Jacques Lavigne did not think of it as "criminal activity," not in the usual sense. He did not think of himself as a criminal. Criminals put guns in people's ribs and broke heads. Jacques considered himself "a nonviolent person." He did, however, think of himself as a man who liked a "challenge" when money was to be made. Around the tracks he had that reputation, and he did not play it down. His formal education was sketchy, but he said he had "mingled with intelligent people" all his life, and many of them knew the dark attraction of fast money. He said, "You can't beat experience."
Jacques, whose thick, horn-rimmed glasses gave him a steady, opaque look, was the son of a Maine lumberman of French-Canadian descent and had grown up around racetracks. He had been a groom and an assistant trainer, and he had made an inauspicious stab at jockeying, until he grew too big. Then he moved to Florida and into the racing plants themselves as a ticket seller and, finally, a calculator in the mutuels rooms, the fiscal centers of pari-mutuel betting.
The line between right and wrong tends to blur around betting establishments. Money floats. Anglers flower. In the landscape of potential larceny, Jacques found a natural habitat. He was thrilled by stories of older men who had found ways to "manage" the system and bring down some of that floating money. The artful dodgers and cozeners who claimed to have found an edge told how they "handled" the old calculators, penciling in a winning ticket here, a place ticket there, and cashing in. Jacques was fascinated, but he judged it "nickel-and-dime stuff." Jacques liked to measure a score in terms of "six goose eggs." If he ever got the chance, he said, he would "go for the maximum."
It was when the computers came in that Jacques' interest peaked. The computers streamlined the betting process; like a mind with a thousand hands, they translated the indiscriminate bets and channeled the money. From his vantage point in the mutuels room at Flagler Dog Track in Miami, Jacques could see the computer operator's skilled fingers activate the electronic magic of the machines in the room next door, and the magic transferred into the tote boards in the infield, the lights flashing. Like the lights, Jacques' nerve endings jangled. He said he "caught on to it like really quick."
He began feeling around, the line of his questions arousing interest among possible allies. He became friendly with a computer operator named William Deal, who seemed to see life as he did. One night while working in the mutuels room he asked Deal if it were possible to "pull tickets" out of the computers. Deal said he'd look into it. A few days later, between performances at the dog track, Deal joined Jacques for a walk in Coral Gables and told him he "had it." Deal said the computers could indeed be fixed. He explained to Jacques how he could do it.
Jacques was not convinced. The perfect scam at a racetrack, he said, must have a "safety factor," must "arouse no suspicions" when you make a big score, not from the track officials who balance the books, not from the fans who can see their bets working on the tote board. Knowledgeable fans are wary of wildly fluctuating odds.
And then one day in the early 1970s, as Jacques peered out onto the infield, watching the lights through his jelly-jar glasses, he saw the flaw. It was something, actually, that he did not see, but it was clear to him nonetheless, like an opening in a scrimmage line. The safety factor he needed. He saw it and he knew he was going to make a potful of money.
He knew he was going to make it because he could take it and the track would not know or, for that matter, care. The track would not lose a dime. He was going to steal it from the betting public, take money right out of the bettors' hands without their knowing it. Without their ever knowing. He would steal their money even as they happily counted their winnings, because it was from their winnings that he would steal.
And he would not get caught because there would be no alarming shifts on the tote board, no somersaulting odds to catch the eye. And at the end of the day there would be no shortages, no counterfeit tickets lying around and no evidence that could not be destroyed.
What Jacques saw as the final lubricant for his plan was a flaw in the system the tracks themselves had provided. The computers had opened up gimmicky wagers to lure the players, among them the trifecta—a bet requiring the selection of the exact 1-2-3 order of finish in a race. The odds against winning such a bet are immense, but the bait is seductive, the biggest payoffs sometimes running to five figures. However, in an eight-dog field the trifecta creates 336 possible betting combinations. The odds can be handled easily enough in the computer, but it is impossible to post them all on the infield tote board; there is not enough room. The tote board is the player's barometer. On it he can see the odds dance, and he can chart his win, place and show wagers against the betting pools, determining almost to the nickel what they will pay. When he bets the trifecta, however, the odds and the pools are not there; he is blacked out until the payoffs are flashed.
