ITBusiness.ca

Facial recognition software used for national security

Jeff Jonas knows the Las Vegas gambling industry inside and out. As the founder and chief scientist of Systems Research & Development (SRD), Jonas helped build numerous casino systems before 2005 when his company was purchased by IBM.

Big Blue was intrigued by SRD’s NORA system (Non-Obvious Relationship Awareness), a technology that uncovers relationships that can be exploited fraudulently for profit, such as connections between dealers and gamblers.

Now a distinguished engineer and chief scientist for IBM’s Entity Analytic Solutions, Jonas is still based in Las Vegas but is focused more on applying his technology to national security and the banking industry.

Speaking at the O’Reilly ETech conference on emerging technology in San Diego on Thursday, Jonas promised to reveal some, if not all, of the secrets he learned about the casino industry. Before the talk, he called some of his former clients to make sure certain details could be revealed.

“My idea today was to tell more about the casino industry than I ever told,” Jonas said.

After Jonas moved to Vegas in 1990, he met a man who said his job was to cheat casinos.

“I’m like ‘are you a card counter?’ He says ‘you don’t get it! That would be like marijuana. What I do is like heroin!’ I didn’t know anything about this. Then he proceeds to show me his disguises, all these glasses, his mustaches. And I’m like ‘this is going to be crazy.'”

Over the next 15 years Jonas helped pioneer facial recognition technology and various other systems in casinos such as the Bellagio, Treasure Island and Beau Rivage in Mississippi.

“Today possibly half the casinos in the world run something or another that I had my hand in,” he said.

Vegas seems to put an enormous focus on high-tech security, but in some ways the casinos are just doing enough to get by.

“They spend the minimum amount of money on security and surveillance,” Jonas said. “They’d rather buy three more slot machines and make money. They only mess with you if you’re really, really cheating.”

A casino like the Bellagio probably has 2,000 cameras connected to 50 monitors, with just a few people watching live surveillance, Jonas said. But the information is there to be scrutinized when casinos notice players winning unusually large amounts of money.

In one case a dealer — who said his family had been threatened – helped players rake in US$250,000 at a blackjack table when he used a deck of “perfectly ordered cards” that had been handed to him by one of the gamblers, according to Jonas.

“They didn’t detect this as it happened,” Jonas said. “Most of the videos the casinos collect are just used forensically. When the table loses a quarter of a million dollars they go back and replay it nice and slow, see that little piece of video, and it’s time to make some calls. In the old days it was the kneecaps, but those were the old, old days.”

Even if you are cheating, and they know you’re cheating, they might leave you alone if you’re not that good at cheating. Take card counting: While counting cards in one’s head is not illegal, a good card-counter in blackjack gains a statistical advantage over the house, and if the casino decides the counter is making too much money, he or she will be escorted off the premises.

But card counters have to be really good: One mistake an hour could swing the advantage back to the house. And casinos don’t mind that. “If you’re not perfect at card counting, you can still lose money,” Jonas said. “They’ll watch you count cards and if you make any mistake they’ll just let you play.”

They might even let a good counter continue to play if he or she is part of the entourage of a high roller who’s losing millions. If the counter makes $20,000 off the casino while his buddy loses millions, for the casino it’s just part of the cost of doing business.
Card counting teams from MIT made millions off Vegas and have since been immortalized in a book and film. The MIT team’s innovation was to have separate roles for players: Some were there to count and place small bets, others acted as high rollers and made big bets when the decks were favorable.

It’s easy for Vegas to spot a counter working alone based on betting patterns — lots of small bets followed by a few big ones. The MIT team survived longer than most because the separation of roles gave the appearance of legitimate betting patterns.

There are many other methods players use to shift the odds in their favor, with varying levels of legality. Jonas discussed a few, in some cases showing photos and video from his own collection. Here’s a sampling:

 

Exit mobile version