Ask the Wizard: Blackjack - Shuffling
What is your opinion of the continuous shuffle machines now being used at the blackjack tables in Las Vegas? Do these machines give the house more of and edge even when a person is using basic strategy?
— John from London, Canada
For those who don't understand what you're asking, there are new machines that take the blackjack discards and place them randomly back in the deck after each hand. If you are using basic strategy, then the shufflers actually lower the house edge slightly, due to the omission of the cut card effect. It is my understanding that they do provide an honest random shuffle. However, the shuffling machine allows the dealer to waste less time shuffling and spend more time dealing. This means you will spend more time playing, and thus more hands for the house edge to grind you down.
For more information on the mathematical effect of continuous shufflers, please see my blackjack appendix 10
When playing online blackjack, how do you tell when the deck is shuffled? I play Microgaming casinos (which you report as using 1 deck), but I do not know if each time I play if it is a new deck, and there are no signs of knowing when the deck is shuffled.
— Brian from State College, USA
Most online casinos shuffle after every hand. Others shuffle at random times but do not indicate exactly when to the player. I have noticed Microgaming casinos flash the word "shuffling" about one hand in four.
However, if you track the cards between these announcements you will sometimes see the same card twice, which is impossible in a single-deck game, assuming you believe them about when they shuffle. As far as I know, they actually shuffle after every hand, but for reasons I do not understand, only indicate a shuffle occasionally. If I remember correctly, Cryptologic casinos do indeed indicate when they are shuffling their eight-deck shoe.
What do you know about the randomization process that online casinos use to simulate shuffling? How closely does it approximate the actual manual shuffling of cards in a casino? And finally the obvious: wouldn't it be fairly simple to write a randomization (shuffling) program that would give the house a bigger edge -- sort of stack the deck? Enjoy your site. Thanks.
— Jim from Cincinnati, USA
I know that one software company randomly picks two cards in the deck and reverses them, and repeats this numerous times. Since learning of this technique, that is also how I shuffle in my random simulation programs. As long as any method of shuffling is done enough times the deck should be properly randomized.
Manual shuffling is more vulnerable to a biased shuffle and consequently some players try to exploit this by shuffle tracking and card clumping. There are numerous ways an online casino might cheat, but a bad shuffle I don't think is one of them.
I was wondering if you thought continuous shufflers have an effect on basic strategy? I know they speed up the number of hands per hour which is usually bad for the player, but is basic strategy still effective in this instance? Doesn't basic strategy slightly change depending on the number of decks?
— Danny from Mission Viejo, California
I first addressed this topic in my December 1, 2000, newsletter. For those who missed it I just added blackjack appendix 10
to my site, which explains the effect on the house edge under both a cut card and continuous shuffler game. To answer your question, no, the basic strategy does not change. Basic strategy is always developed based on a freshly shuffled shoe, which is always the case when playing against a continuous shuffler.
I’ve noticed that the CSM (Continuous Shuffler Machine) at the blackjack table does not shuffle ALL of the cards at the end of each hand. There are a few cards left in the shoe part of the machine (anywhere from 1 to 20 or so) that are not shuffled. Is there any way this can be used to advantage? For example, I was thinking that there is a lower (but still not zero) probability of having a card repeated two hands in a row. Sit out if there were a lot of high cards last hand . . . bet higher if there are a lot of low cards last hand. The CSM I saw used four decks so, on a full table, there are actually quite a few cards played each hand and you could potentially get a true value of plus/minus one if you made the simplifying assumption that none of those would repeat. Maybe enough to skew the odds?
— Chuck from New York
You’re right, the discards are not mixed among all the cards but can not be placed close to the top of the shoe. I don’t know the exact size of this buffer but it is about 10-20 cards I think. As a card counter it would probably be safe to use a true count from just the last hand played and off the top of a shoe. When converting to the true count you will rarely get anything far from +/-1. If you’re any kind of counter at all I would forget about playing against a CSM, it isn’t worth the bother.
On a CSM blackjack game, 5 deck, what would be the effect of the dealer not putting discards back into the machine every hand if 24 of 52 cards in the discard rack were face cards? What about 48 of 102? What would be the effect if 44 of 52 cards were non-face cards? Can the odds be heavily skewed? I have a feeling that the odds skyrocket in the casino’s favor if the dealer leaves face cards in the discard.
— Doug from Vancouver, Canada
The exact numbers would be difficult to calculate and I won’t get into that. However your speculation is right that the odds favor the dealer if he leaves a lot of high cards in the discard rack yet will put back into play a lot of small cards. This would be the same kind of thing as preferential shuffling, in which the dealer of a hand held game shuffles when the count is good but deals another round on a bad count. Preferential shuffling is something that definitely does happen here in Las Vegas so what you describe would not surprise me either.
I was just wondering if Las Vegas Video Blackjack reshuffles after every hand or after all the decks are played. I know the tables in Las Vegas do it after all the decks are played because if they did it after every hand no one would play. Are the odds that you’ll win worse if there is a shuffle after every hand? Is this even legal or necessary?
