Categories
BLOG

algorytm lotto

The Man Who Cracked the Lottery

When the Iowa attorney general’s office began investigating an unclaimed lottery ticket worth millions, an incredible string of unlikely winners came to light – and a trail that pointed to an inside job.

The Money Issue

The Baby-Formula Crime Ring

It’s pricey, it’s portable, its users need it constantly, and retailers love to buy it at a discount. All of which makes it a perfect product to steal.

The Pain Hustlers

Insys Therapeutics paid millions of dollars to doctors. The company called it a “speaker program,” but prosecutors now call it something else: a kickback scheme.

The Billion-Dollar Bank Job

In 2016, a mysterious syndicate tried to steal $951 million from Bangladesh’s central bank – and laid bare a profound weakness in the system by which money moves around the world.

The Man Who Cracked the Lottery

When the Iowa attorney general’s office began investigating an unclaimed lottery ticket worth millions, an incredible string of unlikely winners came to light – and a trail that pointed to an inside job.

The White-Collar-Crime Cheat Sheet

How the biggest scammers get away with it.

The Man Who Cracked the Lottery

By REID FORGRAVE
Illustrations by FRANCESCO FRANCAVILLA MAY 3, 2018

When the Iowa attorney general’s office began investigating an unclaimed lottery ticket worth millions, an incredible string of unlikely winners came to light – and a trail that pointed to an inside job.

The file landed on Rob Sand’s desk with something less than a thud. Despite holding the contents of an investigation still open after more than two years, the file was barely half an inch thick. “Happy birthday,” his boss said.

It was not Rob Sand’s birthday. His boss, an Iowa deputy attorney general named Thomas H. Miller, was retiring in July 2014 after nearly three decades of prosecuting everything from murder to fraud. He hired Sand about four years earlier and made him the youngest prosecutor in a nine-attorney team that handled challenging cases all over the state. Now Miller was offloading cases to colleagues. This one, having to do with a suspicious lottery ticket worth $16.5 million, was full of dead ends. Investigators didn’t even know if a crime had been committed. The most tantalizing pieces of evidence were on a DVD: two grainy surveillance clips from a gas station. Sand slid the disc into his laptop and pressed play.

A man walked into a QuikTrip convenience store just off Interstate 80 in Des Moines. It was a weekday afternoon, two days before Christmas. The hood of the man’s black sweatshirt was pulled over his head, obscuring his face from two surveillance cameras overhead. Under the hoodie, he appeared to be wearing a ball cap; over the hoodie, he wore a black jacket. The man grabbed a fountain drink and two hot dogs.

“Hello!” the cashier said brightly.

The man replied in a low-pitched drawl, a voice that struck Sand as distinct: “Hell-ooooh.”

“Couple hot dogs?” the cashier asked.

“Yes, sir,” the man replied quietly, his head down.

The man pulled two pieces of paper from his pocket. They were play slips for Hot Lotto, a Powerball-like lottery game available in 14 states and Washington, D.C. A player — or the game’s computer — picked five numbers between 1 and 39 and then a sixth number, known as the Hot Ball, between 1 and 19. The prize for getting the first five numbers right was $10,000. But a much larger prize that varied according to the number of players who bought tickets went to anyone who got all six numbers right. The record Hot Lotto jackpot of nearly $20 million had been claimed in 2007. The jackpot at the time of this video was approaching the record. The stated odds of winning it were one in 10,939,383.

The cashier took the man’s play slips, which had already been filled out with multiple sets of numbers. At 3:24 p.m., the cashier ran the slips through the lottery terminal. An older man with a cane limped by the refrigerated section. A bus drove by. The cashier handed over his change. Once outside, the man pulled down his hood and removed his cap, got into his S.U.V. and drove away. The gas-station parking lot gleamed; there had been snow flurries that afternoon.

Two years into the case, that was virtually all the investigators had. Sand watched the video again and again, trying to pick up every little detail: the S.U.V.’s make; the man’s indistinct appearance: most likely in his 40s, and 100 pounds overweight, maybe more; the tenor of his voice.

Sand, a baby-faced Iowan who turned down Harvard Law School for the University of Iowa College of Law, had a background that seemed perfect for the case: a high school job writing computer code and doing tech support, a specialty in white-collar crime. His recent cases included securities fraud and theft by public officials.

The ticket in the video was purchased on Dec. 23, 2010. Six days later, the winning Hot Lotto numbers were selected: 3, 12, 16, 26, 33, 11. The next day, the Iowa Lottery announced that a QuikTrip in Des Moines had sold the winning ticket. But one month after the numbers were drawn, no one had presented the ticket.

The Iowa Lottery held a news conference. Phone calls poured in; dozens of people claimed to be the winner. Some said they had lost the ticket. Others said it was stolen from them. But lottery officials had crucial evidence that wasn’t publicly available: the serial number on the winning ticket and the video of the man buying it. One by one, they crossed off prospective claimants. One caller said his friend was a regular Hot Lotto player who had just died in a car wreck — should he go to the junkyard to search through his deceased friend’s car?

Three months after the winning ticket was announced, the lottery issued another public reminder. Another followed at six months and again at nine months, each time warning that winners had one year to claim their money. “I was convinced it would never be claimed,” says Mary Neubauer, the Iowa Lottery’s vice president of external relations. Since 1999, she had dealt with around 200 people who had won more than $1 million; she’d never seen a winning million-dollar ticket go unclaimed. “And then comes Nov. 9, 2011.”

