When Grand Hombre made his debut at Gulfstream Park, readers of the Daily Racing Form might have concluded that the colt couldn't outrun a fat man.

The unraced 3-year-old showed three workouts that had been recorded at the Classic Mile Training Center in Ocala, Fla. He had worked a half mile in 53 seconds, another half mile in the identical time, and a half mile in 543/5 seconds. At that speed he would barely get work pulling a Hansom cab in Central Park.

Yet Grand Hombre was getting some support at the betting windows, going to the post at 12 to 1 against an exceptionally strong field of maidens Saturday. One of his rivals was a highly touted first-time starter who had cost $1.7 million as a yearling. Another had finished second in a fast race at Gulfstream last month, earning a speed figure good enough to win the vast majority of maiden races.

But none of them was a match for Grand Hombre. When the gate opened, he shot to the lead and dueled with the favorite, speeding the first half mile in 44.60 seconds. Then he drew away with authority, winning by four lengths while running six furlongs in 1 minute 8.66 seconds -- phenomenal time for any horse, let alone a first-time starter.

Racetrack people wasted no time concluding that trainer Dennis Manning had pulled off the betting coup of a lifetime. It was a perfect scenario: He had the fastest 3-year-old maiden in America. He was stabled at a training center, so few people could have known about the colt's ability -- particularly clockers and their clients. And the system for reporting workouts at these facilities is made to order for a trainer who wants to deceive the public. Much of the time there are no clockers present, and the trainer is responsible for reporting his horse's time to the racetrack.

The people assuming this was a coup included plenty of Manning's acquaintances. "Everybody and his brother has been calling to ask, 'How come you didn't tell me?' " the trainer said. Manning tried to persuade them he had no idea Grand Hombre was capable of running in 1:08.66, and that he only knew this was a nice colt with a chance to win.

Grand Hombre's owner, Earle Mack, bred the son of Grand Slam and intended to sell him at auction. But when the colt didn't bring a sufficient bid, Mack bought him back for $35,000 and turned him over to Manning. At first the trainer wasn't impressed. Grand Hombre had shin problems and he was physically immature -- "a big Baby Huey." So Manning didn't start to train him in earnest until last fall. Then he liked what he saw. "You can have 10 horses training together," he said, "and one will just jump off the page." Grand Hombre jumped off the page.

But Manning insisted that the colt wasn't burning up the track with fast workouts. When he works a youngster, he said, he likes the horse to go slow early and speed the last furlong. Grand Hombre's best work, he said, was a half mile in "49 and change" -- with the last furlong in a blazing 11 seconds. Manning said he had duly reported that workout, but it never showed up in print.

The trainer believes Grand Hombre is cut out to be a distance runner, and he maintained that he was not confident he would win his debut in a sprint. "I thought he had enough quality that he might get up [i.e., come from behind] to win at three-quarters of a mile," Manning said, "but I'd have been very happy if he finished second or third."

I asked him: Did you bet?

Manning replied: "I'm not a gambler. I don't bet $2 on anything. I just try to train my horses properly." As for the Mack, Manning said, "He bet to place" -- on the theory that if the race were too short for Grand Hombre he still might rally and get up for second place.

Manning sounds credible when he talks about his horse's preparation, but most horseplayers are surely convinced those published 53-second workouts were as phony as a $53 bill. This is an integrity issue for which the sport has no good solution. As private training centers have proliferated, the racing industry can't afford to hire clockers to report on every workout, but relying on reports from trainers is a system that invites abuse.

Dave Bailey, Gulfstream's racing secretary, said he is convinced the system works and that published information is largely accurate. "Most of the time it's on the up and up," he maintained. This contention flies in the face of experience and human nature. Ever since Eclipse was preparing for his career debut in the 1600s, thoroughbred trainers have been trying to conceal the ability of first-time starters and hide their identity from clockers.

Bettors understand that this is the way the game is played, and they will always assume they have been deceived when they see a horse win after a dismal string of workouts. And they will feel a grudging admiration for a trainer who can hide a horse capable of running in 1:08.66.