The Revolution In Tennis Stats That Didn’t Stick

Metals scientist Bruce S. Old contributed to the Manhattan Project, helped found the Office of Naval Research and advised several U.S. presidents.

In his spare time, he brought a notebook to tennis matches and collected statistics for further analysis — “perhaps the most difficult task I ever undertook,” Old later wrote.¹³

Old, an avid player himself, meticulously charted every shot of every rally to figure out how to play the game optimally. A scientist, he broke down the sport scientifically — then wrote about it in a series of books endorsed by some of the best American tennis players of their day.

Sometimes his son, Randy, watched matches with him. “It was boring,” Randy said, laughing. “He didn’t talk. He had this huge notebook, and he just was concentrating.”

In a typical story of a sports stats pioneer, I’d show how this early work inspired others and led to a revolution in analysis of the game.¹⁴ Not this time. Nearly six decades after Old’s first book hit shelves, no one is producing stats like his. Old’s stats on how often pros hit overheads for winners, or hit their returns down the line, aren’t available in tennis today.

In Old’s time, tennis greats Pancho Gonzales, Arthur Ashe, Chris Evert, Don Budge and Jack Kramer all wrote introductions to the books he wrote with nine-time Grand Slam doubles champion William F. Talbert. Budge and Kramer called Old’s book on doubles the “bible” of the sport.

But these days, Old’s work is mostly forgotten. His books are out of print. Contemporary doubles stars say they haven’t heard of him. The stats they use are much more rudimentary.

Bob and Mike Bryan, the American twins who are the most successful doubles team of all time, both called the book “cool” when I showed it to them at a tournament in London last month. Bob said the stats wouldn’t apply to today’s doubles game. But as he spoke, his brother kept leafing through the pages to get a look at stats he’d never seen, such as which kinds of volleys are most effective.

The making of a stats maven

Old got started in tennis stats thanks to a fortuitous college dorm-room assignment. His freshman-year roommate was Bryan M. (Bitsy) Grant, Jr., “one of the most marvelous athletes by the pound the world has known,” Old wrote in a diary.¹⁵

The 5’5”, 130-pound Grant could outrun and outthink most opponents, but sometimes he needed his roommate’s help with the thinking. Grant wanted details on how opponents tried to pass him when he was at the net. Old’s information on how other players responded to Grant’s approach shots helped Grant beat some of the best players in the world, by Old’s telling.

That memory was in the back of Old’s mind in 1947 when he went looking for a book he could read to improve his own club-level doubles game. He checked the library stacks in his hometown of Concord, Massachusetts, and in Boston, Cleveland, New York and Washington, D.C., but “the story was the same everywhere — there had never been a book written on doubles,” Old recalled.

So he set about writing one. He drew up stats sheets to diagram where each ball was hit and how players moved around the court. Starting in about 1950, he analyzed the U.S. National Doubles Championships in nearby Chestnut Hill, Massachusetts. He and one of his tennis pals charted points. They also watched points at a local club, to add to their sample size and to contrast top-level doubles with the amateur game. It was, Old writes in his diary, “much like an operations research study to gain statistical significance to back up conclusions being formulated.”

Old spent three years charting matches, analyzing his data and writing up the results. Then he stalked the game’s greats at the Longwood Cricket Club in Chestnut Hill, seeking a co-author to lend his findings credibility. Australian Harry Hopman shot him down. Old’s approach to Talbert was more “diplomatic,” as Old tells it in his diary: He told Talbert he was asking him because he was the best current American doubles player. Talbert soon had plenty of time to read, sick in a hospital bed for a week with infectious hepatitis. He was so impressed he started sleeping with Old’s notebook under his pillow lest he lose it.

The publishing firm Henry Holt bought the book and Gerald M. Simons edited it, somewhat harshly: He called some of the prose “barely adequate.” “While that hurt, I had to admit that he improved things, as do most editors,” Old recalled.¹⁶ “The Game of Doubles in Tennis” was published in five countries, sold about 63,000 copies by 1980, and earned Old about $18,000 after expenses.

It also earned Talbert and Old contracts to write three more tennis books, with illustrations by Old’s wife, Katharine “Bunny” Day Old. First, after five years’ work, came a book on singles, ¹⁷ then one on how to hit the ball. The fourth was “Tennis Tactics,” published in 1983, combining and updating their singles and doubles books.¹⁸

Talbert was a good choice for the books’ main focus, tactics. His diabetes limited his endurance and strength, so he put his energy into teamwork and playing the percentages. “There wasn’t much” in the books “that he hadn’t [already] taught me,” Tony Trabert, Talbert’s occasional doubles partner, said in a telephone interview.

Old always pooh-poohed his own tennis game — in his diary he called himself “the worst tennis player ever to write four books on the subject.” But he was a formidable force at his country club. He won 15 titles in singles, 11 in men’s doubles and 12 in mixed doubles¹⁹ over a 20-year period.

“He was incredibly modest — about everything,” his grandson and namesake Bruce said.

His descendants say playing against Old was maddening.

“He never made a mistake,” Randy said. “He always knew where to put it and did it. It was painful.”²⁰

But doubles was Old’s true love, as it was for Talbert. “It is just about the best of all sports,” they wrote.

One factor in its favor, they wrote in the book, was that doubles can be played “from six to sixty, and well beyond.” Old made it nearly seven years beyond before his first of two hip-replacement surgeries, in 1980, ended his competitive career.²¹ Soon after, “Tennis Tactics” was published, and that marked the end of his tennis writing career, “a great hobby that brought much pleasure and many fond memories over about 35 years,” Old wrote in his diary.

The revolution falters

In other sports, statistical pioneers pass the mantle to others who build on their work. In tennis, as far as I know, no one has charted the game like Old did.