As a result—and this is what Jacques saw—the trifecta odds could change dramatically, even after a race, and no one would be the wiser. They could be changed in and by the computer itself as they drifted in the electronic limbo of the computer memory, awaiting the operator's release of the final printout and the posting. Deal assured him that between the finish and the final posting he could "move" tickets from one "address" in the computer to another, manipulate the switches in such a way that tickets purchased on losing dogs could be readdressed to the number of the winner. The final totals would match. He could do it quickly enough to avoid suspicion. And he could do it right under the noses of the men the state paid to stand watch—because they knew zilch about computers.
It fascinated Jacques, the idea of beating those machines. "If man builds 'em, there's a way," he said. "It's like breaking the bank at Monte Carlo." He thought it ironic that the track would, in effect, be party to his larceny. The track's insatiable appetite for ways of getting into the public pocket had made it easy—easier still with the introduction of trifecta "wheel" tickets, tickets that allowed a bettor, at $84 a play, to pick one number and wheel the others in the field, covering all trifecta combinations with that number. To move trifecta wheel tickets, Deal had only to readdress one number instead of three.
A problem for Jacques was keeping the conspiracy safe but wieldy. He and Deal could not work it alone. The simple logistics of moving, printing and then cashing the tickets without raising questions demanded assistance, and there were at least two areas to insulate: the mutuels room, whose personnel was employed by the track, and the computer-totalisator operation, run by Automatic Totalisators. A company incorporated in Delaware and serving 60 wagering plants nationwide, Autotote, serviced all the Flagler betting machinery and ran the computers. Deal was Autotote's computer manager at Flagler.
In the mutuels room, Jacques needed a soothing voice to calm track officers in case of a foul-up, someone in authority to explain any delays to the stewards and the two state agents overseeing the operation. (Jacques had no fear the latter would discover anything on their own. They were "coffee drinkers," he said.)
Jacques had gotten friendly with a fat, wise-talking mutuels calculator named Richard Korn. He considered Korn "a dealer—he likes money." Jacques had no trouble persuading Korn. In turn, Korn enlisted the mutuels manager, a man named Al Johnston. Jacques had "heard stories" about Johnston and had himself heard Johnston talk about the chicanery of "the old days." Jacques sent Korn into Johnston's office to talk. Five minutes later Korn emerged with his thumbs up.
To protect Deal in the computer room, they rang in the assistant computer operator, a man named Robert Watters, and Watters himself was taught the procedure so that he could serve as Deal's backup. A third Autotote employee named Gilles Caisse was enlisted primarily to keep a record of the bets moved by the operator and subsequently to print the corresponding tickets.
None of the conspirators had a criminal record. They were rookie crooks, coming to bat for the first time, with all the attendant qualms. But once under way the scheme functioned perfectly. (When it proved out, Jacques said he experienced a kind of inventor's euphoria, a sensation not unlike levitation.) Deal had a cautious but steady hand and a good memory. The latter was important because he had to choose his address changes quickly and carefully. The moment the dogs bolted from the starting box, when he ordinarily pushed the "race increment" button which would set up for the next race and print out the betting results, Deal switched off Computer "A," one of the two computers used to orchestrate the betting process at the track. In Autotote's dual system, "A" was official; it fed directly into the tote board. Computer "B" ran simultaneously, offering the same information—the amount bet, which windows it was bet through, the odds, the pools, the breakage, etc.—but was used only as a backup.
No one noticed when Deal turned off the computer. There were no telltale lights, no alarms. He waited the minute or so it took the dogs to run around the track, and on a signal from one of his confederates in the mutuels room—giving the winning number as the dog crossed the finish line—he transferred bets recorded on losing numbers to the winners. The transfer involved as many as 20 toggle-switch maneuvers and required upwards of 30 seconds. He then turned the computer back on and hit the "race increment" button to produce the final printout of the betting. The bogus figures were now implanted with the legitimate bets.
The rest was routine. When the race judges verified the order of finish, Deal punched the "compute prices" key, then hit the "display" button to light up the payoffs on the tote board.