A man named Philip Johnston, a lawyer from Quebec, called the Iowa Lottery and gave Neubauer the correct 15-digit serial number on the winning Hot Lotto ticket. Neubauer asked his age — in his 60s, he said — and what he was wearing when he purchased the ticket. His description, a sports coat and gray flannel dress pants, did not match the QuikTrip video. Then, in a subsequent call, the man admitted he had “fibbed”; he said he was helping a client claim the ticket so the client wouldn’t be identified.

This was against the Iowa Lottery rules, which require the identities of winners to be public. Johnston floated the possibility of withdrawing his claim. Neubauer was suspicious: The winner’s anonymity was worth $16.5 million?

One year to the day after the winning numbers popped up on the random-number-generator computers — and less than two hours before the 4 p.m. deadline — representatives from a prominent Des Moines law firm showed up at the Iowa Lottery’s headquarters with the winning ticket. The firm was claiming the ticket on behalf of a trust. Later, the Iowa Lottery learned that the trust’s beneficiary was a corporation in Belize whose president was Philip Johnston, the Canadian attorney. “It just absolutely stunk all over the place,” says Terry Rich, chief executive of the Iowa Lottery. The Iowa attorney general’s office and the Iowa Division of Criminal Investigation opened a case.

In an interview in Quebec City, Johnston told investigators that he had been contacted about the ticket by a Houston attorney named Robert Sonfield. Johnston also pointed investigators toward a Sugar Land, Tex., businessman named Robert Rhodes. A trip to Texas by Iowa investigators proved fruitless; during their several days there, both Sonfield and Rhodes managed to avoid them.

By the time the file ended up on the desk of Rob Sand in 2014, the case had acquired cultlike status in his office. It was spoken about with gallows humor: “We’ll find the guy who bought the ticket ended up getting offed,” Sand said. “That’s what this is going to turn out to be, a murder case.”

Miller had mentored Sand and saw in him a kindred spirit, someone for whom practicing law was a calling. Sometimes Sand’s moral compass was so steady that he came off as a square; his brothers-in-law nicknamed him Baby Jesus. Sand grew up in Decorah, in northeast Iowa, the son of a small-town doctor who still made house calls. He wanted to get into white-collar criminal prosecution because it focused not on crimes of desperation but on crimes of greed. “Crimes against gratitude,” Sand called them.

But all he had was grainy video of a man buying a lottery ticket worth $16.5 million. “We only had one bullet left in our revolver,” Sand says, “and that was releasing the video.”

On Oct. 9, 2014, nearly 46 months after the man in the hoodie left the QuikTrip, the Iowa Division of Criminal Investigation put out a news release that included a link to a 74-second clip of the surveillance footage. A few days later, in Maine, an employee of the Maine Lottery opened an email forwarded from his boss. The employee recognized the distinct voice in the video: It belonged to a man who had spent a week in the Maine Lottery offices a few years earlier conducting a security audit.

In Des Moines, a web developer at the Iowa Lottery who watched the video also recognized that voice: It belonged to a man she had worked alongside for years. A receptionist in another lottery office handed her earbuds to Noelle Krueger, a draw manager, and told her to listen. “Why am I listening to a video or listening to a tape of Eddie?” Krueger replied.

By Eddie, she meant Eddie Tipton, the information-security director for the Multi-State Lottery Association. The organization runs lotteries for 33 different states plus the District of Columbia, Puerto Rico and the U.S. Virgin Islands. It was based in the Des Moines suburbs. Among the games it ran was the Hot Lotto.

Eddie Tipton cut a big, friendly figure around the office of the Multi-State Lottery Association. He grew up in rural Texas, but while his siblings were outside, he was always in his room, fiddling with his computer. He was a paranoid sort who rarely paid with credit cards, who worried about people tracing his identity. But he always wanted people to like him. When a co-worker was in a bad mood, one colleague said, Tipton would pat him on the shoulder and say, “I just want you to know I’m your friend.”

Tipton built a 4,800-square-foot, $540,000 house in the cornfields south of Des Moines. The house had five bedrooms and a huge basement, including a pool table, a shuffleboard table, a stadium-style home theater with couches and a space he considered turning into a basketball court. Friends wondered why a single man needed such a big house and how he could afford it on a salary just shy of $100,000 a year. In private moments, Tipton told them he was lonely and wanted a family more than anything, so he poured his savings into the house he hoped to fill with a wife and children. But the right partner never came. Instead, he hosted office Christmas parties, and he constantly asked friends to visit. His family, still in Texas, checked on him frequently.

His life revolved around his job. The Multi-State Lottery Association was a small organization, and Tipton felt overextended. He wrote software and worked on web pages. He handled network security and firewalls. And he reviewed security for lottery games in nearly three dozen states. He was putting in 60-hour weeks and staying at the office until 11 p.m.

When Ed Stefan, the chief information officer and chief security officer at the Multi-State Lottery Association, saw the surveillance video, he didn’t want to believe it. This wasn’t just some co-worker. This was Eddie Tipton, a man he had known for more than two decades, since they were in calculus class at the University of Houston. Stefan met his future wife while he and Tipton were on a charity bike ride in Texas; Tipton would later be in their wedding. Stefan helped Tipton get his job at the association. They bought some 50 acres of land together and built adjacent houses. They even applied for a joint patent for computer-based lottery security.