— Stephanie from Evans, USA
I don’t know when they shuffle but I would speculate after every hand. From my blackjack appendix 10
you will learn that the player’s odds improve slightly if the dealer plays exactly n hands between shuffles (including one) rather than playing to cut card, finishing the hand, and then shuffling.
First, thank you for the great site. I went to Las Vegas for this first time this past summer and I played double-deck blackjack at the Orleans. I noticed that after a dealer shuffled both decks, the dealer asked the player to cut the deck. Most players refused. I did not mind, so I cut the deck. Is there a blackjack cut superstition that I am not aware of, or is there a better reason why?
— Sam from Price
I would say about 1/3 to 1/2 of players would at least initially decline to cut. However if everyone initially declines somebody has to rise to the occasion and do it. Sometimes when players who refuse to cut will say something like "I don’t want the blame for a bad shoe" or "I’m unlucky." I’ve never seen it put into words but there does seem to be a superstition that the cut is critical to the flow of the shoe, and thus the act should only be done by a competent cutter. Of course this is nonsense. For recreational play it doesn’t make any difference whom cuts or where they cut.
Playing blackjack on a continuous shuffling 5-deck system, are the odds of winning different than playing the dealer with 1 deck or 2 decks?
— Tom from Aurora, CO
For the beneit of other readers, my blackjack appendix 10
explains, the house edge in a five-deck game is 0.028% less if a continuous shuffler is used, as opposed to a hand shuffle. The difference between five decks and two decks, all other rules being equal, is 0.18%. So the two-deck game without a shuffler would be much better. Let’s compare a 5-deck continuous shuffler game to a 4-deck hand shuffled game. As my blackjack calculator
show difference in house edge between four decks and five decks is 0.0329%. So the benefit of a continuous shuffler is worth less than removing a single deck.
While playing blackjack at a locals casino in Las Vegas, a dealer from another locals casino sat at my table. While making small talk, she told me that she could wipe out any player using what she called the "house shuffle." The lady dealing to us, who claimed to have been a dealer for 25 years agreed with her telling me that it’s "all about the shuffle." They were both referring to games dealt by hand as opposed to from a shoe. Is there a way to shuffle that lowers the players chances of winning, and if so wouldn’t this be a form of cheating? Have you ever heard of anything called the house shuffle?
— Jim Y. from Downey, CA
I don’t believe it. Dealers are not the most skeptical group, often believing all the usual gambling myths. Usually the term "house shuffle" refers to the way the dealers are supposed to shuffle. For example, shuffle twice, riffle, and shuffle again. In this context, she seems to be saying she could alter the shuffle to the player’s disadvantage, which I doubt.
I absolutely love your site. I enjoy the strategies and probability discussions as much as, or more than, the actual gambling! I was playing six-deck Blackjack in a St. Louis casino recently. After playing a shoe, the cards were returned to the auto shuffler, which indicated a card was missing. The dealer proceeded to deal the next shoe while the floor person inspected the returned set of cards. Upon completion of this shoe, the missing card from the previous shoe (a king) was found in the un-dealt portion of the second shoe.
Assuming this King was the bottom card and was left in the shuffler, it would have been in play in this first shoe (the cut was in rear portion of the deck). How much of an additional advantage did the house gain on me with this mistake?
— Justin from St. Louis, MO
Thank you for the kind words. I’m going to assume the dealer hits a soft 17, and double after a split is allowed. According to table D17 in Blackjack Attack
by Don Schlesinger, removing one ten per deck increases the house edge by 0.5512%. Dividing that by six, for the six-deck game, the effect is an increase in house edge of 0.09%.
For recreational blackjack players, who use basic strategy, and don’t count, does the house advantage increase as the penetration increases? I believe it does because the deeper you get into the shoe, the greater the absolute value of the count will tend to get, which should trigger count-based strategy changes. Since the non-counter wouldn’t know when and how to make such changes, he would be making more mistakes as the count gets further away from zero. Thus, wouldn’t a non-counter be better off at a table with shallow penetration?
— Jon from Doylestown, PA
In a non-cut-card game, the house advantage is always the same for the non-counter. Clumps of high or low cards are just as likely to appear at the beginning of the shoe, as the middle, as the end. Just because the count is zero at the top of the shoe doesn’t mean you’ll have an exact balance of high and low cards. You seem to be suggesting that the cards are more clumpy at the end of the deck. However, if that were true, then the odds would change if the dealer dealt the cards in reverse order. Surely that is a ridiculous notion.
Let’s say the basic strategy player has 16 against a 10 late in the shoe, and hits. If the count were high, standing would be the right play, resulting in what would look like an error to a counter who was watching. However, if the count were negative then hitting would be all the better. In the end, it averages out, for the basic strategy player.
For reasons I explain in my blackjack appendix 10, the basic strategy player should prefer a game with a continuous shuffler, if his goal is to minimize the house edge. Aside from that, the house edge is not affected by penetration. I should add that with a shallower penetration there will be more time spent shuffling, and thus a lower expected loss on an hourly basis.