Novak spent a very long portion of the year ranked outside the top 3. And he lost very little.

[ponchi101]

12-14% advantage for the higher ranked player is almost nothing, but I suspect it's skewed by matches between players in the 40-100 range. Difference there is really small, probably as little as 4-5%.

I am surprised by your statement. It is NOT a 12-14% advantage. IF the higher ranked player has a 64% chance of winning, the lower ranked player has a 36% chance. And that is a 28 points spread, because sports are a zero-sum game. I said that to Ashkor.
About your guess. Matches excluding players ranked above 40 (2022 ATP):
580 (higher ranked)
422 (lower ranked)
So sure, it gets smaller, but that is still 57.9%, on a sample that is large enough to be representative.

[Suliso]

Indeed I didn't define it correctly. Spread is as you say much larger, but one could also say solely from a higher ranked player's perspective that he/she is winning 14 matches per 100 more than if it was a random chance (50-50 with enough data points). Either way it's a solid advantage, but not an overwhelming one. Nadal won 83% of matches last year.

[ponchi101]

We are also dealing here with a circularity. How do you get a high ranking? By winning a lot. And if you win a lot, you get a high ranking (more or less; if you win a lot of 125's, your ranking goes high, but not in the same way).
Sure, Nadal won 83% of matches last year. And Mr #100, won less than 50%. Is it a standard curve, maybe skewed a bit? Maybe a Pareto effect?
My initial question was simply that: do rankings matter? I say, at least a little. But I am not going to bet my house on IGA, next time she plays ARYNA, simply based on their rankings. But given #20 vs #80, I say go with #20. You have a little bit better chance.

[Suliso]

But given #20 vs #80, I say go with #20. You have a little bit better chance.

More than a little, but that we could say before too

Nevertheless thanks for the analysis. Always nice to see some numbers!

[ashkor87]

So Krejcikova thinks it should be 'big 4'
https://www.wtatennis.com/news/3122377/ ... a-big-four

[ashkor87]

Evidently she doesn't think much of rankings either, neither does Sabalenka it seems..

[ponchi101]

She currently does not hold any slams.
But I said it somewhere in some topic: she is in the conversation, and she is playing solid tennis. Her losses this year have been understandable. To me, she is certainly a more solid player than Garcia, Ons, Coco, Daria or Sakkari, at the moment.
If she does not believe in rankings, it makes sense. She reached #2 and right now she is outside the top ten. So her mentality can't be "I am ranked properly". She must think "I am better than my ranking".
I bet Rafa does not think he is the best 12th player in the world.

[ti-amie]

I am ashamed to say I had to think about who the "Big 3" in the WTA are. Sabalenka, Rybakina and of course Iga. Sabalenka is playing great right now but she can still do a Sakkari and rip defeat from the jaws of victory. Tonight's match will go a long way to establish Krejcikova as a "Big 4" member.

[Deuce]

This talk of a 'big 3' or 'big 4' in the WTA is pure nonsense.
How long were Federer, Nadal, and Djokovic at the top of the game? Slightly longer than the few MONTHS that Swiatek, Sabalenka, and Rybakina have been at the top, I believe.
People tend to exaggerate an awful lot these days (everything is 'awesome' and 'epic', etc.).
The 'big 3/big 4' talk in the WTA should stop and resume only in about 5 years IF there is any long-term consistency at the top at that point.