Stats for 1998 W final (Sampras-Ivanisevic)

krosero

Legend
Sampras d. Ivanisevic, 6-7 (2), 7-6 (11-9), 6-4, 3-6, 6-2

This match actually had more winners from ground strokes than from volleys/overheads, which I think is one reason I enjoyed it more than their 1995 semifinal -- but that doesn't necessarily mean that this match was higher quality.

This was Pete’s 5th Wimbledon and 11th Slam title. It was the only fifth set he ever played in a Slam final. 1998 was his last year finishing #1 on the computer.

Ivanisevic twice had points to go up by two sets, both in the second-set tiebreak. On each one he got a second serve and returned it into the net with his two-hander. He also saved two set points held by Sampras, by drawing return errors. Sampras then got another set point, but this time he got to serve. And the pattern held: another return error. All five set points in the tiebreak were decided that way.

Ivanisevic had beaten Krajicek 15-13 in the fifth set in the semis.


By my count:

Sampras had 12 aces and 8 doubles.
Ivanisevic had 32 aces and 20 doubles.

Sampras got a return error from Ivanisevic with 59 serves -- of which I judged 11 as service winners.

Ivanisevic got a return error from Sampras with 51 serves -- of which I judged 14 as service winners.

Sampras hit 40 clean winners apart from service: 7 FH, 12 BH, 6 FHV, 14 BHV, 1 OH.

Ivanisevic hit 40 clean winners apart from service: 9 FH, 13 BH, 11 FHV, 6 BHV, 1 OH.

Sampras hit 8 service return winners. He got Goran’s first serve with one forehand and one backhand. The other winners were all backhands off Goran’s second serve.

All the returns were passes. On top of those he had 9 passing shots (five from the forehand).

Ivanisevic hit 8 service return winners, too. He got Pete’s first serve with one forehand and two backhands. The other winners were all backhands off Pete’s second serve: and this was the shot that failed him twice on set point in the second tiebreak.

All the returns were passes. On top of those he had 13 passing shots (7 from the backhand).

Ivanisevic pulled off four passing shots out of five points to break Sampras at 2-3 in the fourth set. That was probably the most exciting game of the match.

So Goran ended up with more winners from ground strokes than from volleys, by 22 to 18. Sampras himself almost had that distinction, too, with 19 and 21.

Overall the match featured more ground stroke winners than overheads/volleys, 41 to 39. Some of the overheads – I did not mark them – may have been taken on a bounce rather than in the air, but that would not have changed the big picture.

In their 1995 semifinal, only a third of Pete’s winners, and slightly less than half of Goran’s, were ground strokes.

In '98, each man had fewer aces and more doubles than in '95.

In break points the only significant difference was that Ivanisevic faced 14 in ’98 compared to only 5 in 1995.
 
Last edited:

krosero

Legend
Stats by NBC

Goran had served 161 aces and a tournament-record 78 doubles in his first six matches.

At exactly 2-4 in the third, 221 points had been played, only 16 going to 4 hits or more. At 1-all, 30-15 in the fifth, 306 points had been played, only 23 going to 4 hits or more.

NBC counted only clean aces; they didn't count serves barely grazed by the receiver's racquet. In their running count they got up to Sampras' 11th ace and were silent about his last ace, which shows that the ATP's total number for the match (11) is mistaken and that there is no confusing issue about judgment calls on near-aces.

In the first set, NBC had Sampras serving at 51%, with 0 aces, 4 doubles, 13 winners, 9 unforced errors and 0 of 6 break points converted. Ivanisevic served at 60%, with 13 aces, 9 doubles, 25 winners, 15 unforced errors and 0 of 3 break points converted.

For that set I have Sampras hitting no aces and 8 clean winners, so NBC probably gave him 5 service winners (I gave him 6). I counted 13 aces and 8 other clean winners by Ivanisevic, so NBC probably gave him 4 service winners (I gave him 5).

At 2-all in the fourth, NBC had Sampras at 49 winners. I’d given him 8 aces and 31 other clean winners, so it looks as if NBC had given him 10 service winners (I’d given him 11). NBC had Ivanisevic at 68 winners. I had him at 27 aces and 32 other clean winners, so it looks as if NBC had given him 9 service winners (I’d given him 12).

At exactly 2-all in the fifth, NBC had Sampras serving at 54% with 10 aces, 8 doubles, 57 winners, 19 unforced errors and 2 of 12 break points converted. Ivanisevic was at 56% with 32 aces, 19 doubles, 81 winners, 41 unforced errors and 2 of 9 break points converted.