The maneuver had the effect of diluting the trifecta pool with the addition of illegitimate winners. The unnoticed loss to the legitimate ticket holders could sometimes be spectacular. If, for example, two tickets were moved in a trifecta that had a betting pool of $4,000, and there had been only two legitimate tickets purchased on that number, the fraudulent tickets would rob the lawful bettors of half their winnings, the payoff being split four ways instead of two and the price dropping from $2,000 a ticket to $1,000. In the flush of a $1,000 victory, of course, bettors are not inclined to demand a recount.
Computer "B," meanwhile, clacked on without interruption, undoctored and unattended, spewing out its evidence against the conspirators, which they themselves routinely wadded up and tossed into a wastebasket.
All that remained was to print and cash the tickets. Caisse printed them after the evening program, when Autotote personnel customarily serviced the ticket-issuing machines behind each cashier's window. His presence was not unusual. With the ticket machines no longer hooked up to the computer, it was simple enough to insert a "date stick" (a plastic key that imprints the date and proper code letters on each ticket, with the letters changing from race to race to discourage counterfeiters) and manually run off the tickets.
For accuracy and for accounting purposes, Caisse kept a record of the transfers on a small "code sheet," a list of the day's code sequences given to the tellers to code the machines, and totaled the trifecta payoffs on the sheet.
Jacques himself would go to the past-performance ("outs") window the next day to cash the tickets. He distributed the take at the track that night. Caisse then disposed of his "record," tearing the code sheet into pieces and throwing it in the wastebasket.
In the early days, the crew moved with fear and caution. Pickings were slim. Then one day a race went off and tickets were moved, and suddenly the conspirators were the only "winners" in the trifecta pool—a $26,000 payoff. For the first time, Jacques understood what "going for the maximum" could mean. He was 30 years old, making $14,000 a year, and he could see his life swelling with promise and his family living "comfortably forever."
He told his wife their surprising good fortune was a result of his skill as a handicapper. He had always been an enthusiastic gambler; now, he said, he was a smart one. To cover himself at the track, he made frequent trips to the ticket windows, buying and, occasionally, cashing legitimate tickets. He cautioned his accomplices to avoid "unusual displays of wealth." Nevertheless, Jacques now did things he never did before. He took his family to Disneyworld. He "thought nothing" of paying $30 a pair for his children's tennis shoes. He bought season tickets to Miami Dolphin games, and for the first time in his life went deep-sea fishing. He loved it. He made plans to buy some horses "for somebody else to train."
Like an underground stream, the swindle rolled on unnoticed. For the conspirators it became almost routine, like a second job. They milked the computers like a cow. Despite their fears, there was nary a hitch.
In 1975, however, after a year of manipulation, Deal resigned from Autotote and the scam was temporarily suspended. Jacques did not approach Deal's replacement. Watters worked the scheme in 1976, but in 1977 he was promoted, although he remained in Florida as regional manager, traveling from track to track. The conspirators were not able to get to the computer operator at Flagler that year, but at the Pompano harness track, where he worked that winter, Jacques approached a young computer operator named Richard Hudson. Hudson had assisted at Flagler before and was scheduled to go back there again as the computer manager.
As roguish figures go, Hudson left a lot to be desired. Slightly built, with lank brown hair and nervous eyes, he had a diffident conversational manner and spoke so softly that everything he said sounded confidential. But Jacques was not put off. He sensed something "good" in Hudson. Sitting together under a race monitor in the Pompano clubhouse, Jacques laid it out. He told Hudson there were "ways to move tickets in the computer" and a "lot of money to be made." He showed Hudson the key to a safe-deposit box and hinted at its content.
"You interested?" he asked.
Hudson felt a wetness in his hands. He was 28 years old, working 80 hours a week to make $22,000 a year. Married just a few months, he and his wife were buying a house.
"Yeah," he said, "I'm interested."
"Think about it and I'll talk to you at Flagler. Just don't say anything. You could get a lot of people in trouble."
Hudson thought about it. He thought about the people at the track who bet, people who "really don't care where the money's going. Out there day after day throwing their money away. They're not going to miss a couple bucks."
