Mens Tennis Forums banner

1 - 20 of 32 Posts

·
Registered
Joined
·
2,750 Posts
Discussion Starter #1
Good win Gael, into the quarters. :yeah:
He will play either Safin or Haase next, so this is a good opportunity for him.
21 aces for Gael. :worship:
 

·
Registered
Joined
·
770 Posts
He shouldn't have lost the first set. Spaz. But I like him, no matter how frustrating he is, so I'm glad he won.
 

·
Registered
Joined
·
21,230 Posts
He will get beat by haase or safin.
 

·
Registered
Joined
·
124 Posts
Keep it up, Gael! :yeah:
 

·
Registered
Joined
·
11,530 Posts
This guy has become in one of the biggest server in the game, right now in the top 5 for sure
Karlovic, Ljubicic, Roddick are all above him but then he is pretty much there. We will have to see about the consistency of Querrey and whether or not Isner will continue to bomb down 10 aces a set.
 

·
Registered
Joined
·
2,750 Posts
Discussion Starter #11
Karlovic, Ljubicic, Roddick are all above him but then he is pretty much there. We will have to see about the consistency of Querrey and whether or not Isner will continue to bomb down 10 aces a set.
Yes I agree. As far as hitting aces, lately Gael has been better than Ljubo, but Ljubo is a little more consistent in terms of getting the first serve in.
 

·
Registered
Joined
·
11,530 Posts
Yes I agree. As far as hitting aces, lately Gael has been better than Ljubo, but Ljubo is a little more consistent in terms of getting the first serve in.
This year Ljubicic is ahead by average aces per match by 14-11.5. Karlovic has a monstrous 19.8.

Guccione is also probably a better server than Monfils.
 

·
Registered
Joined
·
2,750 Posts
Discussion Starter #13
This year Ljubicic is ahead by average aces per match by 14-11.5. Karlovic has a monstrous 19.8.

Guccione is also probably a better server than Monfils.
Yeah I know he is. I'm talkin about lately though.
 

·
Registered
Joined
·
1,451 Posts
This year Ljubicic is ahead by average aces per match by 14-11.5. Karlovic has a monstrous 19.8.

Guccione is also probably a better server than Monfils.
To be fair, aces per set is a better measure than aces per match, as aces per match obviously are going to increase with the length of a match, giving an unfair advantage to players who lose one set over players who win in straights (if we're talking three-set matches). Aces per service game is an even better measure, as aces per set give an unfair advantage to players playing longer sets, thus punishing better returners.

For instance, if we compare Karlovic's ace count today to Isner's, we might think Isner had a better serving day than Karlovic, as he had 31 aces in his match, whereas Karlovic "only" had 19. Isner also leads in aces per set - 10+1/3 compared to 9.5. These stats, however are very much misleading, as Isner had 18 service games and three tie breaks. Even if we disregard the tie breaks, this gives Isner 1+13/18 aces per service game (about 1.722, that is). If we include the tie breaks as service games, which is more reasonable, it gives him 10+10/21 aces per service game (about 1.476). Karlovic only had 9 service games, giving him 2+1/9 aces per service game (about 2.111). Clearly, this shows Karlovic had a better serving day than Isner (or simply that Karlovic is a better server). The best measure, however, is aces per first serve. It takes away all bias that comes from the amount of first serves hit. It is, however, based on the faulty assumption that second serve aces never occur. That obviously isn't the case, but second serve aces are so uncommon, even among the really huge servers, that I think it's a reasonable simplification. I compared Karlovic to Roddick using this measure and the results are astounding - Karlovic hits 0.334423930824959 aces per first serve on average (more than an ace on every third first serve, which is pretty sick), compared to "only" 0.223101265822785 for Roddick. These values are based on rather big samples (over 60 matches for both players), so they are pretty reliable.
 

·
Registered
Joined
·
6,071 Posts
if u want a better measure, then consider aces per points
To be fair, aces per set is a better measure than aces per match, as aces per match obviously are going to increase with the length of a match, giving an unfair advantage to players who lose one set over players who win in straights (if we're talking three-set matches). Aces per service game is an even better measure, as aces per set give an unfair advantage to players playing longer sets, thus punishing better returners.

