Stats for 1987 Wimbledon final (Cash-Lendl)

[krosero]

I watched this match a few times to get different kinds of stats from it – including, for the first time, forced and unforced errors.

I never get tired of watching it. From the time I saw it live, it has always thrilled me to see Cash’s skill, athleticism and concentration against Lendl.

Score: 7-6 (5), 6-2, 7-5

Cash served 4 aces and 2 doubles. He got a service return error from Lendl 36 times (of which, by my own judgment, 6 were service winners).

Lendl served 6 aces and 6 doubles. He got a service return error from Cash 29 times (of which, by my own judgment, 12 were service winners).

NBC credited Lendl with 7 aces because they gave him an ace on what today we would call a service winner.

Lendl threw his racquet down in frustration while serving at 1-3 in the second set, a game in which he doubled three times.

The stats on the return errors are deceptive. Cash and Lendl each got a similar number of return errors, but Lendl’s service games were much longer than Cash’s (see the service percentages below). In other words, Cash earned free points with his serve more frequently than Lendl did.

Cash served at 67%, making 54 of 81 first serves.

Lendl served at 61%, making 80 of 131 first serves.

Cash’s first-serve percentage by set:
29/37 (78%)
12/16 (75%)
13/28 (46%)

Lendl’s first-serve percentage by set:
45/63 (71%)
13/31 (42%)
22/37 (59%)

Cash won 66 of 81 points on his serve, Lendl 77 of 131.

Cash won his serve 9 times at love, Lendl 3 times.

Cash won 120 points overall, Lendl 92.

Cash won 4 of 18 break points that he held. Lendl won 1 of 1.

Cash got his first serve into play on the only break point he faced. Lendl got his first serve in on 8 of 18 break points (only 44% of the time).


Cash hit 44 non-service winners: 6 FH, 16 BH, 11 FHV, 6 BHV, and 5 overheads.

Lendl hit 27 non-service winners: 7 FH, 2 BH, 8 FHV, 7 BHV, and 3 overheads.

(At 1-all in the third, the screen went blank for one point in NBC’s coverage, but the sound remained. It sounded like a quick exchange at the end, with Lendl apparently finishing it at net. I credited him on that point with a forehand volley winner).

Cash’s winners by set: 21, 12, 11
Lendl’s winners by set: 15, 5, 7

Cash is nicely balanced, with exactly half of his winners coming off his ground strokes.

Cash returned Lendl’s second serve for a winner 7 times with his backhand, and once with his forehand. He returned Lendl’s first serve cleanly once off each side.

At 4-5 in the first set, when Lendl had not yet been passed cleanly on a service return, it happened to him four times – three times when serving to Cash’s backhand. Lendl managed to save the set then, but he saw two more backhand returns pass by him in the tiebreak. All of these involved second serves.

I could not help wondering how much better Lendl might have done in this match if he’d stayed back on at least some of his second serves and tried to work his way in.

A good example is the one time that Lendl stayed back – on a second serve at 1-5, love-15 in the second set. Again, he went to Cash’s backhand. But since he didn’t rush the net, he got an ordinary return down the line, instead of a sharp backhand directed at his feet or out of his reach. He took that return on the bounce and pummeled it with his forehand; Cash fell trying to get it.

Per the New York Times, Lendl had a losing percentage of points won at net – 43% (compared to 60% for Cash). This is partly because he could not win points coming in behind his second serve: at 1-all in the second set, per NBC, Lendl was winning just 33% of such points.


Apart from service returns, Cash passed Lendl another 6 times – four times off the backhand. He also hit 6 lob winners – again, four times off the backhand.

So all 22 of Cash’s ground stroke winners were passing shots or lobs.

By contrast, only 3 of Lendl’s ground strokes were passing shots: a forehand lob and two service return winners, all in the first set. Lendl did pass Cash one other time by taking a sitter in the air (with a smash, from no-man’s land). And, of course, he forced some volley errors from Cash, just as Cash did to him.


Not counting service returns, I have Cash making 23 forced errors (including 7 volleys), and Lendl 15 (including 6 volleys).

Not counting doubles, I have Cash making 5 unforced errors, and Lendl 15.


Cash won 28 of 43 approaches (13 of 22 in the first set, 7 of 7 in the second, 8 of 14 in the third).

Lendl won 41 of 89 approaches (21 of 44 in the first set, 7 of 18 in the second, 13 of 27 in the third).