Stefan watched the convenience-store video for the first time after a former co-worker sent him the link that had been released by prosecutors. That just can’t be Eddie, he thought. Then: That’s Eddie. Why is he wearing a hoodie? I’ve never seen Eddie in a hoodie. Stefan got sick to his stomach: His friend, a man with deep knowledge of the computers that ran the lottery, was there onscreen buying a ticket that would be worth $16.5 million. Later, Stefan would tell investigators it was like finding out your mother was an ax murderer. He felt betrayed.

Jason Maher was another friend and colleague who didn’t want to believe what he was seeing on the video. He and Tipton had met at Taki, a Japanese restaurant outside Des Moines that they both frequented. The lifelong computer aficionados and gamers hit it off; Tipton joined Maher’s gaming clan, and they spent hours playing the multiplayer online game World of Tanks. Tipton suggested Maher apply for a job at the lottery association as a network engineer. Tipton, Maher told me, “had a heart of gold.”

So when Maher saw the video and heard that familiar low-pitched voice, he did what a computer whiz does. “That night I sat down — there’s no way Eddie did this,” Maher said. “There’s got to be something wrong.” He put the file of the surveillance tape into audio software, removed white noise and isolated the voice. Then he took footage from security cameras in his house — Tipton had just visited the night before — and compared Tipton’s voice in that footage with the convenience-store video. “It was a complete and utter match, sound wave and everything,” Maher said. The next day, he went to the QuikTrip where the ticket was purchased and measured the dimensions of the tiles on the floor, the height of the shelving units, the distance between the door and the cash register. He used the results to compare the hand size, foot size and height of the man in the video with the man he had become friends with.

“When the F.B.I. guys came in, I wanted to be able to tell them it wasn’t Eddie,” Maher said. “Once I did this, it was like, ‘Well, [expletive] — it’s Eddie.’ ”

In November 2014, state investigators showed up at Tipton’s office. They asked him whom he knew in Houston. He told them about his family — mother, sister and brothers, including Tommy, a former sheriff’s deputy turned justice of the peace near the Texas Hill Country. He did not mention Robert Rhodes, the man who initially passed the $16.5 million ticket to an attorney. By searching Tipton’s LinkedIn profile, investigators found that Tipton had been employed at Rhodes’s Texas-based software company, Systems Evolution, for six years as its chief operations officer. In fact, the two were best friends and vacationed together.

Tipton was arrested in January 2015 and charged with two felony counts of fraud. Half a year later, on a hot, sticky July morning, Rob Sand stood before a jury at the Polk County Courthouse. “This is a classic story about an inside job,” he began. “A man who by virtue of his employment is not allowed to play the lottery, nor allowed to win, buys a lottery ticket, wins and passes the ticket along to friends to be claimed by someone unconnected to him. This story, though, has a 21st-century twist.”

The prosecution knew Tipton had bought the winning ticket. The video, specifically the distinct voice that colleagues had recognized, made that pretty clear. So did cellphone records, which showed Tipton was in town that day, not out of town for the holidays as he claimed, and that he had been on the phone for 71 minutes with Robert Rhodes, the man who briefly had possession of the ticket. Investigators believed he’d fixed the lottery. But how?

Jason Maher, Tipton’s gaming buddy, told them about Tipton’s interest in rootkits: malicious software that can be installed via flash drive in order to take control of a computer while masking its existence until it deletes itself later. Sand theorized that Tipton went into the draw room six weeks before the big jackpot and, despite the presence of two colleagues, managed to insert a thumb drive into one of the two computers that select the winning numbers. That thumb drive contained the rootkit; the rootkit allowed Tipton to direct which numbers would win the Hot Lotto on Dec. 29, 2010.

Tipton’s defense attorney, Dean Stowers, called this the “Mission: Impossible” theory. Stowers characterized the story of a malicious, self-destructing rootkit (“magic software”), installed while two colleagues looked on, as preposterous. His closing arguments referenced a quotation attributed to Albert Einstein: “Logic will get you from A to B. Imagination will take you anywhere.”

But Sand called Stowers’s focus on this complicated rootkit theory a red herring. Sand told the jury to focus on the many ways Tipton could have fixed the lottery: He wrote the code. He had access to the random-number-generator machines before they were shipped to other states. You don’t have to understand the exact technology to convict Tipton, Sand argued; you just have to realize the near-impossible coincidence of the lottery security chief’s buying a winning ticket and that ticket’s being passed to his best friend. The prosecution had to prove only that Tipton tried to illegally buy lottery tickets as a Multi-State Lottery Association employee and tried to claim the prize through fraudulent means.

The jury found him guilty on July 20, 2015. He would be sentenced to 10 years in prison, and he would appeal. The State Supreme Court later dismissed his conviction on one charge, tampering with lottery equipment, and the case was sent back to District Court.

Six weeks after the trial concluded, Sand had returned to his desk. It had been a busy summer. But in the back of his mind, he was still thinking about Eddie Tipton. Sand knew white-collar criminals aren’t usually caught on their first attempt. The fact that Tipton’s attorney had demanded a 90-day speedy trial, an unusual maneuver that cut short the prosecution’s time to investigate, made Sand suspicious. His gut said other fraudulent lottery tickets were out there.

One morning Sand’s office phone rang, and an area code he recognized popped up: 281, from Texas, where Tipton used to live. Sand picked up. The caller had a drawl and told Sand he’d seen an article in the La Grange, Tex., newspaper about Tipton’s conviction. “Did y’all know,” the tipster asked, “that Eddie’s brother Tommy Tipton won the lottery, maybe about 10 years back?”

