Sampras Stich 1992 Wimbledon QF stats

[abmk]

The ONLY time these two met in a grand slam. Stich was the defending champion and sampras hadn’t found much success on grass before this wimbledon.

I hadn’t seen this match before . One of Sampras’ best performances easily.. His serving was top-notch and returning very good. Sampras played at his very best, stich was not quite as sharp . Stich’s serving could’ve easily been better and though his returning wasn’t bad, it could’ve been easily better.

The first set, stich started out a bit slow and got broken in his first service game due to some excellent returning from Sampras. Missed a backhand pass on BP very slightly with Sampras serving for the set.

The second set was a demolition with stich getting broken twice. The third set, the difference was a single break, when stich went off and Sampras capitalized with his excellent returning .

The highlight shot of the match was a stunning running low backhand pass by Sampras at 4-2 in the 2nd set to get a breakpoint. Then he got the break with a forehand return winner.

Stich grumbled quite a bit to the umpire about the line calls more than once.

These are the stats :

This match again shows so clearly what I’ve always felt:

Just the no of aces is no indicator of the quality of serving. The return errors should also be shown, for a more accurate measure. Sampras had ‘just’ 9 aces, but stich got back only 43% of the serves.

Both players SnVed on every serve, 1st and 2nd and there were very few rallies. Which is why there were so few unforced errors.

Sampras had 7 return passing winners, 1 of the FH wing and 6 of the BH wing. He also had 2 other passes, both of the BH wing. Therefore in total 9 passing shot winners, 8 of the BH wing and 1 of the FH wing

Stich had ZERO return passing winners, but had 3 passing shot winners, all of the FH wing

Please note that these are passing shot winners from the baseline only, I didn’t mark the shots at the net when one player passed the other at the net.


The ATP stats:

http://www.atpworldtour.com/Share/Match-Facts-Pop-Up.aspx?t=540&y=1992&r=5&p=S402

The ATP site has the stats wrong. There is a huge difference in the total no of points won.

The return % in the stats table was flashed at the end of the match ( I didn’t compute it ). With the help of that , we get total no of points as 146 ( I got 145 ), ATP has it at 166.

This is how I calculated the total no points from the returns % in stats.

Sampras had 63% return % in, which means he didn’t put 37% of the serves back into play. Now that 37% comprises of 15 return errors, 7 aces and 1 service winner by stich. Total = 23.

Therefore no of points “played” on stich’s serve = 23*100/37 =62

He had 3 DFs, so that makes his no of service points = 65

Stich had 43% returns in, which means he didn’t put 57% of the serves back into play.Now that 57% comprises of 32 return errors,9 aces and 4 service winners form Sampras. Total=45

Therefore no of points “played” on sampras’ serve = 45*100/57=79

He had 2 DFs, so that makes his no of service points = 81

Therefore total no of points = 65+81=146

 

[Moose Malloy]

thanks for posting, I remember Sampras being considered the favorite for the title that year by the oddsmakers after this match(understandable considering the surprising semifinalists that year)

Interesting to see Sampras have so many more winners off the bh side than off the fh side.

I attended Wimbledon that year & was surprised how unknown Sampras was among the British fans, despite already being a USO champion (he was on Court 16 for his 4th round match)
I was told by many fans that Wimbledon was basically the only tennis tournament that received any media attention in England.

how long does it take you to get all these stats? do you watch the match more than once to get all the info?

 

[krosero]

With the help of that , we get total no of points as 146 ( I got 145 ), ATP has it at 166.

This looks like one of the many matches in the 1990s in which the ATP included the aces and df's twice.

There were 21 aces/df's in the match according to the ATP. Subtracting 21 from the ATP's 166 total points yields, in fact, your 145.

 

[krosero]

I've got just one published stat (probably since it was a qf). Per the Associated Press, Sampras served 8 aces, “and his first-serve percentage was 83%.” I'm sure that's a mistake, though, and probably meant to refer to points won on first serve.

The ATP has Sampras winning 85% of the time on first serve -- winning 47 of 55 points (including 9 aces). But ABMK's stats confirm that the ATP included the aces and df's twice. So if we presume that Pete's 9 aces were all first serves, and we subtract them once, it leaves Pete winning 38 of 46 points on first serve: 83%.