At that point in the match, I had Sampras at 10 aces and 37 other clean winners, so it looks as if NBC had given him 10 service winners (I’d given him 11). I had Ivanisevic at 32 aces and 39 other clean winners, so it looks as if NBC had given him 10 service winners (I’d given him 14).


[...]


At 3-all, 15-all in the first set, NBC had Sampras winning 11 of 18 “net points”, Ivanisevic 12 of 19. At love-1 in the third, they had Sampras at 54 of 77, Ivanisevic at 42 of 62.

I counted net points myself up to the display of the first graphic. Like NBC, I have each player losing 7 net points – always when he was in front of the service line and made an error or got passed cleanly.

The difficulty is in the points won. I have Sampras winning 3 net points, and Ivanisevic 7, in clear-cut ways: the player who won the point either hit a winner of some kind after rushing forward, or his opponent made an error.

If I include all the serves that drew errors (since Pete and Goran were following all their serves in), then Sampras moves up to 15 points won, and Ivanisevic to 14. But if I then remove the few serves that I judged personally as service winners, my numbers line up exactly with NBC’s: 11 of 18 net points won by Sampras, 12 of 19 by Ivanisevic.

Note to Moose Malloy: maybe, just maybe, there is a hint here of something I once suggested: that net points can include serve-and-volley approaches in which you judge that there is a chance to return the serve successfully, but should not include whatever you judge as service winners, since those are serves that win points outright without any added pressure from the net-rushing.

In case you're interested, of the errors by the net-rusher that I marked down on my sheet, none were errors made on approach shots; all were ground strokes or volleys in front of the service line. There was no instance, in this small sample, of an error made on an attempted approach shot. So there’s no confusing issue here concerning how approach errors should be counted in net stats.

NBC's net stats look similar to SI's final numbers.
 
Last edited:

krosero

Legend
Stats from Sports Illustrated

The boxscore and article are here: http://sportsillustrated.cnn.com/tennis/1998/wimbledon/

2448583455_2317e89973_o.jpg


Compared with the ATP stats, Sampras has 12 aces, rather than 11. Ivanisevic’s service percentage is 56 rather than 57 at the ATP site (where it rounds down to 57). Sampras won 80% of points started on his first serve rather than 82%, and Ivanisevic is at 83% rather than 87%.

And of course SI’s stat for return points won is mistaken. It might be just an online typo. I wonder if the printed magazine originally had another number.

I have exactly the same number of “Service Winners” as SI, if I add all the return errors to the aces.

This is the first time I can remember that aces were included as service winners. So now I’ve seen 3 different definitions of service winners:

1) serves judged unreturnable;

2) all serves that the receiver gets a racquet on but cannot put back in play (we usually call these “unreturned serves” or “return errors”); and

3) all unreturned serves, including aces

Moose first showed me this boxscore. I found our discussion from before. I guessed that "service winners" included only the return errors, but he read it correctly as including aces.
 

krosero

Legend
And of course SI’s stat for return points won is mistaken.
Actually just realized, because I did the service return winners late: I have 8 service return winners by each player. This boxscore has 8 "return points won" by each player.

Could that be the cause of the error?
 

Moose Malloy

G.O.A.T.
Sampras hit 40 winners apart from service: 7 FH, 12 BH, 6 FHV, 14 BHV, 1 O.

Ivanisevic hit 40 winners apart from service: 9 FH, 13 BH, 11 FHV, 6 BHV, 1 O.

At exactly 2-all in the fifth, NBC had Sampras serving at 54% with 10 aces, 8 doubles, 57 winners, 19 unforced errors and 2 of 12 break points converted. Ivanisevic was at 56% with 32 aces, 19 doubles, 81 winners, 41 unforced errors and 2 of 9 break points converted.

Interesting to see that Goran had such a big lead on Sampras in winners(though that was mainly due to the big difference in aces)
I wonder how often Sampras had less winners than his opponent in matches that he won. I would think not very often.

I just got the Sampras-Goran match from the '96 Year End Championship, judging from the stats on atp tennis, this one could have some high winner numbers.

Still surprised we haven't come up with high service winner numbers on a Goran match yet, guess I underestimated the rest of his game.

NBC's net stats look similar to SI's final numbers.

looks like the net stats that flashed onscreen in the '95 match were way off. I wonder what their definition of net points was.

Ivanisevic pulled off four passing shots out of five points to break Sampras at 2-3 in the fourth set. That was probably the most exciting game of the match.