Richard Rennison’s phone rang at the F.B.I. office in Texas City, a port town on the shore of Galveston Bay. Sand was on the line, inquiring about a case that Rennison, a special agent for the bureau, investigated a decade before. At the time, it turned out to be nothing. But the case still stuck in Rennison’s mind. “Hey,” Rennison replied, “that’s my Bigfoot case.”

A man named Tom Bargas had contacted local law-enforcement authorities in early 2006 with a suspicious story. Bargas owned 44 fireworks stands in Texas. Twice a year — after the Fourth of July and after New Year’s — he had to handle enormous amounts of cash, more than a half-million dollars at once. A local justice of the peace who shod Bargas’s horses called him around New Year’s. The justice of the peace caught Bargas off guard: “I got half a million in cash that I want to swap with your money.” “What’s wrong with your money?” Bargas replied.

What’s a justice of the peace who makes around $35,000 a year doing with that much cash? Bargas thought. He called the sheriff and the police, who called the F.B.I. Soon, the bureau contacted Bargas. Federal agents outfitted him with a wire. Bargas met with the man, who pulled out a briefcase filled with $450,000 in cash, still in their Federal Reserve wrappers. As the F.B.I. listened, Bargas swapped $100,000 of worn, circulated bills for $100,000 of the man’s crisp, unused bills. To the F.B.I., this smelled like public corruption, and they went to work investigating the serial numbers on the bills.

One day a couple of months later, Rennison got a call from the Fayette County sheriff in La Grange, a place best known for the Chicken Ranch, the brothel that inspired “The Best Little Whorehouse in Texas.” The sheriff was laughing so hard he could hardly speak. He told Rennison the justice of the peace was holed up in a Houston hospital with two shattered legs. He had fallen 31 feet out of a tree. He had been hunting Bigfoot.

“My grandmother was raised on a farm in Arkansas where this creature would come in and harass all the farm animals,” this man later told investigators. “My grandmother would tell me all these stories of this animal that harassed my family.” He went on: “I started hitting the woods. It was always that doubt in your mind. And then something happened to me in Louisiana where I actually watched these animals for a couple of hours, and I’ve been hooked ever since.”

Rennison visited the man in the hospital and then set up an interview once he was discharged. The man was a member of the Bigfoot Field Researchers Organization. He told Rennison he’d won the lottery in Colorado while on a Bigfoot hunt. He was on the outs with his wife and was trying to keep the lottery winnings from her. A Bigfoot-hunting friend claimed the prize in exchange for 10 percent of the money. It all checked out. Case closed.

“Right before I leave, we’re still sitting down at this nice conference table, and he looks over at his attorneys and says, ‘Can I show him?’ ” Rennison recalled. “Hanging off the back of his chair is a plastic grocery bag. He pulls out a plaster cast of a footprint.”

Rennison put the footprint next to his own foot. They were roughly the same size. “That doesn’t look like Bigfoot,” the F.B.I. agent said. “It was a juvenile,” the man snapped. The man’s name was Tommy Tipton.

Now the hunt was on for more illicitly claimed tickets. Iowa investigators noticed that the friend who claimed the $568,990 Colorado Lottery prize for Tommy Tipton, a man named Alexander Hicks, was dead. “We first thought, Whoa — this is our first body related to this case,” Sand says. It wasn’t. Hicks had died of cancer.

The investigators collected a decade’s worth of winners from lotteries around the country associated with the Multi-State Lottery Association. They loaded data from approximately 45,000 winning tickets into Microsoft Excel spreadsheets and searched for any connections to Eddie Tipton. They reviewed Tipton’s Facebook friends, pulled phone records and looked for matches with the spreadsheet.

In September 2015, they learned that a $783,257.72 payout for Wisconsin’s Very Own Megabucks Game had been claimed in early 2008 by a Texas man named Robert Rhodes, who wanted to deposit it into the account of a limited-liability corporation. That drawing took place on Dec. 29, 2007 — the same day the winning numbers on Tipton’s $16.5 million Iowa ticket were selected three years later. Rhodes was Eddie Tipton’s best friend. Another hit.

One evening over the holidays, Sand was at his parents’ house working on his laptop, sifting through records, using search commands on his computer again and again. He noticed that a Kyle Conn from Hemphill, Tex., won a $644,478 jackpot in the Oklahoma Lottery some years back. Tommy Tipton had three Facebook friends named Conn. Sand got a list of possible phone numbers for a Kyle Conn and cross-referenced them with Tommy Tipton’s cellphone records. Another hit.

Investigators noticed two winning Kansas Lottery tickets for $15,402 apiece were purchased on Dec. 23, 2010 — the same day Tipton had purchased the Iowa ticket, and the same day that cellphone records indicated he was driving through Kansas on the way to Texas for the holidays. One of the winning tickets was claimed by a Texan named Christopher McCoulskey, the other by an Iowa woman named Amy Warrick. Each was a friend of Eddie Tipton’s.

Early one morning, Sand and an investigator knocked on the woman’s door. She told them she’d gone on one date with Tipton, but their relationship became platonic. Tipton told her he wasn’t able to claim a winning lottery ticket because of his job. If she could claim it, Tipton said, she could keep a significant portion as a gift for her recent engagement.

“You have these honest dupes,” Sand says. “All these people are being offered thousands of dollars for doing something that’s a little bit sneaky but not illegal.” Investigators in Iowa now had six tickets they figured were part of a bigger scam. But the question remained: How did it work?

Investigators in Wisconsin discovered they still had the random-number-generator computers used for the 2007 jackpot sitting in storage. Unlike Iowa’s computers, the hard drives had not been wiped clean; their software was the same as the day Robert Rhodes won $783,257.72. Wisconsin enlisted a computer expert named Sean McLinden to conduct an investigation that included forensic analysis and reverse engineering.

