Photo finish examiner details how trials race was deemed a dead heat
EUGENE, Ore. -- At roughly 20 minutes past six last Saturday evening at Hayward Field, eight women ran the 100 meters at the U.S. Olympic track and field trials. The first three finishers would earn individual places on the U.S. team for the London Olympic Games. It's an overwritten truth that such trials as these are among the most emotionally punishing contests across the landscape of all sports: Four years of training, one race to succeed or fail. Four years -- or forever -- to live with it. But make no mistake, it's a truth just the same. There's desperation in the air in those last few minutes before the gun.
In the 11 seconds it took to run that 100-meter race, and in the approximately 45 minutes that followed, track and field detoured into a very odd place and remains there to this day, a resolution no closer than it was in the slanting shadows of an early summer night. Carmelita Jeter, who has been the fastest woman in the world for the last three years, won the race. Tianna Madison, a one-time world champion in the long jump suddenly turned sprinter, was second.
Behind them came athletes at opposite ends of the professional spectrum (yet who train together on the same track at UCLA, with the same coach, Bobby Kersee). In Lane Two was Allyson Felix, a two-time Olympic silver medalist in the 200 meters, three-time world champion and one of the most prominent track Olympians of the last decade. In Lane One was Jeneba Tarmoh, 22, who made her first worlds team last summer. Their compensation and endorsement packages are hundreds of thousands of dollars apart.
Their finish was declared a dead heat that will be broken in the near future (or some time before the 100 meters begins Aug. 3 in London). This story is not about that resolution, or about the fact that USA Track and Field did not have in place a procedure for breaking such a tie, which is unconscionable. This story, in these words and video above, is about the minutes that followed the finish of the race, a story that's told in hundredths of a second, but stretches across two decades of American track and field history, from a Manhattan hotel room to a crow's nest at the top of Hayward Field. It's about how the most sensitive technology on earth still requires human assistance in track and field. It's about the difference between a torso, the wheel of a bicycle or the nose of a horse.
Every day of the Olympic trials, Roger Jennings, 45, sits at the far right end of the officials' booth at the top of the Hayward Field grandstand, a majestic, green structure with a cantilevered roof that keeps everyone under it dry. Jennings is the chief photo finish examiner for Company X, which handles results and timing for not just the trials, but dozens of other major competitions. This is the subcontractor for Company Z, which created and manufactures the photo finish technology. Because of highly restrictive sponsor arrangements that could damage their business, I have been asked not to name their companies. But if you Google Roger Jennings of Company X or marketing director Giles Norton of Company Z in conjunction with photo finish technology, you'll find their company names.
Jennings' father, Tom Jennings, founded Company X in 1992. A native of Long Beach, Calif., Tom Jennings, 70, was in the 1970s the manager of the influential Pacific Coast Track Club, which included the likes of high-jumper Dwight Stones and shot-putter George Woods. Many in the track world consider Jennings the first "agent" in the sport, in the days when athletes were paid under the table by meet promoters.
In 1992, Tom Jennings was in New York City for the national indoor championships on the boards at Madison Square Garden. As Jennings recalls, he ran into Howard Schmertz, the legendary director of the Millrose Games at the meet headquarters hotel, the Penta, on Seventh Avenue across from the Garden. Jennings said, "Howard told me, 'You gotta go see what they've got in room whatever, 237 or something."'
That was veteran track official and computer technology specialist Bob Podkaminer's room. Podkaminer was letting University of Maine graduate Doug DeAngelis sleep on his floor and use the room to show off a new photo finish system that he had developed, and which would become the foundation of Company Z. (Feel free to Google DeAngelis, too). "At that time, we were driving around to track meets in Doug's Volkswagen," Norton said. "And sleeping on a lot of hotel room floors."
Remember the name of DeAngelis's host: Bob Podkaminer. He will be back soon. Track and field is a small town.