Yeah, I remember that game, it may have been the best non service game of Goran's game, incredible passes.
 

krosero

Legend
ATP stats mistaken

NBC provided the total number of points played (306) as of 1-all, 30-15 in the fifth set. I counted 27 more points till the end, for a total of 333, compared to the 404 reported at the ATP site.

As I noted above, the ATP's service percentages are only a little bit different from those in the SI boxscore, and NBC had similar numbers. I did not doubt that the ATP stats were at least close to accurate, and did not think to compare NBC's figure for total points with the ATP's.

I completed the count by getting the first-serve percentages, so I now have these new stats:

In the fifth set, Sampras served at 71% (making 15 of 21 first serves). Ivanisevic served at 52% (making 11 of 21). Sampras won 28 points overall, Ivanisevic 14.

We no longer have (good) figures for the service percentages in their 1995 semifinal. But in this match, with four games to go NBC put Sampras at 54%, Ivanisevic at 56%.

And SI put Sampras at 55%, Ivanisevic at 56%.

Not a particularly good percentage for Pete.
 

krosero

Legend
I have to take back my original argument that NBC made a mistake in its stats.

The boxscore and article are here: http://sportsillustrated.cnn.com/tennis/1998/wimbledon/

2448583455_2317e89973_o.jpg


[....] I have exactly the same number of “Service Winners” as SI, if I add all the return errors to the aces.

This is the first time I can remember that aces were included as service winners. So now I’ve seen 3 different definitions of service winners:

1) serves judged unreturnable;

2) all serves that the receiver gets a racquet on but cannot put back in play (we usually call these “unreturned serves” or “return errors”); and

3) all unreturned serves, including aces

Well now I know of a fourth definition of "service winner", after getting my own stats for the 2005 USO final. When CBS displays the "service winners" during that match, they appear to be including the aces along with all the serves they had judged to be service winners -- but not all unreturned serves.

That's actually a definition of "service winner" that makes a lot of sense and should have been intuitive. When a player is credited with a total # of winners, he has winners apart from service, and he has service winners. Those service winners consist of his aces in addition to serves that are judged to be winners. Judged as such, they end up in his total winner count.

But that is not the definition in Sports Illustrated's boxscore, and maybe I had that boxscore too much in mind. Whatever the reason, I believe I missed what NBC was doing. I ended up supposing that NBC was making two different sets of judgment calls -- and that they were way off on Sampras' service winners.

I've taken my original argument out of my post, but I'm quoting it here for the record:

However in this match NBC actually presented a stat called “service winners”, and it wasn’t the same: this one was the same as all the return errors, a much larger number. A graphic at 4-all in the second set put Ivanisevic at 24 service winners. I know this represented all return errors because NBC then displayed another graphic showing that Ivanisevic had put 66 serves into play, with only 24 being “returned.” They arrived at that number by subtracting from 66 his 18 aces up to that point and the 24 “service winners” they had just reported.

But I think that this stat, 24 "service winners", is slightly inaccurate, because by my count Ivanisevic had gotten only 22 return errors from Sampras. And I think that my number is the correct one, because I proofed all my stats for the first two sets, and because my numbers for the aces and the return errors for the entire match line up exactly with stats provided by Sports Illustrated (see below).

And for Sampras, NBC was wildly mistaken. They put him at only 9 service winners. (When he drew an ordinary return error on the next point, Enberg said that it was his “tenth service winner”). But that number is impossible. By my count he had already gotten 30 return errors from Ivanisevic.

I wondered if perhaps NBC’s number for Sampras was using the other, common definition of service winners as those unreturned serves that the receiver gets a racquet on but that are judged as unreturnable. As I estimated above, if we use that definition it looks like NBC had given Sampras 5 service winners in the first set according to their judgment; so at 4-all in the second set it’s plausible that they could have been up to 9. Yet by my estimate NBC was still only at ten, as of 2-all in the fourth set, so I’m inclined to think of Pete’s “9 service winners” as a simple mistake.

It sounds strange that NBC would be using two definitions of “service winner,” for two different purposes. But that’s what appears to be necessary, because as noted I proofed my work for the first two sets and there is no way to account for the winner-totals that NBC gave for the first set if you count only clean aces and other clean winners. Something else has to be added – not only to the first set but to all of NBC’s midmatch winner-totals. Whatever NBC was adding, it tracks well with my own numbers of (judged) service winners.

This was too complex and speculative. Simply include the aces in NBC's "service winners", and restrict the remaining serves to judgment calls. Then the numbers make sense.