At today’s big tournaments, Hawk-Eye Innovations rings courts with video cameras to estimate the trajectory of the ball so players can contest calls. Singles players can also glean such stats as the average speed of groundstrokes and the average distance by which each player’s shots clear the net.²² The cameras could be used to show the same stats for doubles, but Hawk-Eye isn’t producing the data.²³ Even if it were, the Hawk-Eye stats made available to the media for singles don’t match Old’s level of detail on spin and depth of shots.

That leaves doubles players with only the basic stats that date back decades, such as percentage of first serves in. Some players also employ coaches to chart video — when video exists. Colin Fleming, who was ranked 17th in the world in doubles last year, said he sometimes has friends take video of his matches, then sends the video via Dropbox to his coach, Louis Cayer. “It’s frustrating. It’d be good to have the mean statistics for each area of your game,” Fleming said.

Last month at the World Tour Finals in London, I showed the doubles book to the Bryans, to Fleming, and to Daniel Nestor, a 12-time Grand Slam champion. None had seen stats like Old kept — on winning percentage on volleys depending on where they were struck, or on which shots were most likely to yield winners.

They said some data-driven tenets from the book still held: Come to net.²⁴ Crowd the middle of the court. Keep returns low. Others, like taking speed off the first serve to allow time to come to net, don’t apply in today’s power-dominated game, the current stars agreed.²⁵

Some basic stats hadn’t changed much, either. Talbert and Old wrote that in 1950s doubles, serving teams won 70 percent of points and seven of eight games. At the 2014 tour finals, played on a slow court that reduces servers’ advantage, serving teams won 66 percent of points and six of seven games.

At the tour finals, I tried charting the same stats Old recorded more than 60 years ago, to learn more about how the game has changed.²⁶ What I found corroborated what today’s doubles stars told me: The increased power of the serve and return have made those shots more important. Old found that 19 percent of points were decided by serves, either aces or unreturned serves. At the tour finals, twice as many points were. And 11 percent of points in Old’s time were settled by a return that was either a winner or elicited an error. That figure was 15 percent in the matches I charted. Serves and returns gained at the expense of volleys, which decided 22 percent of points, down from 40 percent in Old’s charting.

But I can’t make any great claims to accuracy. I tried to simplify by skipping depth and spin of serves and depth of volleys, but when rallies went past four shots, I was lost. I missed at least one detail on one in five points. The points went too fast, and with four players on the court, there was too much to record.

It’s hard to know whether Old struggled with the same problem. His descendants haven’t been able to dig up for me the books Old used to chart matches. Based on his diary, it sounds like he drew points out on court-shaped diagrams, then extracted the stats later. His son Randy is pretty sure he wasn’t using video, at least not for the first book. He was just taking notes like crazy.

“He’d go whipping through [his book] as the point went on,” Randy said.

Randy spoke by phone about his father for 35 minutes the morning he was to be sworn in as a newly elected member of the Vero Beach, Florida, city council. He didn’t mind the imposition on the busy day. “It’s a kick talking about him,” Randy said. “It’s one of those things that was important for a while, and then disappeared.”

A Federer fan

Years after publishing his last book with Talbert, Old kept watching tennis keenly. Randy said his father’s favorite modern player was Roger Federer, who won his first Grand Slam title at Wimbledon a couple of months before Old died, just shy of 90, in 2003.

Old’s grandson Bruce recalls watching Wimbledon at his grandparents’ Cape Cod home. The rule was, the old television there was only ever switched on during the annual tournament. “People would be kind of rotating in and out of the living room with scores updates,” he said in a phone interview. “There were three generations sharing couch and floor.”

Some of Old’s writing comes off as, well, old-fashioned. He calls winners “placements,” and topspin groundstrokes “topped drives.” More notably, he writes critically of women’s tennis. “The men rightly tend to look down upon the game of the weaker sex,” Old and Talbert wrote in the book on singles. They added, “In general, the ladies play a type of singles which is a notch well below that of the men, as they cannot hit as severely or cover the court as well. Thus, they have not contributed greatly to the development of the game from a technical standpoint. Instead, they tend to follow the lead of the males.”

Randy said those sentiments weren’t sexist, but were accurate reflections of the relative states of men’s and women’s tennis at the time. “He was not biased, by any means,” Randy said, “but the women’s game was not strong in those days.”²⁷

Bruce said his grandfather came to appreciate the modern women’s game more than men’s tennis. “It spoke more to him as a tactician because it wasn’t so centered on the serve,” he said.

We’ll never know how Old would have broken down the modern game: He didn’t chart matches or analyze numbers regularly after “Tennis Tactics” appeared. But there’s a fascinating passage in the Talbert and Old singles book in which the authors consider how former greats would compare to contemporary stars. Many analysts today use the same line of reasoning.

Beyond any question, the big game as played today far outstrips any all-court game of the past. Does this mean that the more modern players, Kramer and Gonzales, are head-and-shoulders above [Bill] Tilden as tennis players? We would like to answer this question in two ways. First, we believe that a Kramer or Gonzales employing the big game of today would consistently beat a Tilden on grass playing his all-court game of 1920-25. Second, we are convinced that had Tilden played in the 1960 era, he likewise would have adopted and mastered the big game style. So equipped, the Tilden of 1960 could have murdered the Tilden of 1920-1925. There is no denying the remarkable skills of the greats of the past, but a Doherty of 1900 could provide little competition for a Doherty of 1960.²⁸ We agree, however, that it is impossible to compare accurately a Tilden, Budge or Kramer employing the 1960 style of tennis with a Gonzales of 1960. Yes, tennis has made great forward strides since 1874. The development of the game has been entirely logical and consistent. And the best is yet to come.

The Old of 1956, I’d bet, would remain at the forefront of tennis stats in 2014.