On Jan. 7, 2016, Sand’s phone rang. It was David Maas, an assistant attorney general in Wisconsin. He told Sand to check his email. Maas had sent him an attachment with 21 lines of “pseudocode,” a common-language translation of McLinden’s forensic analysis that showed part of Tipton’s malicious computer code. The code was small enough that it would not radically change the size of the file, which might create suspicion. And the code hadn’t been hidden. You just needed to know what to look for.

“This,” Maas says, “was finding the smoking gun.”

The smoking gun would help lead to a guilty plea from Tipton. In the plea deal, Sand insisted that Tipton come clean about how he fixed the lottery. This could help the lottery industry improve its security. If Tipton lied — or if another fraudulent ticket were found later — the deal would be voided, and Tipton would be subject to further charges.

Tipton’s program was called QVRNG.dll: Quantum Vision Random Number Generator. In Tipton’s telling, his wasn’t an evil plan to get rich. This was just a computer nerd’s attempt to crack the system. “It was never my intent to start a full-out ticket scam,” Tipton told investigators. “It occurred to me, like, Wow, I could do this, I could be making a living doing this.” He went on: “If this was, like, some mob-related thing, I’d just give this information to the mob and they would go out and win the lotteries left and right. Nobody would know. But that — I don’t have any mob ties. I don’t know anybody. I gave tickets to friends or family.”

More than a decade ago, Tipton told them, he walked past one of the organization’s accountants at the Multi-State Lottery Association. Tipton was conservative, the accountant liberal, and they often ribbed each other.

“Hey, did you put your secret numbers in there?” the accountant said, teasing Tipton.

“What do you mean?”

“Well, you know, you can set numbers on any given day since you wrote the software.”

And that’s when the idea first came. “Just like a little seed that was planted,” Tipton said in his proffer. “And then during one slow period I just had a — had a thought that it’s possible, and I tried it and I put it in.” The code wasn’t a brazen “Mission: Impossible” stunt of sneaking into the draw room with a malicious thumb drive. It was a simple piece of code, partly copied from an internet source, inserted by the one man responsible for information security at an organization that runs three dozen United States lotteries.

Here’s how the Multi-State Lottery Association’s random-number generators were supposed to work: The computer takes a reading from a Geiger counter that measures radiation in the surrounding air, specifically the radioactive isotope Americium-241. The reading is expressed as a long number of code; that number gives the generator its true randomness. The random number is called the seed, and the seed is plugged into the algorithm, a pseudorandom number generator called the Mersenne Twister. At the end, the computer spits out the winning lottery numbers.

Tipton’s extra lines of code first checked to see if the coming lottery drawing fulfilled Tipton’s narrow circumstances. It had to be on a Wednesday or a Saturday evening, and one of three dates in a nonleap year: the 147th day of the year (May 27), the 327th day (Nov. 23) or the 363rd day (Dec. 29). Investigators noticed those dates generally fell around holidays — Memorial Day, Thanksgiving and Christmas — when Tipton was often on vacation. If those criteria were satisfied, the random-number generator was diverted to a different track. Instead, the algorithm would use a predetermined seed number that restricted the pool of potential winning numbers to a much smaller, predictable set of numbers.

So Tipton knew what no one else knew: For the Iowa Hot Lotto drawing on Dec. 29, 2010, there weren’t really 10,939,383 sets of possible winning numbers. There were only a few hundred. Late at night before a draw that fulfilled his criteria, Tipton stayed in his messy, computer-filled office. He set a test computer to the date and time of the coming draw, and he ran the program over and over again. For the first lottery he rigged, the Nov. 23, 2005, drawing in Colorado, Tipton wrote down each potential set of winning numbers on a yellow legal pad. He handed the pad — each sheet had 35 or so sets of six numbers — to his brother.

It was a cheat sheet; instead of playing every possible number combination to ensure one combination won, he had to play only a few hundred. “If you want a chance to win, you need to play all of these,” Tipton told his brother. “I don’t know if any of them will win, but you’re going anyway” — his brother was about to go on a Bigfoot-hunting trip to Colorado — and “these have a good chance of winning based on my analysis.

“Play them,” he said. “Play them all.”

On a clear, blue summer day in Des Moines last year, Eddie Tipton, a square-shaped, balding man who was then 54, trudged up the stairs of the Polk County Courthouse. He wore bluejeans and a short-sleeved salmon-colored button-up shirt, untucked and unbuttoned, with a blue T-shirt underneath. His hands were shoved in his pockets, and his head was down. He had accepted a plea agreement for masterminding the largest lottery scam in American history: one count of ongoing criminal conduct, part of a package deal that allowed his brother to be sentenced to only 75 days. Tipton was here for his sentencing.

In statements to prosecutors, Tipton painted himself in the most generous way possible, a kind of coding Robin Hood, stealing from the lottery and helping people in need: his brother who had five daughters, his friend who’d just gotten engaged. “I didn’t really need the money,” Tipton said. The judge noted that Tipton seemed to rationalize his actions — that Tipton didn’t think it was necessarily illegal, just a taking advantage of a hole in the lottery’s system. It wasn’t all that different, Tipton believed, from insider trading, except laws didn’t specifically prohibit him from fiddling with the random-number-generator code. His attorney equated what he did with counting cards at a casino. Tipton wasn’t robbing the casino at gunpoint; he was cheating the house.

