Sunday, April 01, 2007

Atlantic Lotto Ripoff

Check out the Atlantic Lotto ripoff story. It turns out retailers and other insiders are cashing in large prizes at a statistically impossible rate.

My wife's theory, which the Chronicle Herald story also implied, is that dishonest retailers are telling unsophisticated customers that their winning tickets are losers. The retailer tosses the "losing" ticket into the trash and the mark leaves the store empty handed. The retailer then gets the winning ticket out of the trash and cashes it as his own.

This is most likely true. I am pretty smart, and I find scratch and win tickets confusing, unintuitive, intimidating and difficult to read. It is not at all easy at a glance to tell if a ticket is a winner. 6/49 is only slightly better. If someone has less than high school education (which sadly is a large portion of lotto players), then that player may not be able to tell himself if his ticket is a winner, he has to rely on the retailer. That makes him an easy target for a dishonest retailer to cheat him out of his rare big prize.

Years ago when I was in university I was on a crew helping to clean up a residence on a temp job. One of the regulars from the cleaning company said when he was cleaning offices and it was not busy then he would sometimes recheck discarded lotto tickets. He said he often found winners, including a $50 win. That means people either can't or don't check carefully enough and may not be aware their ticket is a winner.

Make no mistake, retailers will attempt to cheat unsuspecting customers out of winning tickets. I don't play the lotto myself of course. Why would I put money into a game which has a negative expectation? That doesn't make any sense. Anyway sometimes I get tickets for my birthday or Christmas. One time a couple of years ago I had around 5 scratch and win tickets. 4 were losers and one was a $2 winner. I went to the kiosk across from the motor vehicles registry in West End Mall Halifax with all 5 tickets. I had checked the tickets carefully and knew there was 1 small winner and 4 losers.

The clerk laid the tickets out in front of him on the counter and quickly glanced at them. I was watching him. I could tell he noticed the winner among the tickets. He put the tickets together on the counter. Then one by one he put the tickets under the bar code machine. After each checked one he put it back into the original pile, mixing the checked and unchecked together. He checked the losing tickets but not the winning one.

After the losing tickets were checked he said to me sorry you didn't win. I told him he missed one. He tried to act stupid like he didn't know what I was talking about. I specifically pointed out the winning ticket he hadn't scanned. He reluctantly checked it and of course it was good. I got my $2 and left and didn't say anything more about it until now.

So whatever, he tried to cheat me out of $2 and failed. What if I didn't know I had a winner, and it was $200, or $25,000? Based on my direct first hand experience it surprises me not at all to read that retailers on average have an absurdly high win rate.

The thing about the retailers win rate, the true rate is actually even higher. Consider my own case above. Suppose that retailer cheats someone out of a $25,000 winning ticket. If I was that retailer, I would not try to cash that ticket myself. That could bring unwanted attention. I'd get an accomplice such as a spouse, relative or close friend to cash it and divide the proceeds. That way there's no obvious link between me and the big winning ticket. So the stated 10X win rate is just the times the retailers were stupid or sloppy and cashed out the tickets themselves. I expect they would also use accomplices to cash their scammed winning tickets to conceal their ill gotten gains and make it difficult to impossible for the lotto corporation to link back to them.


One angle that I was concerned about is the bar codes on scratch and win. In order to scan a ticket the bar code must be checked. Thus there must be a database somewhere linking the bar codes to the actual value of the ticket. It would seem that given the bar codes it would be possible to determine the value of a ticket without scratching it. I would expect that Atlantic Lotto corporation does not allow casual access to this database. Still for the data to go in or come out it has to be known so it may be possible for insiders to obtain this. This knowledge could be used to skim the tickets to pick out the rare big winners, leaving the general public with the crumbs.

It could work something like this. An insider at the lotto company somehow gets access to the database matching the scratch and win bar codes against the actual value of the tickets. An accomplice retailer gets his regular delivery of tickets. The retailer tells his bar code numbers to the lotto corp insider. The insider checks the database for big winners. Usually there's nothing interesting and the tickets are sold as usual. However say once every 2-3 years the retailer has a $10K+ ticket.

In that case the lotto company employee tips off the retailer about the ticket. The retailer himself or a third accomplice then makes a standard clean purchase of the marked ticket over the counter from the retailer, takes the lucky ticket home and of course hits the jackpot. The proceeds are split and the big ticket is never sold into the general public.


I love in the article how the ALC spokesman said there's no evidence of wrongdoing despite insiders winning at a statistically impossible rate. When they say that, the lotto company is either very stupid or lying through their teeth to the public. I'm not sure which is worse.