DRAFT / ABSTRACT

A major university recently completed a study of the on-line poker site FullTiltPoker.Net, the “free” side of Full Tilt Poker.

Over the course of the one-year study students entered into, but did not play in, over 10,000 9-person Sit & Go No Limit Hold’em 250 (play money) tournaments. Player stats were recorded, and later analyzed. Over 57,000 individual hands made up the study sample.

The study revealed that the dealing of cards at Full Tilt Poker is not random. The site has a tendency to deal a statistically abnormal amount of “big” hands, including straights, flushes, full houses, four of a kind, etc.

The odds of “hitting” a particular hand when seven cards are dealt from a 52-card deck are easily calculated. We won’t go into the gory mathematical details, but you are certainly invited to look them up for yourself if you don’t believe us.

For example, when seven cards are dealt from a 52-card deck, the odds of hitting a royal flush are approximately 500,000 to 1. That’s why you don’t see very many of them in the real world, let alone get one yourself. Yet, on Full Tilt we saw a royal flush in approximately 1 out of every 4,432 hands dealt, over 1,000 times as often as statistical probability would dictate. Several of our players were dealt more than one royal flush over the course of the study.

Oddly, a non-royal straight flush came up in 1 out of 7,203 hands during our study. With a random deal, the probability of hitting any kind of non-royal straight flush is approximately 83,000 to 1. That probability is multiplied by a factor of 10 at Full Tilt.

The most common winning hand in our tournaments at Full Tilt was 2 pairs. It came up on average 1 out of every 13 hands, about twice as often as statistical probability would dictate. With a random seven-card deal from a 52-card deck, a 2-pair hand should come up every 21 hands on average.

Likewise, all the other “big hands” occurred far more often on Full Tilt than one would expect from a truly random (and honest) deal. A straight came up in 1 out of every 54 hands dealt. Normal odds for a straight are 1 in every 283 seven-card hands. Likewise, a flush came up in 1 out of every 54 hands, the same as straights. In the real world, we would expect to see a flush in approximately 1 out of every 500 hands, or about half as often as a straight. On Full Tilt straights and flushes come up with almost equal frequency, yet another statistical anomaly.

Interestingly, the most common straights on Full Tilt are the ace high straight and the five high straight. They came up more than four times as often as any other kind of straight. This, in and of itself, represents another glaring statistical anomaly.

Hands with three-of-a-kind were observed in 1 out of every 27 hands, almost twice the 1 out of 47 hands one expects to see in the real world.
A full house came up on average 1 in every 71 hands dealt. This far exceeds the 1 in every 694 hands you will see in the real world.

Hands with four-of-a-kind occurred on Full Tilt in one 1 of every 281 hands, far more often then the 1 out of every 4,166 hands one expects from an random deal.

We can only conclude that Full Tilt does not deal cards randomly, but is programmed to produce big hands. Perhaps this makes for more “exciting” play, but it is less than honest.