For instance, if we compare Karlovic's ace count today to Isner's, we might think Isner had a better serving day than Karlovic, as he had 31 aces in his match, whereas Karlovic "only" had 19. Isner also leads in aces per set - 10+1/3 compared to 9.5. These stats, however are very much misleading, as Isner had 18 service games and three tie breaks. Even if we disregard the tie breaks, this gives Isner 1+13/18 aces per service game (about 1.722, that is). If we include the tie breaks as service games, which is more reasonable, it gives him 10+10/21 aces per service game (about 1.476). Karlovic only had 9 service games, giving him 2+1/9 aces per service game (about 2.111). Clearly, this shows Karlovic had a better serving day than Isner (or simply that Karlovic is a better server). The best measure, however, is aces per first serve. It takes away all bias that comes from the amount of first serves hit. It is, however, based on the faulty assumption that second serve aces never occur. That obviously isn't the case, but second serve aces are so uncommon, even among the really huge servers, that I think it's a reasonable simplification. I compared Karlovic to Roddick using this measure and the results are astounding - Karlovic hits 0.334423930824959 aces per first serve on average (more than an ace on every third first serve, which is pretty sick), compared to "only" 0,223101265822785 for Roddick. These values are based on rather big samples (over 60 matches for both players), so they are pretty reliable.
 

·
Registered
Joined
·
2,750 Posts
Discussion Starter #16
To be fair, aces per set is a better measure than aces per match, as aces per match obviously are going to increase with the length of a match, giving an unfair advantage to players who lose one set over players who win in straights (if we're talking three-set matches). Aces per service game is an even better measure, as aces per set give an unfair advantage to players playing longer sets, thus punishing better returners.

For instance, if we compare Karlovic's ace count today to Isner's, we might think Isner had a better serving day than Karlovic, as he had 31 aces in his match, whereas Karlovic "only" had 19. Isner also leads in aces per set - 10+1/3 compared to 9.5. These stats, however are very much misleading, as Isner had 18 service games and three tie breaks. Even if we disregard the tie breaks, this gives Isner 1+13/18 aces per service game (about 1.722, that is). If we include the tie breaks as service games, which is more reasonable, it gives him 10+10/21 aces per service game (about 1.476). Karlovic only had 9 service games, giving him 2+1/9 aces per service game (about 2.111). Clearly, this shows Karlovic had a better serving day than Isner (or simply that Karlovic is a better server). The best measure, however, is aces per first serve. It takes away all bias that comes from the amount of first serves hit. It is, however, based on the faulty assumption that second serve aces never occur. That obviously isn't the case, but second serve aces are so uncommon, even among the really huge servers, that I think it's a reasonable simplification. I compared Karlovic to Roddick using this measure and the results are astounding - Karlovic hits 0.334423930824959 aces per first serve on average (more than an ace on every third first serve, which is pretty sick), compared to "only" 0,223101265822785 for Roddick. These values are based on rather big samples (over 60 matches for both players), so they are pretty reliable.
Dang, how long were you working on that man?
 

·
Registered
Joined
·
1,451 Posts
if u want a better measure, then consider aces per points
Well, that would actually not be as good a measure, as more than 95 % of all aces are first serve aces. Aces per serves hit (which is what I assume you mean when you say "aces per points") would in many cases just reflect a higher first serve percentage and not a higher ability to hit aces.
 

·
Registered
Joined
·
1,451 Posts
Dang, how long were you working on that man?
You mean the Karlovic-Roddick comparison? It took a while, actually. I had to check the amount of first serves and the amount of aces for all the matches using the ATP stats feature and enter them into Excel manually. I also included first serve points won (Karlovic dominates Roddick there too, obviously), so I had to check the amount of those too. I had to do it because some retard claimed Roddick had a better serve than Karlovic. While the very thought of that is silly, I had to scientifically prove him wrong, so I constructed a hypothesis test, testing the hypothesis that they had an equal proportion of aces per first serve against the one-sided alternative that Karlovic would have a higher proportion. I don't remember at what levels the test was significant, but I recall it was a pretty good one, not surprisingly.
 

·
Registered
Joined
·
2,750 Posts
Discussion Starter #19
You mean the Karlovic-Roddick comparison? It took a while, actually. I had to check the amount of first serves and the amount of aces for all the matches using the ATP stats feature and enter them into Excel manually. I also included first serve points won (Karlovic dominates Roddick there too, obviously), so I had to check the amount of those too. I had to do it because some retard claimed Roddick had a better serve than Karlovic. While the very thought of that is silly, I had to scientifically prove him wrong, so I constructed a hypothesis test, testing the hypothesis that they had an equal proportion of aces per first serve against the one-sided alternative that Karlovic would have a higher proportion. I don't remember at what levels the test was significant, but I recall it was a pretty good one, not surprisingly.
:bowdown:
 
1 - 20 of 32 Posts
Top