Stats guys: 55% games won; 70% matches won

[Maui19]

I was looking at my stats for this season, and I don't really know what to make of it, so I am hoping some of you stats guys could offer some opinions. This season, when playing at my rating level, I won 70% of my matches, and 55% of my games. Now, I'm pretty happy I've won 70% of my matches--that's an excellent outcome for me. But I was surprised to see I won just 55% of my games. Then I was thinking maybe that is pretty normal--that even lopsided matches tend to be closer than you would think. So do people who win 30% of their matches typically win 45% of their games?

Is it typical in tennis matches that the game score tends to be much closer than the match score?

 

[schmke]

Yes, statistically one would generally expect games won % to be closer to 50% than matches won % in most cases.

Consider someone who wins all their matches. Their match won % is 100% but unless they were to win all their matches 6-0,6-0 their games won % will certainly be less than 100%.

Now, this won't always be the case. Consider someone that wins exactly 50% of their matches. It is possible but not likely that they win exactly 50% of their games, so in this case the general rule wouldn't hold. But the farther the match win % is from 50%, the more likely the game win % will be closer to that break even number.

 

[J_R_B]

If you look at points, it's even closer. If you win like 55% or 60% of the points, you win like 85% or 90% of matches, or something like that.

 

[Topspin Shot]

I think winning 65% of points gives you a legitimate shot at a double bagel assuming you and your opponent aren't big servers.

 

[Carsomyr]

I think winning 65% of points gives you a legitimate shot at a double bagel assuming you and your opponent aren't big servers.

Not an expert on league stats or anything, but I think Federer won something like 54-55% of the points played against him in 2005: went 81-4. The further you move away from matches when scoring (i.e. into sets, games, and points) - the stats will trend closer toward the mean, regardless of whether you're blowing people out or not.

EDIT:

If you look at points, it's even closer. If you win like 55% or 60% of the points, you win like 85% or 90% of matches, or something like that.

Haha, didn't see this, but that's what I'm getting at.

 

[RobFL]

I can attest to the other end of the curve. When going up against a 5.5 i can win about 35% of the points and it still equates to a 0, 1 thumping, every time.

 

[LeeD]

Well, in your defense, at least you got one out of 3 points, and one game, against a 5.5.
That's better than 98% of us can do, so kudos to you.

 

[schmke]

So, I calculated a few scenarios, so here goes.

Scenario 1: You win 70% of the points in your service games and 40% in your return games.

Points won: 55%
Service games won: 87%
Return games won: 29%
Games won: 59%
Sets won: 78%
Matches won: 88%

38% of the sets will be a 6-3 or 6-4 win for you

Scenario 2: You win 60% of the points in your service games and 45% in your return games.

Points won: 52%
Service games won: 71%
Return games won: 39%
Games won: 55%
Sets won: 65%
Matches won: 72%

30% of the sets will be a 6-3 or 6-4 win for you.

Scenario 3: You win 53% of the points in your service games and 49% in your return games.

Points won: 51%
Service games won: 57%
Return games won: 43%
Games won: 52%
Sets won: 56%
Matches won: 59%

25% of the sets will be a 6-3 or 6-4 win for you.


In all cases, match % > set % > game % > points %.

 

[geom]

I run a contract time group of 4.5/5.0 guys. One year I logged the match results into a database and tracked stats. Here's how they looked...

 

[dode]

I like stats a lot and use them a lot for my teams. From my experience, wins and losses tell only part of the story. Game win percentage can really say a lot (and in some cases more) too. From my experience, the better players in a given NTRP level win between 55% and 60% of their total games played given a decent number of data points. If they are winning 60% to 65% of games played, they have a very good chance of getting bumped up, and over 65% they are really playing out of level. Obviously these numbers are dependent on who you play and who you play with in doubles, but I have found them to be quite true over the last three years at two different levels up through the national level. Here is an example from one of my winter doubles only teams.

We lost in the first round of single elimination district playoffs because three of those guys were playing up a level and I had to play two of them just to field a team.

I would say the stats you listed actually make sense.

John

 

[Maui19]

Great info guys. Thanks very much--this was exactly the kind of info I was hoping to find. Very interesting.