Jennings, then 51, threw his professional energies into developing a company that could use DeAngelis' system. The first time he used it formally was in the spring of 1992 at a meet involving high schools from Hanover, N.H.; Essex Junction, Vt.; and Nashua, N.H. (Jennings was living in New Hampshire at the time). Roger Jennings had grown up a track geek, devouring statistics and reading his father's full pre-Internet collection of
Roger Jennings has since become one of the most prominent photo finish judges in the world, and the only American member of the international track federation's seven-member photo finish judges' panel.
Last Saturday afternoon, Jennings watched the women's 100 meters with his naked eyes. "I try to watch the first-place finisher," he said Tuesday. "I'm still a track fan." As soon as the race finished, Jennings did what he always does, and what he has done, literally, thousands of times. He turned his eyes to the laptop in front of him and looked at the photo finish picture that's frozen on the screen.
The picture, taken with a camera that shoots 3,000 frames per second, captures and freezes each of the runners at the instant they cross the finish line. Yet defining that moment is what becomes complex and distinctly more imprecise than one might expect. Here is why: In track and field, an athlete is defined as having crossed the finish line when his or her torso breaks the plane of the finish. But since humans are built differently from each other, and move differently from each other, even the most sensitive technology cannot determine flawlessly when that moment occurs. Often arms, feet or Gail Devers' hair crosses the line first, but none of those things count. Which is where Jennings' eyes come into the picture.
"In horse racing, which I've done, it's easy: You go by the horse's nose," Jennings said. "In auto racing, the front of the car. In speedskating, by the skate. In cycling, by the wheel. It's the first thing that crosses the line. In track and field, it's the torso. And there is subjectivity in determining where the torso is. That's what we got into in the women's 100 meters."
(An aside here: Sports fans are increasingly strident in their demand for technological solutions to competitive uncertainties. Track races would seem to be ripe for this type of intervention but, in fact, the torso dilemma makes it impossible. There is a distinct human element).
Tuesday morning in the conference room at his Eugene hotel, Jennings walked SI.com through the process that he undertook following the controversial race.
(And right here, props to the running website LetsRun.com for first posting a cursory interview with Jennings after the race last Saturday night; the attempt here is to tell a fuller story. But they got him first.).
Jennings pulled up the finish photo on his 17-inch screen, just as it popped up automatically Saturday as the race finished. "I've seen this a million times," he said. To determine times and places, Jennings uses a wireless mouse to place a vertical line on athletes' torsos, from first to last. Once the line is in place, he inputs the lane number of the athlete, hits "enter," and the software spits out a time, and that time is immediately posted to the stadium scoreboard. He re-creates the process here: "[Lane] Six (Jeter) was easy (for first), [Lane] Four (Madison) was easy (for second)," Jennings said. "Then I saw this ... " He scrolls over to the image of Tarmoh and Jeter, clearly very close. "I went, 'Oooooffff.'"
Jennings also said, into his headset, which is connected to the infield officials (another headset is wired to the NBC broadcast support team, for whom Jennings provides immediate finish information), "Oh, sh--." He recalled Tuesday: "This is exactly what you don't want at the Olympic trials."
But Jennings went immediately to work. Times for Jeter (10.92 seconds) and Madison (10.96) were posted as soon as he had clicked enter. The camera that captures his primary image is on the infield, above the finish line. "Allyson Felix's shoulders were pretty square, so it was easy to find the torso," Jennings said. "I could see right away that Tarmoh was twisted, throwing one of her shoulders forward." This makes it more difficult to find the torso. Jennings enlarged a secondary image from the lower right corner of his laptop screen that was shot from the outside of the track. "Look at this," he said Tuesday.
Tarmoh's torso is totally obscured by her arm in the photo taken from the outside angle. "It shows nothing," Jennings said. "I'm hoping it will show me a torso position, but her torso is totally blocked. I can't see a thing."
At this point, 15 to 20 seconds have passed since the runners went through the finish line. Felix and Tarmoh are staring up at the scoreboard, awaiting their fate. Hayward is buzzing and, pointedly, so is Jennings' headset. NBC's telecast has already run over, and while no one has suggested that the network was pushing Jennings for an unduly swift resolution, referee -- wait for it -- Bob Podkaminer said, "As soon as there was a delay, there were voices in that booth who wanted something. The meet announcer wanted something. NBC wanted something. Those voices were loud."