The other side disagreed. Tipton was “nothing but a common thief who happened to be handed the keys to the candy store,” Miller, Sand’s former boss, told me. “It’s not a case of Sherlock Holmes’s archnemesis, Moriarty, being a criminal genius. This is just a regular schlub, who is a thief who happens to have knowledge of computer security.”

From Tipton’s point of view, it was complicated. He had done something to see if he could do it. To his surprise, it worked. He said he inserted that code only once; after the code was approved by Gaming Laboratories International, machines containing it were shipped all over the country. He had created a beast and sent it into the world. “You plant that money tree in your backyard,” as Maas, the Wisconsin prosecutor, put it, “and it’s hard not to keep picking at it.”

In interviews, investigators had asked Tipton if he was proud of the success of his code. “It was more like I’m ready for it to be gone,” Tipton said. “It was never my intent to go out there and start winning all these lotteries. It was just, like I said, step by step it happened.”

At sentencing, the judge asked if Tipton had anything to say. After a long pause, Tipton cleared his throat. Family members and former co-workers were in the courtroom. “Well,” Tipton said matter-of-factly, “I certainly regret my actions. It’s difficult to say that with all the people behind me that I hurt. And I regret it. I’m sorry.” As the case was being litigated, Tipton had confessed to friends that he was racked with guilt. At another point during the proceedings, Tipton leaned across a divide and extended his hand to Sand. Sand took the handshake as a sign of respect, as if Tipton had thought he outsmarted the system but the system figured him out. On the day of his sentencing, Tipton told the judge he’d been taking classes to go into ministry. A deputy placed Tipton in handcuffs and led him away.

Earlier in the summer, Tipton sat in a conference room with Sand and law-enforcement and lottery officials to give his full confession, as promised in his plea agreement. Eventually he would head to Clarinda Correctional Facility in southern Iowa, near the Missouri border, where he remains today, Offender No. 6832975. (Through his attorney, he declined interview requests for this article; Tipton did not respond to nearly a dozen emails through the prison email system.)

During a lunch break in Tipton’s hourslong confession, Sand and others involved in the prosecution walked a few blocks to the High Life Lounge. They ordered bacon-wrapped tater tots to celebrate. “Eddie sees himself as much brighter than the rest of the world, the sharpest tool in the toolbox,” Sand says. “It’s the kind of thing I see in white-collar case after white-collar case, people who think they’re better than everybody else, that people trust them and love them, and that no one will be able to figure this out.”

The judge sentenced Tipton to a maximum of 25 years in prison. His restitution payments to the various state lotteries came to $2.2 million even though, according to his attorney, Tipton pocketed only around $350,000 from the scam, the rest going to those who claimed the tickets. (Prosecutors did not believe that, pointing to Tipton’s massive house as well as the fact that Tipton and his brother owned 11 pieces of property either jointly or individually in Fayette County, Tex.) In Iowa, which has indeterminate sentencing, a 25-year sentence could mean Tipton is released much sooner; Sand expects Tipton to be released by the Iowa Board of Parole within seven years.

Sand says he felt a deep intellectual satisfaction in solving the puzzle: “The justice system at its best is really about a search for truth.” But it couldn’t go back in time and correct wrongs. At the end of this yearslong case, he came to a realization: He had grown weary of dealing with criminals. “In so much darkness,” Sand says, “I started to lose my light.”

A few months after the highest-profile case of his career, Sand went up to his boss and quit. He had decided to run for state auditor in the coming November election, so he could make positive changes. If he wins, he will be investigating government waste, abuse and fraud. “There’s no way I would make a move to get away from the darkness of prosecution without finishing this case first,” Sand says. “So finishing it, to me, was not merely satisfying. It was liberating.”

Sign up for our newsletter to get the best of The New York Times Magazine delivered to your inbox every week.

Reid Forgrave is a writer based in Minnesota. He last wrote for the magazine about the fight over proposed mines near the Boundary Waters.

When the Iowa attorney general’s office began investigating an unclaimed lottery ticket worth millions, an incredible string of unlikely winners came to light – and a trail that pointed to an inside job.

Algorytm lotto

Churov V.E., Arlazarov V.L., Solov’ev A.V. “Results of election. Analysis of electoral preferences”

In the article methodological approach to definition of dominant tendencies of electoral preferences on basis of statistical data is stated. Results of analysis of final reports of regional electoral commissions on basis of factor of territorial accessory voters are presented. Conclusions about far directions of researches of electoral preferences are made, and also brief description of construction of mathematical model of electoral preferences is given.

Keywords: electoral preferences, mathematical model, statistical data, regional electoral commission.

Khovansky A.G. “Polyhedrons and algebra”

Theory of polyhedrons of Newton links geometry of polyhedrons with algebraic geometry. It is told in the article about some results of this theory.

Keywords: theory of polyhedrons of Newton, algebraic geometry.

Podrabinovich A.Ya. “About completeness of strategies of search of conclusion for set of generalized horn clauses”

In the article generalization of concept of set Horn clauses (Hn-set) is considered. Generalization is offered to strategy of entrance resolution (Gn-resolution). Completeness of Gn-resolution on Hn-set is proved.

Keywords: set Horn clauses (Hn-set), entrance resolution (Gn-resolution), сomplete-ness of Gn-resolution.

Danilenko A.Yu. “Maintenance of safe functioning of big information systems”

This article is devoted to analysis of approaches to information systems architecture and development strategy to achieve necessary level of data security.

Keywords: information systems architecture, data security.