On the fourth night of the Flagler meeting, Watters came in to show Hudson how to manipulate the computer. Watters made a bobble or two. Hudson took the controls, but also made mistakes. In correcting, however, he was able to move two tickets for a small payoff. When that session was over, he told Watters he thought he was going to throw up.
"That's the way it is," Watters said. "You tighten up. Your heart pounds. You get a headache." Watters said he got that way every time. He said it didn't make it any easier that passersby could watch the computer operation through a picture window on the ground floor of the Flagler grandstand. Their fraud, however specialized, was consistently on display.
But Hudson soon settled down. His fears subsided, to be replaced by a decisiveness under fire that the group had not known before. He got better and better, and was soon completing ticket transfers in less than 30 seconds.
"The kid's all right," Johnston told Jacques.
"Richie's the best," said Jacques.
As the payoffs mushroomed, Jacques became aware of a new risk: his regular appearances at the outs window. He began using "10-percenters," track hangers-on who cash tickets for high rollers and income-tax evaders. Most frequently he used an insurance man named Wayne Murray, who seemed to spend more time at the track than he did selling insurance. Jacques also put an outs-window cashier on the payroll, a man named Bill Bobson. Neither Murray nor Bobson knew where the tickets were coming from. When payoffs went over $600 and were not cashed by Murray, Bobson saw to it that the IRS forms, known by their number as "5754s," were signed, again by 10-percenters willing to risk arrest for a buck or two. Usually they signed the forms in advance and never saw the figures entered on them. The forms were clipped to the tickets after they were presented to the cashier for payment.
Jacques handled this part of the operation. As orchestrator of the scheme, he knew everybody involved, but he did not tell everybody all that was going on. Only Hudson and Caisse knew that Murray, the truant insurance man, was not getting 10% for cashing tickets; he was getting 5%. By the same token, Hudson and Caisse had only Jacques' word for what he paid the others. Watters, traveling a lot, had no way of staying informed. He was actually in on a pass, anyway, a bonus for his previous involvement. On the other hand, Korn and Johnston had no way of knowing how many tickets were printed. They also had to take Jacques' word. Hudson knew because he transferred the bets; Caisse knew because he printed the tickets.
Usually, Caisse printed the tickets at a particular trifecta window, No. 507, that was convenient and provided a degree of privacy. In his confidence (Jacques would later call it laziness), Caisse came habitually to 507. On Aug. 30, 1977, jacked up "to put more in," Hudson moved 35 tickets—rigging every trifecta payoff on the matinee and evening programs. Four of the tickets were for more than $600. Caisse printed them all at Window 507, without bothering to check the computer printout to see if any tickets had actually been sold at that window. The next day, Murray signed the 5754 forms for the IRS, cashed the tickets at Bobson's outs window, and turned the money over to Jacques.
By that time, computer operator Hudson had a problem—not at the track but at home. Originally, he had kept his share of the take in a suit jacket hanging in a closet. One day his wife, Doris, found the money and demanded an explanation. "I won it," Hudson said. "Good," said Doris. "Just don't lose."
But as the money accumulated, Hudson's discomfort grew. Watters, the only member of the group he saw socially, cautioned him regularly against "splurging." In Hudson's case, there was no need. Never having had money, Hudson was in a position to buy for the first time but, unlike Jacques, he was afraid to. The role of thief nagged at him. His excesses were modest, and mostly attributable to his less inhibited wife. Doris began to collect expensive Hümmel figurines.
One night Hudson came out of Flagler with $10,000 in his pocket. "This ain't real," he said to himself. "If I get rolled, the lucky sonsabitches won't know whether to kill me or kiss me." He opened a safe-deposit box at a bank in Hollywood, Fla. and put the bulk of his money there. He stowed additional money in a plastic bag and buried it under the patio where he sometimes fed his dog.
When he could no longer explain away his incredible good fortune, Hudson confided in Doris. Until that moment, no one outside the conspiracy had known of the scheme. Doris was the first. She would not be the last. Unlike her husband, Doris had a wide circle of friends and was not a reluctant conversationalist. Without going into details, Hudson told her he was "doing things" with the computer and "messing with the outs book." Doris said, "You better hope to God you don't get caught."