The ATP has Sampras making 55 of 93 first serves (59%), so following a similar process would leave Sampras making 46 of 82 first serves (or 56%). Note: in this case his 2 double-faults as well as his 9 aces are subtracted from his total number of attempted first serves.

I don't think that adjusting the ATP figures like this will always work out to the exact true figures; and it involves the assumption that all aces are first serves (Sampras after all was known to serve aces on second serves). But it gives at least a good idea of what the true figures are -- if you know (as we do thanks to ABMK) that the ATP stats are off for a particular match.

 

[abmk]

thanks for posting, I remember Sampras being considered the favorite for the title that year by the oddsmakers after this match(understandable considering the surprising semifinalists that year)

Interesting to see Sampras have so many more winners off the bh side than off the fh side.

I attended Wimbledon that year & was surprised how unknown Sampras was among the British fans, despite already being a USO champion (he was on Court 16 for his 4th round match)
I was told by many fans that Wimbledon was basically the only tennis tournament that received any media attention in England.

how long does it take you to get all these stats? do you watch the match more than once to get all the info?

well most of those were return passing winners - stich served majorly to that side ..... as pete's returning was sharp, he got those winners.

I'm not surprised at hearing that sampras was 'unknown' to the British fans that time at all ... He hadn't done anything much at wimbledon before ...

As far as the stat collection goes, I do it mostly watching the match once, pausing here and there if required.

Several of those categories, I compute at the end from the components ( winners, total UEs etc .. ) ....

As far as return errors are concerned, I mark down the error as R in the forced or unforced errors category if it is a return error. Then later compute the total no of return errors from this.

Usually the pausing/resuming and the computation and counting etc results in taking half an hour to an hour more than the actual length of the match

 

[abmk]

This looks like one of the many matches in the 1990s in which the ATP included the aces and df's twice.

There were 21 aces/df's in the match according to the ATP. Subtracting 21 from the ATP's 166 total points yields, in fact, your 145.

I've got just one published stat (probably since it was a qf). Per the Associated Press, Sampras served 8 aces, “and his first-serve percentage was 83%.” I'm sure that's a mistake, though, and probably meant to refer to points won on first serve.

The ATP has Sampras winning 85% of the time on first serve -- winning 47 of 55 points (including 9 aces). But ABMK's stats confirm that the ATP included the aces and df's twice. So if we presume that Pete's 9 aces were all first serves, and we subtract them once, it leaves Pete winning 38 of 46 points on first serve: 83%.

The ATP has Sampras making 55 of 93 first serves (59%), so following a similar process would leave Sampras making 46 of 82 first serves (or 56%). Note: in this case his 2 double-faults as well as his 9 aces are subtracted from his total number of attempted first serves.

I don't think that adjusting the ATP figures like this will always work out to the exact true figures; and it involves the assumption that all aces are first serves (Sampras after all was known to serve aces on second serves). But it gives at least a good idea of what the true figures are -- if you know (as we do thanks to ABMK) that the ATP stats are off for a particular match.

yeah, looks like the ATP counted the aces and DFs twice, which is why those figures are off.

sampras did not have any second serve aces in this match ...

 

[krosero]

Both players SnVed on every serve, 1st and 2nd and there were very few rallies. Which is why there were so few unforced errors.

Including their double-faults (which is the convention today and, I'm pretty sure, was in '92), Sampras had only 6 UE, Stich 4.

So in a match like this all the difference is in the forced errors and the winners.

About your net stats: what are you doing with aces and service winners? Are you counting those as net points?

Please note that these are passing shot winners from the baseline only, I didn’t mark the shots at the net when one player passed the other at the net.

I usually count any groundstroke winner as a pass, so long as the other person is at net. But if the person hitting the groundstroke is up at net, he's essentially playing a net point. So does it makes sense to credit him with a passing shot? I'm beginning to think no; the way you're doing it makes a lot of sense. A player is either playing a net point or a baseline point; crediting him with a net approach and a passing shot at the same time may be a contradiction.