Romanov B.L., Slobodetsky D.Ya. “Architecture of module of software interface for euphrates document manage-ment system”

In the work problem of development of program module, describing classes of objects for information system subject area, and realizing software interface (API) for interaction with them is considered. Basic purpose of such module-branch of algorithms, realizing basic operations with objects from user interface and, the very, simplification of user interface development, and also increase of system reliability. Problems are described, arising at development of user system’s interface, and also way of them decision with the help appropriate organization of module interfaces.

Keywords: architecture, software interface, EUPHRATES Document Management System.

Akimova G.P., Pashkina E.V., Solov’ev A.V. “Methodology of modeling of information systems behavior under influence of catastrophic factors”

We describe methodology of modeling of large geographically distributed information systems under influence of catastrophic factors.

Keywords: stable against catastrophe, human factor, risk estimate, information sys-tem, modeling, probability, methodology, destabilizing factor.

Bogdanov D.A., Lagaeva Yu.A., Maksimov D.A., Pliskin E.L. “A distributed document-oriented budgeting system: an office solution”

Presented in the article is an office software based solution for the distributed document-oriented financial system for the planning and execution of corporate budget. End user experience is based on MS Excel 2003 application enhanced with a VSTO 2.0 based custom program extension. This program supports data exchange between corporate headquarters (HQ) and distant regional branches by means of e-mail or an FTP. An asymmetrical document version control method is implemented which gives preference to HQ staff in changing (approving) the documents received originally from the regional branch to be returned back to the branch office. Each location makes use of a separate quarterly data base. A special procedure for aligning quarterly data is implemented where necessary for moving and copying documents between databases. This procedure accounts for integrity of linked documents in each database when the document is executed in the next quarter than the one it was originated in. Also presented in the article is an algorithm for merging changes in the documents flowing between corporate headquarters and regional locations.

Keywords: office software, distributed document-oriented system, corporate budget.

Antonova А.А., Misyurev A.V. с”Cognitive dwarf 2.0: english and russian dependency parser”

This document describes a refined version of Cognitive Dwarf parsing program, which is designed for automatic morphological and syntactic analysis of English or Russian plain text. In its present version the program includes a new parsing algorithm, an improved set of syntactic relations and a simplier output format.

Keywords: рarsing program, automatic morphological and syntactic analysis.

Slobodetsky D.Ya., Polevoy D.V. “Practical aspects of formalization of complex data objects on example of problem of analysis of flat representation of tabulated information”

The article is devoted to the problems of building formal models for complex information objects. Tabular data representation is used as example.

Keywords: building formal models for complex information objects, tabular data rep-resentation.

Reingold L.A., Oleg A. Slavin “Transformation information resources of tenders from account in marketing”

At present legislation ed in Russia in sphere of the undertaking state and municipal bulk purchase. In these condition turns out to be claimed introducing the facilities to automations for increase the velocities and quality of the undertaking the competitive sale. Usually automation of the sale is considered in connection with account activity of the organizer and participant of the sale. However application appearing in process of the sale to information greatly broader. She can be used not only in context of the tendering, but also for decision of the other problems. Competitive information in the whole her(its) collections it is necessary to consider as new information resource, which particularities of the use are considered in article. Complex use given about mechanism of the shaping and satisfactions of requirements in goods and service capable to form the new level of the knowledges about operation social-economic system on practical and scientific level.

Keywords: automation of the sale, tender, marketing.

Tarkhanov I.A. “Theoretical model of not suppotred relational database”

In the article theoretical relational database model is considered, that becomes not supported in operation. Is considered as arising in such database problems influence on relation scheme. Probabilistic approach for estimation of search data is offered. Search algorithm, based on this model is resulted.

Keywords: not supported relational database, estimation of search data, search algorithm.

Reingold L.A. “Repository of spatial data concerning to region management”

Developing of new informational technologies brings new types of information for management practices and also for common life. For example, spatial data technolo-gies was using exclusively by specialists some time ago but today they getting wide spread occurrence for solving practical needs. There are different sources of spatial data and they could have different accuracy, re-liability, minuteness. That’s why it’s important to implement the data which becomes a reason for taking management decisions and for monitoring which is realizing on significant objects. Regions are one of the level of public administration where processing of spatial data becomes the most relevant. The most important questions of processing spatial data on region level are represented in the article. Also are shown problems of data im-plementation and creating of Repository of spatial data is suggested as an infrastruc-ture for consolidation spatial data and semantic which belongs to it for solving cur-rent and future tasks.

Keywords: repository, spatial data, region management, information technologies.

Akimova G.P., Pashkina E.V., Solov’ev A.V. “Methodology of estimates of quality of random number and series of numbers”

We describe methodology of finding probability of number or series of numbers generated by different types of random number generator to be random at the example of lotto machine.

Keywords: casting of lots, lotto machine, series of random numbers, statistical test, probability, uniform distribution, mean of distribution, random number generator, methodology.

Fedotov A.I. “Analysis of models of construction of pension systems”

In given article analysis of models of construction of pension systems with account of functionality and parameters is leaded. Basic attributes of pension systems of devel-oped countries are selected.

Keywords: model of construction of pension systems, functionality, developed countries.

Kostin A.S. “About concept of introduction erp-systems”

In given article general concept of introduction of standard Erp-System of industrial class is considered. Basic concepts of vital cycle of system behaviour are described. Processes and phases of introduction ERP-System according to methodology Oracle are resulted. The most important standards of introduction of information annexes are listed. Organization of designing of information system in accordance with state standards of Russian federation is described.