The call to the Dade County State Attorney's office came in midafternoon of Sept. 2, a Friday, as the staff was breaking up for the Labor Day weekend. The governor had a report that persons unknown might be working some kind of massive fraud at the Flagler Dog Track. Details were sketchy. They were provided mainly by an informer, a somewhat dubious track regular seeking reward money. State Attorney Richard Gerstein immediately called Martin Dardis, his chief investigator. "Don't leave," he told Dardis.
For Dardis, it was a familiar summons. He was Gerstein's prize bloodhound. He had been a cop 30 of his 50 years and had a national reputation for what he called "de-scheming schemers."
A ruddy, barrel-chested, quick-witted Irishman, Dardis had grown up tough in upstate New York. Subsequently he moved to Florida to investigate for the attorney general, became a police chief and then Gerstein's chief investigator. Gerstein said Dardis had "the greatest gift I've ever seen for putting the pieces of a crime puzzle together."
Dardis knew his way around race tracks. He had de-schemed fixers who fed drugged meatballs to dogs at Bonita Springs in the '60s and had busted a ring operating a "wire service" to past-post bookies at Hialeah in 1963. He had solved each case with stunning speed. "I'm not a step-over-the-body cop," Dardis said. "At five o'clock I don't go home."
Gerstein arranged a meeting late that afternoon with the Flagler attorney and a representative from the state pari-mutuel wagering division. The informer had contacted David Hecht, owner and general manager of Flagler, a week before, telling him somebody was "messing around with the outs book," cashing what must be fraudulent tickets at the past-performance window. The informer said they were "using the computer." He said he knew because he had a "close friendship" with the wife of one of the computer operators, a man named Hudson. He implied that he and Mrs. Hudson were not so close that he would refuse a reward for supplying evidence. He said he had no idea how long the scam had been going on or how many people were involved, but it was "big." He said that Hudson's wife mentioned the names "Caisse" and "Jacques."
Hecht had tried to run it down but drew a blank. The track was not short any money. All figures balanced. Counterfeit tickets were out of the question because they would have created deficits. Fearing a more sinister influence, Hecht had approached the IRS and the FBI. The former said, "It would take two years to run this down." The latter said it was outside its jurisdiction. At Gerstein's meeting, ominous possibilities were raised: "nationwide implications" and "millions and millions involved," with "Mafia involvement" suspected.
Dardis said that his inclination "when people start talking Star Wars" is to get out and get closer to the source. Keith Dodwell, executive vice-president of Autotote, had been flown in to aid in the investigation and was in a hotel near the airport. Dardis left the meeting and at 6 p.m. he was in Dodwell's hotel room.
"They're doing something with the computer," Dodwell said. "They're my people, and they're doing something."
"How do you know that?"
Dodwell gestured to a pile of papers on his bed, many of them torn and rumpled. He said it was the debris from the trash can in the computer room at Flagler. A week's collection of garbage.
He placed two sheets of printed material side by side on the bed, identifying them as Autotote computer printouts from an August race at Flagler. "These are supposed to be identical," Dodwell said. He ran his finger down the columns indicating the trifecta prices paid on the race. Computer "A" showed a winning payoff of $2,905, Computer "B" $14,527. "A" showed five winning tickets, "B" showed one.
" 'A' is official, but I think it's a phony," said Dodwell.
"Is it possible to change the figures in the computer?"
"Maybe you better tell me how everything works, Dardis said."
Dodwell said that in Autotote's dual computer system every bet was fed directly from the ticket windows into the computer memory via a scanner that monitored the various selling locations. Every 90 seconds the computers disgorged printouts containing every piece of betting information on the race.
Computer "A," said Dodwell, dispensed three copies, one for the mutuels room to check against the cash brought in after each race, one for the state to verify tax revenue, the track's take, breakage, etc. and one as a permanent record. Computer "B," he said, was a backup, used only in emergencies. The "B" copy was generally thrown away.