And Jennings wanted to give them something. The right something, based on his experience. "At this point, it's been like 25 seconds," Jennings said. "That's about as long as I can take. If you're an umpire, you can't take 20 minutes to call a pitch a ball or a strike. And the torso is very much like the strike zone. I had to make a call." I asked him if he felt pressure. "There is some pressure," he said. "But I've been doing this so many years, I don't even think about it."
In any race, when Jennings can't, literally, see the athlete's torso, he uses two known data points to interpolate where the most forward point in the torso probably is. (Given Jennings' level of expertise, and after watching him work, this is a very strong use of the word "probably."). On his laptop, he adroitly located a point on Tarmoh's paper ID bib on the right side of her chest and another spot on her thrown-forward right bicep. Those were data points A and B. He determined that data point C would represent the forward-most spot on Tarmoh's torso (in the photo, it's obscured by Tarmoh's head), lined up his cursor and clicked. The scoreboard spit out 11.067 seconds. Then he clicked on the readily visible spot on Felix's torso, and the scoreboard illuminated 11.068 seconds.
Tarmoh celebrated and began a victory lap. Felix wept. Jennings immediately turned to his right to USATF official Duffy Mahoney and said, "We need to get referees in here."
(It's very important to understand here that the numbers put up on the scoreboard were unofficial, although nothing on the board conveyed this fact to the crowd, the celebrating Tarmoh or the crying Felix. Those instantaneous numbers are always unofficial. Three days after the fact, Jennings was still wondering if there was something he could have done differently. "I have control of the scoreboard," Jennings said. "I could have typed UNOFFICIAL over the whole finish order. Maybe I should have done that.")
In the ensuing minutes, several meet referees entered the booth, and it was decided by acclimation that one referee would consult with Jennings and make the determination: Bob Podkaminer, 70, who first officiated a track meet in Northern California in 1965. Podkaminer's official title on the day of the meet was men's chief running referee (he is also secretary of the USATF Rules Committee), but he has extensive experience in reviewing photo finish evidence. "Roger asked for me specifically because of the experience I have in reviewing photo finishes," Podkaminer said Tuesday in a phone interview. "If Tom Jennings bought the first (Company Z) system in 1992, I probably bought the second."
The consultation lasted more than 20 minutes. Podkaminer did not dispute Jennings' interpolation of the photo. In nearly any other instance, that result would have stood. But Podkaminer, as the representative of USA Track and Field, did not want to put himself, Jennings or the USATF in the position of defending an interpolation against an appeal. "In many cases, you can interpolate," Podkaminer said. "But these are the Olympic trials, where there has to be a more exacting standard. An interpolation is not fact. At some point, I might be asked to stand up and justify what I decided."
Jennings said, "In the end, my read was subjective. The involvement of the torso is always subjective to some degree. They (USATF) went with what they could actually see. I was overruled, and I certainly signed off on their decision. But I did my job. I called what I saw. I try to stay consistent. If I went back and read that photo 100 times, I would call it the same way every time." (That is, he would call Tarmoh the winner, based on an interpolation of where her torso was at the finish).
What took place after Podkaminer's ruling is now an unseemly chapter in track history. USATF spokeswoman Jill Geer informed media that the race was a dead heat, and that there were no procedures in place to settle it. More than 24 hours later, Geer again stood up in front of the same media and announced
Yet the frenzy that led to the tie is clearer. In the retelling, Roger Jennings comes across as unfailingly professional. I watched him work Monday night in his crow's nest perch. He is responsible for not just finish photos, but clicking on athletes to provide splits as they run around the track and spotting for television. It is an intensely pressurized environment, intimately linked to the spectators' understanding of the meet. In the heat of battle on Saturday, he flagged himself to protect the integrity of the trials. (Not, it should be said, to protect Allyson Felix, the more famous of the sprinters in limbo. "They're pixels to us," Giles Norton said. "Not people.").
It is bittersweet, as well. Roger Jennings stood in that hotel room Tuesday, with his laptop under his arm and said, "If I hadn't protested myself, I'm not sure anyone else would have."