Card Clumping

I stumbled across the following article recently (excerpt below). Would any of you care to comment please ?

"Blackjack: A Winner�s Handbook discusses the problem of card clumping as it relates to short term play. Simply put, clumping wipes out the card counters edge in many games by inverting the information generated by the count. Instead of an advantage, as the count may indicate, the counter is playing to a disadvantage as he or she bets up into a low-card clump and draws the unfavorable low cards.

Not only that, the counter is committing the very sin for which so many card counting authors criticize the player who uses a non-count system: just like many gamblers, he is betting up on successive losses.

There is another problem, however, which all but nullifies any potential theoretical advantage the short-term counter expects to realize: bankroll swings versus small advantage.

Consider the short-term card counter in his or her attempt to grind out profits using a card counting system. First, the counter must contend with boredom because betting with the count is a waiting game.

It has been shown in many card counting books that high bets make up fewer than 5 to 10% of the hands. So when those high bet opportunities present themselves, the adrenalin starts flowing as the big bet is pushed out. On a loss, another big bet as the count increases into a low card clump, and more adrenalin.

If the counter is using a betting spread of 12 to 1, which many traditional blackjack books recommend for the multi-deck shoe game, it is quite easy to lose 40 units or more in a clumped shoe. This ratio between high bet and low bet is necessary to overcome the house edge and give the counter his 1% theoretical advantage.

But these swings in the counter�s bankroll are devastating and in some instances catastrophic to a player with the standard bankroll of 200 units. Many blackjack card counting authors recommend bankrolls of 200 units, and, in my opinion, even a 200-unit roll is not safe in a heavily clumped shoe game. It is interesting to note, however, that books published late in the 20th century, have upped this units per bankroll number considerably � Vancura in Knock Out Blackjack to 1000 and Ian Andersen is Burning The Tables in Las Vegas to as high as 2000.

Kenny Uston writes about downswings in Million Dollar Blackjack:

� . . . if you were to play and enjoy a 2% advantage on every hand (which is unrealistically optimistic), after 2,500 hands you would have a 20% chance of losing. This statistic is continually borne out by our actual playing experiences . . .�

He then goes on to describe a 22-day period �in some of the most favorable games I�ve ever experienced. The interval included five days of playing only positive four-deck shoes at the Fremont (team play), five days of juicy single-deck game at the Dunes and six days of playing only positive shoes at the Desert Inn.�

At the end of this period, betting optimally to his large bankroll and with the true count, Kenny was down $35,000.

Now you have an understanding of why I stressed two bankroll divisor numbers in Chapter 3 of Blackjack: A Winner�s Handbook; today�s player, playing to the short-term, with winning strategies in this book, can start with a bankroll 1/5th or even 1/10th the size of the traditional card counters� bankroll.

Here, then, are my recommendations for the traditional card counter in confronting the realities of today�s multi-deck shoe games:

1. Do not bet up with the count on successive losses as most other blackjack books recommend; set a rigid table stop-loss amount and adhere to it count or no count. Table departure is the best decision you can make under these losing conditions because it is quite possible that this same low-card clump may come back to whack you in the next shoe." Unquote.

That last sentence seems to endorse what I saw recently :

Hot dealer wiping out players (on very high counts) shoe after shoe (6D) from 1pm - 5pm. I gave up after a few shoes as it was getting too expensive. Had a break and at night went to see if that table had improved. Back counted, didn't play. Incredible .....very high counts so often - and right to the end of shoes too - and it was still a massacre for players.

You can lose in a high count situation because of the specific cards dealt and the order in which they are dealt. A card counting system can identify situations in which there is an excess of player-favorable cards, i.e. aces, tens and nines, remaining in the shoe. It cannot tell the location of such cards or which specific cards will be dealt to the counter. Shuffle-tracking can provide the user with information as to the location of concentrations of such cards. Shuffle-tracking is, however, an advanced and difficult card counting method.

Jerry Patterson, who wrote "Blackjack: A Winner�s Handbook," did not teach shuffle-tracking. He sold a system called TARGET. It is not a card counting system. I think it purported to identify player-biased and dealer-biased tables caused by card clumping resulting from non-random shuffles. See Arnold Snyder's review of TARGET. To the best of my knowledge, no independent test of TARGET proved it to be effective.

Be extremely wary of any gambling system that is purported to always win, be beyond the mathematics of the game, or have a much higher win rate than proven systems, e.g. conventional card counting. Do not purchase or use any such system unless the claims made about it have been independently verified.

Nobody denies that there are bad streaks but Patterson makes an unwarranted link between these fluctuations and low card clumps. I say that a downswing doesn't need a single low card to occur.
Imagine a game with only tens and aces. Even in this highly favorable game the player will suffer severe setbacks when the dealer gets blackjack after blackjack.

Francis Salmon

Do you believe the Vancura and Andersen advice and the Uston anecdote were referring to the effects of card clumping?

And when reading Patterson, you will do well to skip anything having to do with TARGET.

Random distributions of cards does not mean evenly distributed. Many players make a good living off of clumping. Players get massacred in high counts because they have big bets on the felt. You can't get crushed when you're making small bets. Part of randomness is that sometimes the dealer gets better cards than you do regardless of the count.

Neither traditional card counting methods nor TARGET can advise the player as to the location of a clump in the shoe, the size of the clump, whether such clump favors the player or the dealer or whether the player will actually win when cards from this clump are being dealt. Shuffle tracking, which is an advanced and difficult card counting method, allows the player to follow where the cards after the cut card in the prior shoe are and to what extent they are diluted by other cards shuffled within them. Don't confuse shuffle tracking with TARGET, BIAS or other systems which have either not been independently proven to be effective, or, which have been proven to be ineffective.

I checked out "TARGET" when it first came out.

When I ran it against a computer simulated "stutter shuffle," which was the one in primary use in A.C. at the time, it was the worst thing I'd ever seen -- FAR worse than simple flat betting.

In a nutshell, you get plus points for getting a blackjack (which is nonsense because those use up good cards), plus points for winning a hand and extra plus points for winning a hand you "should" have lost. You get minus points when the dealer gets a blackjack and for losing a hand, and extra minus points for losing a hand you "should" have won. There are also plus and minus points for various dealer up-cards.

As others have pointed out here, "card clumping" DOES exist -- if it didn't exist, that would be proof the shuffle was other than random, not the other way around as Patterson alleges.

To boil the whole thing down, "TARGET" says follow the age-old gambler's advice: When you're winning, bet more; when you're losing bet less.

...I ran. I forget whether that was part of "TARGET" or not.

Still, all the same voodoo.

Thanks all for your comments and attached articles giving much insight into the subject.

Regarding Patterson's "strategy", at no time was I convinced enough to follow up on what he's trying to sell. I saw through the sale pitch, and being a believer in maths-proven card counting I thought his system lacked substance.

But being a novice who have seen in just a short span of time, some of the most monstrous shoes through high counts, I thought he was describing well the predicament new card counters find themselves in.

Incidentally, on that night when I counted & didn't play, one guy lost 2K just flat-betting $25 (through successive high count shoes). He was slaughtered, and not because he was betting big.

The character of those high-count shoes was such that it was just too hard to beat. So does Patterson have a point in saying, leave the table ?

If the casino you saw this happen in was small or Native American, be careful. Maybe the shoe had some of its faces and aces pulled out and replaced with 4,5 and 6 cards. However, if the shoe has all the correct cards, there is never a reason to run away during high count situations even if Patterson says to do so.