Playing dice with lottery tickets
Lotteries date back to the Han Dynasty in China in around 200BCE and were popularised in Roman times. Since then, the method of generating random winners has changed very little. Lottery random number generation systems involve complex entropy systems, such as ball machines and roulette wheels. These machines are not truly random, but rely on a system very sensitive to the initial conditions. This sensitivity makes the outcome very hard to predict, but in theory it could be determined if all conditions of the system were known. More importantly, these physical classical systems can be easily influenced by external factors, such as magnets, or calculated using a computer. Any tampering may be difficult or impossible to detect, enabling the control, or predictability of a game.
Serialisation is an important tool to track tickets or betting chips, but if the serial numbers are not truly random, they could be exploited. Serial numbers on scratchcards are used for distribution and security purposes, to track them. However, they have been used as an identifying marker for winning tickets. Numbers on scratchcards in Canada gave away winning tickets, allowing a consultant statistician to predict correctly the outcome of 19 out of 20 cards!
This predictability of systems that aren’t truly random can and have been exploited. One woman is reported to have won multiple millions of dollars from scratchcards in the USA four times! A former maths Professor from Stanford, it has not been confirmed how she managed to do it, but three of her wins were all bought from the same location. It is thought she has determined the pseudo-random algorithm that distributes where the winning tickets are located in each run. With no true random number generation in the distribution of tickets, it was possible to predict a winning scratchcard from its position in the run of tickets.
In 1999, a software company discovered several exploits in an online poker game to determine which hands were being dealt every time. The pseudo-random number generator determining the shuffles had a flaw, using the system clock as the seed. This left 86,400,000 possible shuffles for the deck. They were then able to synchronise their clock with that of the poker game, allowing them to predict which shuffle had occurred, based on the time and the cards they had been dealt.
Quantum Base has developed a simple random number generator. The numbers it produces will help to prevent the problems highlighted here. Find out more on this page.
Do you want to know how to use physics to help you win at roulette? More here.