Hacking the Casinos for $1million

[shadowman]

That IS the way RNGs work. They are really called pseudo-RNGs because of this nature. The other factor that is important is the RNGs keep running constantly. Hence it would be difficult to predict even if you knew the algorithm. Add in today's processor speeds and you might see a million numbers pass by each second.

[jm002546]

  So how did they pattern the RNG and systematically win $1,000,000?

[cddenver]

That IS the way RNGs work. They are really called pseudo-RNGs because of this nature. The other factor that is important is the RNGs keep running constantly. Hence it would be difficult to predict even if you knew the algorithm. Add in today's processor speeds and you might see a million numbers pass by each second.
 
Nowadays the RNG's are required by state law to be operating continuously even when a machine is not being played.  That particular requirement has been part of our gaming regs here in CO since I first began checking the regs about seven years ago, and I don't know whether that provision was part of the original gaming act or added in later.  I just read the referenced article and liked the part about where the hackers would try to keep players away from a given machine while they figured out when the next RF was going to come up based on the last hand played since the RNG wasn't running when the machine wasn't being played......wouldn't work nowadays.
 
The pseudo-RNG stuff is hard for people to understand.  There isn't any way to just tell a processor to "pick a random number".  The use of code of any kind means structure and regularity - unrandomness!

[shadowman]

  So how did they pattern the RNG and systematically win $1,000,000?
 
They knew the speed of the processor and the algorithm after taking the machine apart. They would see a particular sequence of cards which would tell them exactly where they were in the RNG sequence. Therefore they could predict the amount of time until a RF hand was available. They would sit there and time it and try to hit the deal button at that exact moment in time. Since the machines were much slower back then they could be successful quite often.
 
I think they even commented that they were surprised how often they were successful.

[faygo]

I read further that they were using a remote PC linked to what the player was seeing and the remote PC was was actually predicting the next hand or drawn cards.  Again, as mentioned only possible because the processor sat idle after the initial 10 card draw.

[jm002546]

  I recall the timing part re royals, but then they shifted focus to just take smaller wins by knowing what to draw-hold. How did that work?

[shadowman]

First of all, the machine did not sit idle, it just didn't change the 10 cards. It did keep cycling through the numbers. This is why they needed to spend extra time on their "count down timer". They knew where they were in the RNG and the exact time. From this they could compute the exact time a RF would appear. If I remember right they hit it about 1/3 of the time.
Second, since they can figure out where they are in the RNG they can know all ten cards that were selected when they hit deal. Therefore, they could always hold whatever cards would lead to the best result. For example they may be dealt 3 kings, an ace and a four. However, if they know 3 aces are the next cards coming up they can toss the 3 kings that everyone else would hold. Low and behold they have 4 aces with a kicker.