It may seem strange that I got Lendl approaching the net twice as much as Cash. But so did the New York Times, as I’ll describe in the post below.

Both men followed all their serves into the net. And each man got to net on his opponent’s serve only a handful of times. So basically they went to net about as frequently as they served. And Lendl served a lot more points than Cash did: 131 compared to 81.

That is how he came to net so much more than Cash did.


Some stats from NBC.

In the first set, Cash served at 78%, Lendl at 71%. At 1-love in the third, Cash was serving at 77%, Lendl at 62%.

In the first set, Cash lost 8 points on serve, Lendl 24. At 1-love in the third, Cash had lost 8 points on serve, Lendl 40.

At 1-all in the second, Cash was winning 80% of points on his second serve, Lendl just 33%.

In the first set, Cash made 2 unforced errors, Lendl 8. At 1-love in the third, Cash had made 3 unforced errors, Lendl 18.

NBC gave Cash 2 unforced errors in the first set; I gave him only one. They gave Lendl 8, while I gave him 7.

In the second game of the third set, NBC had Lendl at 18 unforced errors. I had him only at 9.

[edited a typo on the double-faults]

 

[krosero]

Published stats

I learned a few things from studying a New York Times boxscore for this match.

1) The stats published in the Times – and perhaps elsewhere, since I don’t know who actually produced them – refer to all service return errors as service winners.

2) Points that ended with service return errors were not included in the count of the net approaches -- not in the stats published by the Times. However, NBC did include such points when counting net approaches.

3) “Placement winner” remains a problematic category, even as late as 1987-88.


Here is the boxscore:

SERVICE WINNERS

The “service winners” in the box line up exactly with my own count of the return errors. This is the first time I’m seeing this, where all return errors are referred to as service winners.

Even the year before, the Times distinguished between service winners and other return errors. In reference to Becker’s victory over Lendl in the final:

The West German had 15 aces and 13 clear service winners. In addition, there were numerous points that he won when Lendl's return of serve sailed wide or sank into the net.

So this case looks unique, to me. However, MooseMalloy dug up a boxscore from Sports Illustrated for the 1998 final between Sampras and Ivanisevic, in which I thought that all the return errors were being counted as service winners.

But the only way to know is to watch the match and count them.


PLACEMENT WINNERS

The “placement winners” in the Times box for Cash and Lendl are a lot higher than the clean winners that I counted. I have Cash at 44 non-service winners, while this box has him at 52 placement winners. I have Lendl at 27 non-service winners, while this box has him at 41.

There was a similar problem with Times boxscores from 1976-77, concerning the U.S. Open. But in that case, those stats were simply unreliable, with service percentages off by double-digits, and the reported number of total points in the match falling short of the actual number by dozens.

In the Cash-Lendl box, everything looks mostly okay. Cash is listed as having 5 aces when he actually had 4 (by my count and NBC’s), but that is no big deal. The real problem is that this box has 22 more non-service winners than I counted.

I don’t think you can chalk that up to mere sloppy counting. Yet I don’t know the reason for it.

I did confirm, however, that there's an excess of 23 points in the stat box.

The box is missing only one category: forced errors. (And by that I mean errors on strokes other than service returns; the errors on those are already accounted for, as “service winners”). If you had that category, you would be able to add up all the numbers in the box and see if there's an excess over the total number of points reported. That total number is 212, with Cash winning 120 -- the same as my own count. So I did my own count of the forced and unforced errors (see post above), in order to fill in the missing category.

Of course, the difference between a forced and an unforced error is a judgment call. But the total number of combined errors is not. I counted 58. If I add enough forced errors (29) to the unforced errors in the stat box to yield 58, I find that the numbers in the stat box exceed the total points by 23.

This excess is made up of two things. One is the extra ace I mentioned, credited by mistake to Cash. The remaining 22 extra points are the 22 extra placement winners.

That confirms that my count of the winners was correct. But it’s a mystery why the placement winners exceeded the number of winners by 22. Perhaps some of the forced errors (22 of 29, to be exact) were also counted as placement winners. I simply don’t know.

But I do think that the term “placement winner” refers to clean winners. There are three other boxscores available online in the Times archives for this period: the 1988 U.S. Open final between Graf and Sabatini and the two men's semifinals, Wilander-Cahill and Lendl-Agassi. I can't find Wilander-Cahill, but I got an exact match between my count of winners and the placement winners reported for Sabatini, Lendl and Agassi. Graf was reported at 26 placement winners while I counted only 21. Relatively speaking, I don’t consider that too big a problem – partly because two of the points she won are missing from my DVD.