"If this is what I think it is," Dodwell said, holding up the two printouts, "they're tampering with the trifecta winners." He said the track had "the name of a guy who cashed a whole lot of trifecta wheel tickets at the outs window. I don't know who the guy is. They haven't told me. They're not telling me who the informer is, either." He said everybody suspected everybody else, and that Hecht was blaming Autotote for everything.
Dodwell said that if figures were altered in the computer, that had to be done by the operator, a man named Richard Hudson. Or his assistant, a man named Gilles Caisse.
"What about a 'Jacques'?"
"There's a guy named Jacques Lavigne in the mutuels room. I don't know him. He's not one of our employees."
The two men ordered sandwiches from room service and began rummaging through the litter on the bed. Past midnight, Dardis began to puzzle over a small sheet of paper with twin columns of letters, one under the word MAT, the other under EVE. A date was imprinted at the top of the sheet. On either side of each column were handwritten figures that had been totaled at the bottom. Dodwell said the slip of paper that he had pieced together earlier was an official "code sheet," run off daily for each machine. The columns of letters—BCLT, RSVD, etc.—indicated which code stick had been placed in the mutuel machines before the matinee (MAT) and evening (EVE) programs on the date indicated.
Using figures written down on another sheet by one of the conspirators from a Computer "A" printout, Dodwell compared the figures noted on the code sheet with the trifecta payoffs for that day, allowing for multiple ticket "purchases" in some races. The figures matched. He said the totals at the bottom apparently represented the amount "won" by the figure writer—over $16,000 for the day.
On Saturday morning, Sept. 3, the last day of the Flagler meeting, Dardis arrived at the track early and slipped in through a side door and up a private elevator to Hecht's office. If fraudulent tickets were being cashed, he not only had to know how, but he also had to link the tickets with the computer room, and he had to do it in a hurry.
He stayed the weekend, sleeping on the floor of a conference room Hecht provided. Hecht also provided ticket stubs and 5754 forms signed by the man who had cashed the unusual number of trifecta tickets: an insurance salesman named Wayne Murray. Murray had enjoyed a breathtaking run of luck. He had cashed trifecta tickets worth upwards of $150,000.
Sealed in the conference room, not sure what he was looking for, Dardis began weeding through large boxes of printouts, 5754 forms and cashed tickets. It was slow, painstaking work. Ticket after ticket, form after form. He called in two assistant investigators to help. Late Sunday night he found it.
In the columns of ciphers on the computer printout for the seventh race of the Flagler matinee of Aug. 30, he found that Machine No. 507, located in a trifecta ticket window in the grandstand, had sold no trifecta wheel tickets: the entry said "0." But when Dardis matched that figure against his collection of 5754 forms, he saw that Murray had cashed four trifecta tickets worth $12,000 on that race—and each ticket wheel bore the signet of Machine No. 507.
Dardis unclipped the tickets from the 5754 form and compared them with the test ticket run off at each machine before each race program. The test tickets are kept on file. The date and code numbers were consistent, but the impressions were not. With the naked eye, Dardis could see subtle differences. He called in a lab expert, who confirmed the suspicion: yes, the tickets were printed by the same machine. No, they were not printed at the same time. The code was the same but the stick had been changed.
For Dardis, this proved two things: 1) Someone had come back to Machine 507 after the computers had been turned off, reinserted a code stick and manually printed the four tickets Murray had cashed; 2) The thief did not think to look on the printout to see whether any legitimate tickets were sold at Machine 507 for that race.
This meant that Wayne Murray's tickets were fraudulent.
Labor Day morning, Dardis called Gerstein. Although he had not slept, he was wide awake. Dardis said, "I not only know they're doing it, I know how. And here's the hell of it—they're not cheating the track, they're cheating the other bettors. They could go on forever and nobody would ever know."
He said he had an IRS form showing that a part-time carpenter named Leon E. Rodriguez had won a $2,900 trifecta on the seventh race on Aug. 30. "Mr. Rodriguez probably jumped up and down for joy," Dardis said, "but what he didn't know was he had the only legitimate winning trifecta ticket bought on that race. The race had a trifecta pool of nearly $15,000. His ticket wound up with four electronic partners that cheated him out of $9,000. The track got its share, the state got its share. But Mr. Rodriguez was taken, and doesn't even know it."