 

[abmk]

Including their double-faults (which is the convention today and, I'm pretty sure, was in '92), Sampras had only 6 UE, Stich 4.

So in a match like this all the difference is in the forced errors and the winners.

yeah, even I do it that way ( include the DFs in the UE count ). I forgot to add the DFs to the UE count this time ....

About your net stats: what are you doing with aces and service winners? Are you counting those as net points?

no, not counting those as net points

I usually count any groundstroke winner as a pass, so long as the other person is at net. But if the person hitting the groundstroke is up at net, he's essentially playing a net point. So does it makes sense to credit him with a passing shot? I'm beginning to think no; the way you're doing it makes a lot of sense. A player is either playing a net point or a baseline point; crediting him with a net approach and a passing shot at the same time may be a contradiction.

that's what I thought, but how do the official statisticians actually do it ?

 

[krosero]

no, not counting those as net points

Definitely agree with that.

Can't remember the particular matches, but I think Moose and I have found indications of official net stats being done both ways.

that's what I thought, but how do the official statisticians actually do it ?

That I don't know, but if past experience is any guide, they probably disagree among themselves.

 

[quest01]

That was a great match, I actually have that match on DVD plus many more Sampras matches. I believe Sampras mentioned in his autobiography that Stitch was his toughest rival or had the most potential to challenge him if he played longer on tour. The stats on that match was unreal, the unforced error count was crazy. I'm surprised none of them had any overhead winners seeing that they both served and volleyed.

 

[abmk]

Definitely agree with that.

Can't remember the particular matches, but I think Moose and I have found indications of official net stats being done both ways.

But why ? What does a net approach have to do with the returner not being able to get his racquet on the ball (ace) or barely get his racquet on the ball ( service winner ) ? even if he didn't approach the net, he'd win the point anyways ...

yes, return errors can be forced by coming to the net, but that isn't applicable to aces/service winners !

That I don't know, but if past experience is any guide, they probably disagree among themselves.

hmm, ok

 

[krosero]

But why ? What does a net approach have to do with the returner not being able to get his racquet on the ball (ace) or barely get his racquet on the ball ( service winner ) ? even if he didn't approach the net, he'd win the point anyways ...

yes, return errors can be forced by coming to the net, but that isn't applicable to aces/service winners !

I agree; it's just we found one match where it seemed that the official net stats included everything. I think they were BBC stats, I just can't recall which match.

 

[krosero]

You know what, ABMK, I'm looking through my notes and I can't find an instance of aces and service winners being included in net stats. So it might be faulty memory. I was thinking there were two ways of doing net counts, and one is the way you and I agreed on (I think Moose also agrees): counting unreturned serves, except the aces and service winners. That must be a common way of doing it, and for example, I think NBC did it that way for the 98 W final (I have a thread on it).

I'm looking over old boxscores and I'm seeing a second way of doing it, which is to actually not count unreturned serves.

Here's an example: 1987, Cash-Lendl.

Cash served on 81 points, Lendl on 131. Both men were coming in behind all serves, first and second.

If we subtract from Cash’s total his 40 unreturned serves (including 4 aces)and 2 doubles, we get 39 points in which he approached the net. He also came in 4 times during Lendl's service games. That gives us 43 net rushes by Cash – an exact match with a boxscore in the New York Times.

Similarly if we subtract from Lendl’s service points his 35 unreturned serves (including 6 aces) and 6 doubles, we get 90 points in which he could have rushed the net – an exact match with the boxscore. And he didn't get to net on Cash's serve. However he did stay back once on his own second serve, so there's a minor discrepancy.

There was a Sports Illustrated boxscore for Sampras-Rafter in 2000 that seemed to be doing the same thing with its net stats. It's hard to tell for sure, but that looks true for a lot of boxscores at Wimbledon -- at least in the SV days. Today you can't work on the assumption that players are coming in behind all serves, so it's harder to tell.

Anyway, you would think that at least some return errors take place because of the pressure of the incoming server. So that's why I mention these boxscores: they're examples where all return errors seem to be taken out of the net stats. Whether that was done as a fixed rule, or because the statistician genuinely judged all the return errors to be due to the difficulty of the serve alone, is another question. But it's interesting.