Overall, it looks like “placement winner” means what we would call a clean winner and that the Cash-Lendl boxscore is either working with some definition that I can’t account for when I look at the numbers, or is mistaken – by a lot.


NET STATS

In the stat box Lendl won 39 of 90 approaches, Cash 26 of only 43. That was over the course of 33 games. By contrast, after only 22 games in his semifinal win over Connors, Cash had already approached 75 times, per NBC.

The difference is there because NBC was including service return errors, as I confirmed by doing my own net stats for the Connors match.

When NBC had Cash winning 23 of 32 net points as of the third game of the second set, I had him at 13 of 16 approaches. If I add the 9 times that he drew a return error, he goes up to 22 of 25.

When NBC had Cash at 53 of 75 as of the third game of the third set, I had him at 38 of 56. If I add the 13 times that he drew a return error, he goes up to 51 of 69.

Not an exact match, but close enough.

One thing I can’t say for sure is whether NBC counted all return errors among the net points (assuming, of course, that the server followed his serves in). It looks that way from the stats. But the reason I hesitate is that I have no reason to think that NBC regarded all return errors as service winners; by far the most common practice is to judge only some return errors as service winners. Those winners are then commonly reported next to the aces; you see that a lot in articles from the 80s. And since I have no reason to think that aces were included among net rushes (anymore than double-faults were), I think it’s possible that NBC might have judged some return errors as service winners and explicitly excluded those from the net stats. After all, if a serve is good enough to be judged as service winner, the fact that the player followed the serve into the net had nothing to do with inducing the error. At least, that is how I see a service winner.

Unfortunately the Cash-Connors match is not a good test, because there was only one serve in the whole match (by Cash) that I would have judged a service winner.

And NBC gave no net stats for the final.

But in the final you can see how the Times-published stats were produced.

Cash served on 81 points, Lendl on 131.

If we subtract from Cash’s total his 4 aces, 2 doubles, and the 36 return errors that he induced, we get 39 points in which he approached the net, since he didn’t stay back on any of his serves. And by my count he came to net four times on Lendl’s serve. That gives us 43 net rushes by Cash – an exact match with the stat box.

Similarly if we subtract from Lendl’s service points his 6 aces, 29 service winners and 6 doubles, we get 90 points in which he rushed the net – an exact match with the stat box. And he didn’t get to net on Cash’s serve. However, as noted above, he stayed back once on a second serve. So it’s not really an exact match, just close enough.

When I counted the approaches myself I got exactly the same totals: 43 by Cash and 89 by Lendl.

One discrepancy is in the number of approaches that they won. The stat box has Cash winning only 26, and Lendl only 39; I have Cash winning 28 and Lendl 41.

Again, close enough.

 

[hoodjem]

I remember that match. Cash played great.

One peculiar thing that has stuck in my mind about it was how Cash seemed to get down to every ball no matter how low, he would dig it out of the grass. Indeed it seemed that Cash played in a bent-knee crouch the whole match. Particularly when up at the net, Lendl couldn't get anything by him because he was so low.

It taught me how important it is to bend your knees in tennis, particularly when volleying.

 

[krosero]

Cash-Connors

Score: 6-4, 6-4, 6-1

SERVICE

A chief difference between Cash's semifinal and final was that Connors -- when he got his racquet on Cash's serve -- returned it better than Lendl did.

Lendl made 36 service return errors over 33 games. Connors made only 15 errors over 27 games.

Connors himself got 13 return errors from Cash, compared to the 29 that Lendl got. That is no surprise: Lendl's serve was clearly superior to Connors'.


Cash served 10 aces against Connors (compared to only 4 against Lendl), and 2 doubles.

Connors served 1 ace and 4 doubles.

Cash served at 52%, making 37 of 71 first serves. NBC seemed to come back too late for one of his serves; I took a guess and marked it down as a first serve. My count was close to NBC’s percentage for him (60%) after the first game of the second set.

Connors served at 72%, making 58 of 81 first serves. I took a guess on five of his serves, marking three of them down as first serves. This way happens to line up my count with NBC’s percentage for him (74%) as of the second game of the second set.

Cash’s first-serve percentage by set:
13/21 (62%)
10/29 (34%)
14/21 (67%)

Connors’ first-serve percentage by set:
17/23 (74%)
28/39 (72%)
13/19 (68%)

Cash won 55 of 71 points on his serve, Connors 44 of 81.

