Mens Tennis Forums banner

1 - 20 of 52 Posts

·
Rankings Master
Joined
·
15,649 Posts
Discussion Starter #1 (Edited)
Iteration 3 of this stat after all of R4 matches have been played:

It appears Monfils falls down as last with the smallest percentage of W+FE, while Raonic maintains himself as the top aggressive player of the tournament.

QF version:


SF version:


zippar's QF/SF version:


Changes vs Slasher's version:
-I excluded the opponents' DFs
-I did not include the SW measure since when I checked some of the detailed data on the tournaments' websites, it did not seem completely reliable(I'm assuming they record specific data and then that data is aggregated automatically on the overall stats page for the match but sometimes the sum of specific data does not equal the aggregate data and it seems they plug for the difference...so, I will check a few more matches to see what they are doing exactly)

New things:
-Serve Advantage Measure
Rationale - some players that have stronger serves gain a significant advantage during the point. This advantage gives them more opportunity to hit winners or to put pressure and force errors from the opponent.
*The major assumptions
1. I assumed that the server gains on average a material advantage only for 1st serve, not 2nd serve
2. I assumed that the average between the points on 2nd serve(both serving and returning) give a good approximation of the rallies we need to focus on(those that for the most part start from a neutral position)
So, the calculation that i used is Serve Advantage = % of 1st serve points won(aces excluded) - the average % of points won on 2nd serve(both while serving and returning)

-Return Disadvantage Measure
Rationale - some players that have weaker returns start the point at a disadvantage. This disadvantage makes them less likely to hit W or force the opponent to make an error.
*The major assumptions
1. I assumed that the returner has a material disadvantage only when returning 1st serve, not 2nd serve
2. Again used the points on 2nd serves to get an approximation for the neutral rallies
So, the calculation that i used is Return DisAdvantage = % of 1st serve return points won(aces excluded) - the average % of points won on 2nd serve(both while serving and returning)

After I got the SA and RDA %, I reflowed the distribution of W,UE,FE etc.

So, I ended up with a 4 main measure - one without any adjustment for SA, RDA, one with each separate and one with both adjustments. I also included the % of total points each player won because it is reasonable to assume that a player with a higher percentage(more dominant in his matches) will have a different distribution of W,UE etc...
 

·
Registered
Joined
·
1,513 Posts
Re: Who hit and who pushed their way into RG QF [W+FE %]

Is it possible to add an extra column showing (W+FE-A-SW)%?
 

·
Registered
Joined
·
10,560 Posts
Re: Who hit and who pushed their way into RG QF [W+FE %]

How has Nadal hit one ace? That's pathetic considering his opposition.
It is surprising considering his serve ATM.
 

·
Registered
Joined
·
3,325 Posts
Re: Who hit and who pushed their way into RG QF [W+FE %]

The service winner stat for every match in RG is not maintained anywhere. Find me this stat, and sure.
There seems to be the stat on RG official page when you select individual match stats and then "Serve Stats" tab, but I really doubt that those stats are correct as for, for example, Gulbis - Federer match they are showing only 5 SW for Gulbis and 2 for Federer and that seems way too low to me. But the stat is there.
 

·
Registered
Joined
·
1,513 Posts
Re: Who hit and who pushed their way into RG QF [W+FE %]

The service winner stat for every match in RG is not maintained anywhere. Find me this stat, and sure.
Actually, they seem to have it under serve stats in the completed matches section. Give me a few minutes and I'll tell you the totals for each of the 8 players.

EDIT: No, I think they must be incorrect as R0B1S says. Raonic is listed with only eight service winners in his first four matches
 

·
Rankings Master
Joined
·
15,649 Posts
Discussion Starter #10
Re: Who hit and who pushed their way into RG QF [W+FE %]

There seems to be the stat on RG official page when you select individual match stats and then "Serve Stats" tab, but I really doubt that those stats are correct as for, for example, Gulbis - Federer match they are showing only 5 SW for Gulbis and 2 for Federer and that seems way too low to me. But the stat is there.
Thanks for pointing this out. Let me see if I can encompass this.
 

·
Registered
Joined
·
1,597 Posts
Re: Who hit and who pushed their way into RG QF [W+FE %]

At first glance Nadal with a highest percentage of FE? Wouldn't hurt to include percentages as well. :)

Not to be a dick but service winners remain an issue.
 

·
Rankings Master
Joined
·
15,649 Posts
Discussion Starter #12
Re: Who hit and who pushed their way into RG QF [W+FE %]

Alright, the service winners are now included in the table, and a new W+FE-A-SW column is used now for sorting.

Unfortunately for the people wishing this change would have any effect, there isn't much change.
 

·
Registered
Joined
·
1,022 Posts
Re: Who hit and who pushed their way into RG QF [W+FE %]

As I pointed out a couple of days ago, the stats do more harm than good unless we find at least *******(representative) for "quality of serve" and "quality of return". We should try to incorporate the data we can obtain(e.g. aces, % of 1st serve won, % of return points won etc.). It won't be perfect but it was never supposed to be when trying to measure an abstract concept.

EDIT: Why it does not let me use the word *****(representative)? :) weird
 

·
Registered
Joined
·
1,513 Posts
Re: Who hit and who pushed their way into RG QF [W+FE %]

Alright, the service winners are now included in the table, and a new W+FE-A-SW column is used now for sorting.

Unfortunately for the people wishing this change would have any effect, there isn't much change.
It's hard to believe those service winners are correct over 4 matches. Perhaps they are just counting those ones where the opponent gets the frame of the racket to the ball rather than all points where the opponent doesn't get the return in play after touching the ball with the racket. Maybe best to not use this stat in this case.
 

·
Registered
Joined
·
3,325 Posts
Re: Who hit and who pushed their way into RG QF [W+FE %]

I just can't believe those "Service winner" stats. If we define service winner as "a serve that is touched by the opponent, but not returned" then there is no way Raonic or Gulbis have just 8 of those in 4 matches.
 

·
Rankings Master
Joined
·
15,649 Posts
Discussion Starter #17
Re: Who hit and who pushed their way into RG QF [W+FE %]

I just can't believe those "Service winner" stats. If we define service winner as "a serve that is touched by the opponent, but not returned" then there is no way Raonic or Gulbis have just 8 of those in 4 matches.
That is why it is usually not included in any statistic.

A clear definition of service winner DOES NOT exist. If it would be included in the ITF rules of tennis, maybe it would be clear for tournaments. Some define it as the returner unable to return the ball in the direction of the opposite court, whilst others use a more general, returner unable to get the ball passed the net and imaginary net extension.
 

·
Registered
Joined
·
13,451 Posts
Re: Who hit and who pushed their way into RG QF [W+FE %]

Hmmm...so Raonic is not a servebot but a bashbot? Haha. I think these stats could be really good for Raonic.
Milos has huge power and attacks on nearly every point. Shouldn't it be expected that he'd be at the top of this chart?
 

·
Registered
Joined
·
1,513 Posts
Re: Who hit and who pushed their way into RG QF [W+FE %]

That is why it is usually not included in any statistic.

A clear definition of service winner DOES NOT exist. If it would be included in the ITF rules of tennis, maybe it would be clear for tournaments. Some define it as the returner unable to return the ball in the direction of the opposite court, whilst others use a more general, returner unable to get the ball passed the net and imaginary net extension.
The RG stats page also has a stat for baseline forced errors. Presumably total forced errors minus this number would give service winners under R0B1S's definition.
 
1 - 20 of 52 Posts
Top