 

[abmk]

You know what, ABMK, I'm looking through my notes and I can't find an instance of aces and service winners being included in net stats. So it might be faulty memory. I was thinking there were two ways of doing net counts, and one is the way you and I agreed on (I think Moose also agrees): counting unreturned serves, except the aces and service winners. That must be a common way of doing it, and for example, I think NBC did it that way for the 98 W final (I have a thread on it).

I'm looking over old boxscores and I'm seeing a second way of doing it, which is to actually not count unreturned serves.

Here's an example: 1987, Cash-Lendl.

Cash served on 81 points, Lendl on 131. Both men were coming in behind all serves, first and second.

If we subtract from Cash’s total his 40 unreturned serves (including 4 aces)and 2 doubles, we get 39 points in which he approached the net. He also came in 4 times during Lendl's service games. That gives us 43 net rushes by Cash – an exact match with a boxscore in the New York Times.

Similarly if we subtract from Lendl’s service points his 35 unreturned serves (including 6 aces) and 6 doubles, we get 90 points in which he could have rushed the net – an exact match with the boxscore. And he didn't get to net on Cash's serve. However he did stay back once on his own second serve, so there's a minor discrepancy.

There was a Sports Illustrated boxscore for Sampras-Rafter in 2000 that seemed to be doing the same thing with its net stats. It's hard to tell for sure, but that looks true for a lot of boxscores at Wimbledon -- at least in the SV days. Today you can't work on the assumption that players are coming in behind all serves, so it's harder to tell.

Anyway, you would think that at least some return errors take place because of the pressure of the incoming server. So that's why I mention these boxscores: they're examples where all return errors seem to be taken out of the net stats. Whether that was done as a fixed rule, or because the statistician genuinely judged all the return errors to be due to the difficulty of the serve alone, is another question. But it's interesting.

yeah, I thought of the second way too - thought that would've been done in several matches ... But I find the first method more logical ...

 

[krosero]

Baltimore Sun:

"To do well here and in the French [reaching the quarterfinals] has shown me that I can play well on any surface," said Sampras, who won the U.S. Open in 1990. "You know, there is a little bit of pressure on me. The first year I came here, I wasn't experienced on grass, never played grass-court tennis. I did not play well. I was going to come back here every year and give it my best shot. I kind of got the breaks this week and last week."

He certainly got the breaks yesterday against Stich. He broke Stich in the second game of the opening set and never looked back. Sampras broke him twice in the second set, including once at love. He broke Stich once more in the final set.

But it was the way Sampras tore into his serve, especially his second serve, that frustrated Stich. He only had nine aces, but his accuracy (94 percent first serves in the second set) was just as effective.​

 

[NatF]

Baltimore Sun:

"To do well here and in the French [reaching the quarterfinals] has shown me that I can play well on any surface," said Sampras, who won the U.S. Open in 1990. "You know, there is a little bit of pressure on me. The first year I came here, I wasn't experienced on grass, never played grass-court tennis. I did not play well. I was going to come back here every year and give it my best shot. I kind of got the breaks this week and last week."

He certainly got the breaks yesterday against Stich. He broke Stich in the second game of the opening set and never looked back. Sampras broke him twice in the second set, including once at love. He broke Stich once more in the final set.

But it was the way Sampras tore into his serve, especially his second serve, that frustrated Stich. He only had nine aces, but his accuracy (94 percent first serves in the second set) was just as effective.​

94% first serves damn that's impressive.

 

[NonP]

Some more stats taken by moi. (Thanks again to @abmk for uploading the entire match to YouTube.) First off I don't see how The Baltimore Sun got 94% for Pete on 1st serves in the 2nd or any other set. (I suspect they meant % of 1st-serve points won, but if they did they probably should've pick another term than "accuracy.") Here are the correct 1st-serve %s by set:

PS - 57.1% (20/35), 54.5% (12/22), 56% (14/25)
MS - 47.4% (9/19), 70% (14/20), 62.5% (15/24)

So Pete notched 56.1% (46/82) and Mike 60.3% (38/63) overall. (These numbers didn't seem to be in the OP so I thought I'd list them here.)