They have Sampras at 9 service winners; I had him at 11 (that is, 4 aces and 7 judgment calls).

They had Ivanisevic at 24 service winners; I had him at 25 (including 18 aces and 7 judgment calls).

This way there is no need to suppose that NBC used different counts of service winners for different purposes. And there is no need to suppose they made a mistake.

In any case, my apologies to Leo Levin for saying that his work on this match was "wildly mistaken."
 
Last edited:

krosero

Legend
Other stats

The Boston Globe had Sampras making 19 unforced errors for the match, and only 3 in the last set. And since NBC already had him at 19 as of 2-all in the fourth set, it looks like he made none in the last four games.

I couldn't find the total winners, but at 2-all in the fifth, NBC put Sampras at 57, Ivanisevic at 81. For the remainder of the match I've got Sampras hitting 2 aces and 4 other winners, including a judgment call on a BH. And I've got Ivanisevic hitting one BH return winner.

From that I'd guess that NBC gave Sampras a match total of 63 winners, and Ivanisevic 82.
 

380pistol

Banned
The Boston Globe had Sampras making 19 unforced errors for the match, and only 3 in the last set. And since NBC already had him at 19 as of 2-all in the fourth set, it looks like he made none in the last four games.

I couldn't find the total winners, but at 2-all in the fifth, NBC put Sampras at 57, Ivanisevic at 81. For the remainder of the match I've got Sampras hitting 2 aces and 4 other winners, including a judgment call on a BH. And I've got Ivanisevic hitting one BH return winner.

From that I'd guess that NBC gave Sampras a match total of 63 winners, and Ivanisevic 82.

You'd be right NBC had Sampras finishing with 63 winners and 19 unforced errors, but yet again NBC (or maybe it's just me) could not get the final stats on his opponent (Ivanisevic this time).

The 1994 final had Pete finishing with 49 winners, 10 unforced errors and 17 aces, while they had Ivanisevic at 27 winners, 7 unforced erros and 16 aces ifor the 1st set. Samras according to NBC had 20 winners, 6 unforced and 9 aces at the same juncture.
 

krosero

Legend
Sampras made 8 of 14 first serves in the tiebreaks (57%). By tiebreak:

3 of 4
5 of 10 (he won all 5 and lost only one point on a second serve)

Ivanisevic made 11 of 15 first serves in the tiebreaks (73%). By tiebreak:

5 of 5 (no mini-breaks)
6 of 10 (he won all 6)

So in the tiebreaks Goran won every point when he got a first serve in.

From 4-5 in the second tiebreak neither player got a serve back until Sampras finally did at 9-9 (a run of 9 straight unreturned serves). Sampras then finished the tiebreak with another unreturned serve, so there were unreturned serves on all 5 set points played in the breaker.
 

Nadal_Power

Semi-Pro
I guess this is why so many people called Sampras/Ivanisevic boring tennis.

I know that in 1994 final points were insane fast, but this was not much better

Will make full stats soon, but in first set there were 2 rallyes with 6 strokes, in second one ''long'' rally of 8 strokes, no more than 5 strokes in 3rd and 4th set, and one with 7 strokes in 5th
 

Indio

Semi-Pro
The Boston Globe had Sampras making 19 unforced errors for the match, and only 3 in the last set. And since NBC already had him at 19 as of 2-all in the fourth set, it looks like he made none in the last four games.

I'm under the impression that failed passing-shot attempts are seldom, if ever, considered to be unforced errors. Would that be correct?
 

krosero

Legend
From 20 points in tie break 13 were service winners, Goran made 1 double fault so in 6 points we saw return
This stat would look a little better except that Goran missed a couple of makeable returns of second serves.

Looking forward to your '94 stats. Feel free to create a new thread, that match deserves its own.
 
Last edited:

NonP

Legend
Thought some of you might be interested in TA's stats (compiled by user Edo) on this match:

http://www.tennisabstract.com/charting/19980705-M-Wimbledon-F-Goran_Ivanisevic-Pete_Sampras.html

FYI I'm generally not a fan of TA's own charts - for example this one on the '90 USO final has Agassi's serves at 99 rather than the correct 100 and his 1st- and 2nd-serves at 75 and 24 respectively when they should be 77 and 23, and in fact if you add up Dre's 99 and Pete's 72 you're left one point short of USA Today's total of 172 - but this one looks pretty solid.

P.S. TA has this one annoying habit of showing serve %s by set but not the raw numbers. Wonder if there's actually an easy way to look up the latter.
 
Top