Keywords: Erp-System of industrial class, vital cycle, designing.

Bezmaternyh P.V., Nikolaev D.P., Postnikov V.V. “Fast algorithm for handwriting detection”

In this paper a problem of handwritten fragment detection in the prescribed zone of document is examined. Detection algorithm is proposed, that simply and computa-tionally effective solves stated problem in the presence of noise of different kinds.

Keywords: handwritten fragment, document, recognition, algorithm, noise.

Gavrikov M.B., Pestryakov N.V., Uskov A.V., Farsobina V.V. “Estimation for character recognition method based on polynomial regression”

This article is devoted to one of modern exact methods of character recognition, based on regression analysis. Presence of laws in exhibiting of estimations of recog-nition is revealed.

Keywords: character recognition, polynomial regression, estimation.

Gavrikov M.B., Pestryakova N.V., Uskov A.V., Farsobina V.V. “Average raster and vector–concepts for character recognition method based on polynomial regression”

For research of one of modern exact methods of character recognition based on poly-nomial regression, concepts of average raster and vector are entered. In the article analysis of character of recognition in context of these values is executed.

Keywords: character recognition, polynomial regression, average raster and vector.

Gavrikov M.B., Pestryakov N.V., Uskov A.V., Farsobina V.V. “Dependence of recognition accuracy and estimation from measure of discrep-ancy between bases of training and recognition”

In the article dependence of accuracy and estimation of recognition from degree of distinction between bases of training and recognition is researched. Studying is leaded for recog-nitions bases, that are received using four models of training base destruction.

Keywords: accuracy and estimation of recognition, bases of training and recognition.

Kotovich N.V. “Algorithms to clusterizations image symbol”

In article is described algorithms to clusterizations image recognized printed symbol, used in mechanism of the adaptive recognition of the document. Will Enumerated known methods to clusterizations object. The Described algorithm of the chain un-rolling, is brought proof to his(its) capacity to work, consisting in row of the positions about randomness of the choice of the initial object to clusterizations. It Is Discussed difficulty of the algorithm of the chain unrolling and acceptance to optimization to difficulties. It Is Considered question of the choice to functions of the distance. They Are Discussed acceptance to optimization of the calculation to functions of the dis-tance, being pseudosimmetrics of Hausdorf.

Keywords: klusterizations, adaptive recognition, algorithm of the chain unrolling, functions of the distance.

Kotovich N.V., Slavin O.A. “The algorithms of the recognition font in printed documen”

Algorithms of the building font is described In article on base result recognitions printed symbol, used in mechanism of the adaptive recognition of the document. They Are Given determinations of the font, allowing build the algorithms of the rec-ognition font, by means of which is produced the most further clusterization image symbol. It Is Paid attention particularity to primary categorization image symbol, necessary font for building, using to clusterizations image symbol and use linguistical mechanism. The considered problems, in decision which can be used mechanism of searching for font.

Keywords: klusterizations, adaptive recognition.

Slavin O.A. “Multipass recognition mixed printed text on example russian-english recognition”

In article is described algorithms of the recognition printed multilingual text, consid-ered on example of the recognition russian-english document. Considered using as font-driven, so and omnifont document. Using of self-training is Paid attention on base of the clusterizations image recognized symbol and use linguistical mechanism.

Keywords: klusterizations, self-training, linguistical mechanism.

Rakhmankulov V.Z., Akhrem A.A., Gerasimov V.V., Barsegov A.A. “Multidimensional wavelet analysis of images”

Basic properties of multidimensional spaces are described in the article. Some prob-lems of multidimentional wavelet functions application to the image analysis and pat-tern recognition are considered.

Keywords: wavelet function, image analysis, pattern recognition.

Rakhmankulov V.Z., Akhrem A.A., Barsegov A.A., Gerasimov V.V., Lebedev V.V. “Image compression using the wavelet theory for the case of images of the stamped parts”

Theoretical and practical issues of the image compression on the basis of the wavelet theory are considered in this article. The algorithm and example of image compres-sion are presented.

Keywords: wavelets, image compression.

Akhrem A.A. “Typical properties of some classes of design models applied to complex techno-logical systems”

Typical fundamental properties of the special classes of dynamic design models are investigated in this article. For these models various algebraic and geometrical crite-ria for asymptotic stability have been proved.

Keywords: dynamic design models, asymptotic stability, linear models, differential models.

Nikolayev P.P., Karpenko S.M., Nikolaev D.P.”Spectral models of color constancy: selection rules”

In this paper common principles of spectral models construction for color constancy are considered. Particularly, role of coordinate systems in spectral space is discussed. Uniqueness problem for solution of color constancy task is investigated. Finally, new class of hybrid color constancy models is suggested.

Keywords: color constancy, spectral models.

Gudkov V.Ju. “Mathematical description and identification of finger-print images”

In the work ways of description of images of fingerprints, advantages and lacks of ways are resulted, conceptual mathematical models of fingerprints. Key problems of description are selected. Methods of identification on mathematical models are resulted.

Keywords: fingerprint, identification.

Gudkov V.Ju. “Accuracy rating of automatic minutiae coding”

This paper presents a method to estimate the accuracy of minutiae detection. Method is used to preliminary learn the algorithms of fingerprints recognition.

Keywords: fingerprint, accuracy, estimation.

Algorytm lotto Churov V.E., Arlazarov V.L., Solov’ev A.V. “Results of election. Analysis of electoral preferences” In the article methodological approach to definition of dominant tendencies