Jerry Patterson’s Break the Dealer: Bad Information on Shuffles
by Arnold Snyder
(From Blackjack Forum VII #1, March 1987)
© 1987 Blackjack Forum
All of the “non-random” shuffle exploitation gurus including Jerry Patterson, Eddie Olsen and Doug Grant claim that Patterson’s systems won’t work vs. computer-random shuffles. Yet, none of you have ever published any computer simulation data on your systems in which the computer uses a non-random human-style shuffle.
There is nothing magical nor mystical nor impossible about having a computer shuffle like a human dealer. Instead of having a random number generator put the cards in order, one would simply program the computer to shuffle the cards like a human being would. You have it break the deck(s) into the approximate sized portions that a dealer would, mixing the cards by riffling, stripping and boxing the deck as is done in the casino. The computer can do this as perfectly or as sloppily as any human dealer would. Computers do what you program them to do.
The only computer studies I’ve seen on the effects of poor shuffling on the game of blackjack were the studies published by Stanford Wong in his Blackjack World newsletter in August, 1983. In Wong’s study, he decided to use the poorest shuffling of all, i.e., no shuffle.
The cards were, of course, being slightly shuffled, because as hands were played by the computer, Wong had his computer pick the cards up and place them into the discard rack in the same order that a human dealer would do this. At the end of each shoe, he had his computer just start playing again with these discards. But this standard casino discard ordering was the only “shuffling” that Wong used.
He was using this computer simulation to test the theory that a multi-deck shoe would tend to stay “hot” or “cold” from shoe to shoe if the shuffle were inadequate. His finding was that there was no basis for this belief. Even with no shuffle, the mixing up of the cards caused by placing them into the discard rack as a dealer would was a sufficient shuffle to make the results of the following shoe unpredictable.
This test of Wong’s did not, obviously, test all of the various theories about the possibilities of exploiting a non-random shuffle. Wong did not test the length or frequency of winning or losing streaks or compare such data to random results. He did not input any specific streak-based system to see if it would work. Wong tested just one theory and found it to be invalid.
In your response to Wong’s review of the Bias Barometer, Jerry (elsewhere in this issue), you state that “a few of my students have performed their own independent research activities and written nonrandom shuffle programs to validate B.l.A.S.” So my telling you that a non-random human shuffle is reproducable via computer is nothing new to you. You claim to have access to this type of computer data, yet you have never published any of it.
Jerry Patterson’s Break the Dealer: Even Worse on Playing Ace-Rich Decks
And what you have published, I disagree with. Although you describe some of the elementary methods of shuffle-tracking, which could be of great value to many shoe players, I disagree with your recommendations for how a player should use this data.
To even suggest that any shuffle-tracking player should attempt to create a shortage of aces, or bet low and play his hand as if in a negative count when in fact he is in the middle of a high card clump, strikes me as ignorant, regardless of the fact that the shuffle was non-random. This is the exact opposite strategy that was used by the highly effective concealable “Thor” computer, developed by Keith Taft many years ago for the express purpose of shuffle-tracking.
Keith performed many computer simulation tests of his Thor computer, using “casino-style” non-random shuffles, to develop his strategies. So unless you can produce some data which indicates Keith Taft was wrong—and Taft’s methods are solidly backed by what is currently known about the mathematics of blackjack—my opinion is that you are very badly mistaken about how to get an advantage by shuffle-tracking.
Jerry Patterson’s Break the Dealer And “Game Control Techniques”
As for my “missing the major point” of your book—that casinos are using “game control techniques” to keep players from winning—I don’t think any publication has presented more information on this subject over the years than Blackjack Forum. From warnings and descriptions of preferential shuffles, Sam Case’s and Steve Forte’s exposes of card manipulation, etc., I publish everything of value on this subject that I have access to.
I simply fail to perceive what you describe as “game control techniques” to be what you and Eddie Olsen claim them to be. For instance, you warn players against the “strip” shuffle, especially in single-deck games. You accurately describe the stripping action which may be done by single cards or in clumps of cards, and you note that stripping the deck simply reverses the order of the cards or clumps of cards.
But you write about this common shuffling technique as if it is detrimental to players. It has long been a myth among gamblers that changing the “order of the cards” will change the players’ luck. But I’ve never seen any scientific data to support this notion, and frankly, I don’t believe it. The only players I know who were ever hurt by card stripping were players who were using concealed computers to track the shuffle, and whose computers were not programmed to be able to reverse the order of clumps of cards.
But stripping cards doesn’t hurt human shuffle-trackers, who can easily reverse clump values. Nor does it hurt card counters, according to any accepted mathematical theory.
If you make a claim, you should back it up. Did you computer-simulate stripping a deck to test its effect? If so, why don’t you produce the data? Why not tell us about it? Your whole book is filled with claims that strike me as superstitious nonsense. Let’s see some facts.
And don’t give me that “Blackjack-is-a-short-term-game” nonsense, as you say in your responses to both Wong and myself. Selling players methods for winning in the short run is a cop out. This is what phony system sellers have been doing for years with their betting progression systems, their Martingales, cancellations, stop-losses, double-ups, cut-backs, etc.
It’s easy to create a betting system that will win most of the time in the short run. The simple Martingale (double after a loss – bet one unit after a win) will win 90% of short-run craps sessions if a player has $511 bucks in his pocket, and spreads his bets from $1 to $256 on the pass line. It would take 9 consecutive losses to wipe this player out. It only happens occasionally.
Unfortunately, a loss on the $256 bet, which will put the player $511 behind on his “series,” will occur slightly more frequently in the long run than this player’s expectation of winning 511 series in a row. This player is going to lose 1.41% of all the money he bets in the long run, because that is the house’s advantage. It doesn’t matter that this player will win 90% of his short run sessions. He’s a long run loser.
If your system does not produce a long run player advantage, the house will beat you in the end. They’ve got the time and money to sit through your system. The people who get most caught up in these foolhardy short run systems are compulsive gamblers. They’re always talking about how they usually win. And they’re always behind because of their few big losses.
As for your getting a 100%+ advantage over the casino, which you attribute to exploiting biases caused by non-random shuffles, which you support by claiming to have personally multiplied your bankroll in short run sessions, this is absurd. I often practice card counting on a little Radio Shack computer that deals a single deck out 75%, then “randomly” shuffles. I once multiplied a $100 bankroll to $31,000 in a few hours of play vs. the Radio Shack dealer via a mixture of incredible luck and foolhardy betting. (Unfortunately, I couldn’t get Radio Shack to pay up.) But this phenomenal “winning streak” had nothing to do with biases caused by a human shuffle. There are numerous stories of real life incredible wins like these at casino crap tables, roulette tables, etc.
Jerry, I’d love to find out there was some validity to your TARGET system, or your Bias Barometer, or any of the other streak-based blackjack systems being touted by you or Eddie Olsen or Doug Grant or Charles Einstein, etc. I’d be amazed to find out your theories were right since they seem to violate accepted mathematical theories. But I wouldn’t mind eating humble pie on this if there were something to it, as much as I have enjoyed making fun of you over the years.
So, will you please send me some non-random computer simulation data that might validate your claims? If you have this data, why don’t you publish it? ♠
Never miss another post