Cash won 92 of 152 points in the match.

Cash won 5 of 8 break points, Connors 1 of 3.

Cash got his first serve into play on 1 of 3 break points, Connors on 4 of 8.


WINNERS

Cash made 35 winners: 5 FH, 4 BH, 9 FHV, 9 BHV, and 8 overheads.

Connors made 16 winners: 7 FH, 7 BH, 1 FHV, and 1 BHV.

Cash had more winners from volleys/smashes than from ground strokes, whereas in the final his winners were split down the middle between those two categories. That is because Connors returned better than Lendl and forced him to hit volleys; this match, I think, is a better showcase of his volleying skill than the final.

I can't find a copy of his quarterfinal over Wilander, but I recall reading at the time that he had a very high number of winners.


Some stats from NBC:

At 1-love in the second, Cash was serving at 60%, Connors at 74%.

At 1-love in the second, Cash had 14 winners and 11 unforced errors (mostly off the forehand), Connors 3 and 3.

Cash made 25 volley winners and only 1 unforced volley error in the whole match.

At 1-all in the second, Cash had won 23 of 32 approaches, Connors 2 of 4. At 2-love in the third, Cash had won 53 of 75, Connors 12 of 19.

 

[Moose Malloy]

Not counting service returns, I have Cash making 23 forced errors (including 7 volleys), and Lendl 15 (including 6 volleys).

Not counting doubles, I have Cash making 5 unforced errors, and Lendl 15.

Impressive that you did errors, how did you know how to count these?

I've heard that missed passing shots don't count as unforced errors, so if both players were S&V on 1st & 2nd serves, I'm wondering what kinds of shots were counted as unforced. just volleys?

do you have your error count by set? maybe I'll re-watch a set, just to see what how a 'forced' & 'unforced' error is counted.

 

[krosero]

Impressive that you did errors, how did you know how to count these?

I've heard that missed passing shots don't count as unforced errors, so if both players were S&V on 1st & 2nd serves, I'm wondering what kinds of shots were counted as unforced. just volleys?

I'd never realized that before, but looking over my sheet I see that most of the unforced errors were volleys. I do have Lendl making 3 ue's on his BH, and Cash 2 on his FH.

Generally the ground stroke errors were forced, as I judged them, because the opponent was applying pressure at the net. And yeah, the players followed all their serves in, but there are still a couple of ways you can get some unforced errors that are ground strokes. I think there might have been a few baseline rallies (such as rallies were on grass) after someone got driven back to the baseline with lobs. Also there may have been a time when someone was at net, and they get a ball that lands on the ground near them, but they blow it with too much power or getting too cute with it.

One shot I remember clearly from the first set. Lendl forces a Cash volley to sit up, and he drives it straight into the net with his backhand. I recall putting that down as unforced. I can see following the rule that if anyone is at net, it's pressure, therefore the error is forced. But IMO it's a judgment call. I've heard statisticians describe the unforced error as a shot where you think that a player of Lendl's capability is "supposed" to make, given his skill. That shot was make-able for Lendl. Having Cash at net should force him to try for more than if he was in a baseline rally, but I think to just drive a sitter straight into the tape is an unforced error.

do you have your error count by set? maybe I'll re-watch a set, just to see what how a 'forced' & 'unforced' error is counted.

Set 1

Cash - 13 forced, 1 unforced (on a BHV)
Lendl - 7 forced, 6 unforced

Set 2

Cash - 3 forced, 0 unforced
Lendl - 3 forced, 3 unforced

Set 3

Cash - 7 forced, 4 unforced (two FH's and two FHV's)
Lendl - 5 forced, 6 unforced

 

[Tennis old man]

Thank for the post krosero!

 

[krosero]

That confirms that my count of the winners was correct. But it’s a mystery why the placement winners exceeded the number of winners by 22. Perhaps some of the forced errors (22 of 29, to be exact) were also counted as placement winners. I simply don’t know.

But I do think that the term “placement winner” refers to clean winners. There are three other boxscores available online in the Times archives for this period: the 1988 U.S. Open final between Graf and Sabatini and the two men's semifinals, Wilander-Cahill and Lendl-Agassi. I can't find Wilander-Cahill, but I got an exact match between my count of winners and the placement winners reported for Sabatini, Lendl and Agassi. Graf was reported at 26 placement winners while I counted only 21. Relatively speaking, I don’t consider that too big a problem – partly because two of the points she won are missing from my DVD.