After 2-all in the 2nd set Pete got all of Mike's serves back in play even though most of them were 1st serves, so Stich began the 3rd set differently by taking considerable pace off and almost rolling in all of his five 1st serves. He did win the game comfortably but I thought he went too much for the % and it was only a matter of time before Pete found his groove again. Apparently Mike himself agreed and began his very next service game with a big 1st serve... which I only later found out was voided... which screwed up my counting and forced me to redo my numbers again.

1st-serve % on break points:

PS - 100% (2/2)
MS - 40% (2/5)

While serving for the 1st set at 5-3 Pete stumbled on his 1st BP by missing a fairly easy BH volley at deuce, but saved it thanks to a strong 1st serve and Stich overhitting his BH pass just a couple inches long. And he erased his next BP in the 2nd set with an ace (see below).

Stich OTOH didn't fare as well. In the 1st set he DF'ed on his very 1st BP, and in the next set he threw a 2nd serve on BP #2 which Pete was glad to return with a crushing CC BH to break. Mike did get a 1st serve in on his next BP but again Pete's return was up to the task (this is the FH return winner abmk described in the OP). In the 3rd set Pete set up three BPs with a block return that miraculously floated inside the corner, and paradoxically Stich's 2nd serve finally got the job done on the 1st point but his following 1st serve down the middle proved too safe and Pete broke for the last time to go up 3-2.

Aces by set (there was no 2nd-serve ace by either):

PS - 3, 3 (1 on BP), 3
MS - 2, 2, 3

Double faults:

PS - 1, 0, 1 (on first match point, but he was up 40-0)
MS - 2 (1st on the very 1st BP of the match), 1, 0

And now we come to the granddaddy of them all, unreturned serves again by set but this time I'm also breaking them down by 1st-/2nd serve (in brackets):

Set 1
PS - 57.1% (20/35) [13/20, 7/15]
MS - 31.6% (6/19) [4/9 (didn't hit one on 1st serve till his very 1st ace, but then he fired 2 aces in a row), 2/10]

Set 2
PS - 54.5% (12/22) [9/12, 3/10]
MS - 30% (6/20) [4/14, 2/6]

Set 3
PS - 56% (14/25) [9/14, 5/11]
MS - 41.7% (10/24) [8/15, 2/9]

Match
PS - 56.1% (so Pete won the same % as his 1st-serve %, heh) (46/82), [67.4% (31/46), 41.7% (15/36)]
MS - 34.9% (22/63), [42.1% (16/38), 24% (6/25)]

Any time a player wins over 50% of his points directly behind the serve he's all but unbreakable, and 56.1% is a stellar # even for Sampras. But despite Stich's subpar % by comparison I thought both guys' 2nd serves were generally excellent and often had trouble telling which one is which. Obviously Pete was the superior returner here by far.

A couple parting thoughts. The German station in the YT video flashed this graphic in the middle of the 2nd set showing the tournament's fastest servers with their ace totals (presumably up to the 4th round):

Forget - 210.8 km/h, 89 aces
Stich - 209.2 km/hr, 46
Goran - 207.6 km/hr, 100
Pete - 207.6 km/hr, 78

And we tend to believe Boris when he says (as he did the following year) Stich served bigger than Pete, but that's not what I saw in this particular match. In fact another graphic (at 1-4 with Stich serving in the 1st set) had Pete's and Mike's respective average 1st- and 2nd-serve speeds clocking in at 116 and 99 mph and at 111 and 101 mph. Needless to say with Pete playing at such a high level Stich needed to out-serve or out-return him to have any chance, and given the above numbers it's hard to see any version of Stich getting the better of Sampras on either front at their one and only encounter at Wimbledon.

Pete wasn't yet quite as ruthless at the net as he'd become later, but it's still striking to see just how quick and natural he already was on grass.

And on a final and less serious note, Stich complained a lot to the umpire about what he felt were close calls missed by the linespeople, but what's really funny is that he somehow didn't object when I felt he had better reason to be upset. Wonder how he would've handled the challenge system these days.