Overall, it looks like “placement winner” means what we would call a clean winner and that the Cash-Lendl boxscore is either working with some definition that I can’t account for when I look at the numbers, or is mistaken – by a lot.

I think now I would not say that "placement winner" means a clean winner, because I've since confirmed that judgment calls are made on shots other than aces, and I've found other matches from this time period with winner counts inflated well beyond the number of clean winners.

The Cash-Lendl boxscore has 22 such "extra" winners, and it's hard to say for sure what they are; maybe they're a mistake, or something else; but let's just say they don't look strange to me anymore. I've seen other matches with even higher number of winners from judgment calls (ie, not clean winners).

 

[krosero]

Published stats for Doohan-Becker

Doohan d. Becker 7-6 (4), 4-6, 6-2, 6-4

New York Times:

Doohan won primarily because he returned serve so well. Becker had 14 aces and 21 service winners, but those figures would easily have been doubled against most other players. He double-faulted seven times, an indication that he was trying to add even more to his serve, a typical response, and perhaps the wrong one.

Chicago Tribune:

He produced 21 service winners compared with Doohan's 37 and double-faulted seven times.

Atlanta Journal-Constitution:

An accomplished grass-court player, as all Aussies are, Doohan beat Becker at first-serve percentage (71-58 ), at service winners (37-21) and at points on serve (95-80).

Philadelphia Daily News:

Becker served the ball on 118 points; Doohan returned 83 of them, or 70 percent. That's pretty high against Becker.

 

[krosero]

Semifinal stats

Lendl d. Edberg, 3-6, 6-4, 7-6 (10-8 ), 6-4

Sun Herald (Sydney):

SEMI-FINAL DETAILS
LENDL v EDBERG
Lendl Edberg
Aces 7 6
First serve percentage 67 77
Double faults 2 2
Service games broken 2 3
Total points won 137 132

Cash d. Connors, 6-4, 6-4, 6-1

Atlanta Journal-Constitution:

Cash had 10 aces and 14 other service winners; Connors' total was barely half that on a single ace and 12 winners.

New York Times:

Cash made 53 percent of his first serves and had 10 aces and 14 service winners.

Sun Herald (Sydney) has Cash at a different service percentage:

Cash volleyed 23 winners to the American's six, and buried 10 smashes, whereas his opponent didn't make any.

SEMI-FINAL DETAILS
CASH v CONNORS
Cash Connors
Aces 10 1
First serve percentage 57 73
Double faults 2 4
Service games broken 1 5
Total points won 92 58

Their figures for volley/smash winners are higher than mine, perhaps due to judgment calls (I counted only clean winners).

 

[Moose Malloy]

I can't find a copy of his quarterfinal over Wilander, but I recall reading at the time that he had a very high number of winners

here are my stats on that match

Cash d Wilander 6-3, 7-5, 6-4

Cash served at 78%(62-79)
he won 85% of 1st serve pts(53-62)
and won 53% of 2nd serve pts(9-17)

Wilander served at 73%(69-95)
he won 65% of 1st serve pts(45-69)
and won 38% of 2nd serve pts(10-26)

Cash 47 winners: 6 fh, 11 bh, 14 fhv, 8 bhv, 8 ov
Wilander had 24: 7 fh, 4 bh, 6 fhv, 6 bhv, 1 ov

Cash had 4 aces, 1 df
Wilander had 2 aces, 1 df

I credited Cash with 6 service winners and 9 other unreturned serves.
Gave Wilander 4 service winners and 7 unreturned serves.

I gave both players 6 unforced errors(not counting df's)

Cash was 59-77 at net(77%), S&Ving on all serves, 1st & 2nd.
Wilander was 44-80(55%), pretty much S&V on all serves(the few times he stayed back - all were 2nd serves - Cash came in very quickly during the point)

Cash was 4-9 on break points, Wilander 1-2(his only break came in the 2nd game of the 3rd set, Cash broke back in the next game)

In their 4 set match at '86 Wimbledon, Cash was broken 5 times, Wilander 7

 

[krosero]

So then Cash was broken once by each of his last three opponents: Wilander, Connors, Lendl.

I recall reading that Cash's best match at this Wimbledon was actually against Forget in the R16, according to Fraser, I think. I'm wondering if Cash was unbroken in that match, and how many times he was broken overall for the tournament.

He lost just one set in the event, 2-6 to Schapers in the R32.