Does Card-Clumping Exist?

Part I

by Ron Fitch --- a.k.a. Boris

(originally published - Summer 1995)

If you have access to the Internet newsgroups (specifically rec.gambling.blackjack) you may well have been following the saga (some call it a "flame" war) over the question of whether or not card-clumping (induced by conservative card play and maintained by non-random shuffle procedures) is a fact or merely a statistical aberration. Not surprisingly, I am at the center of this latest newsgroup controversy. In this and subsequent issues of the Journal, we will take a close look at this question.

Many people have asked me why this issue is important in the first place. Let me see if I can sum it up briefly. Let's assume for a moment that most of today's shoe games are indeed not random. To ignore this fact could then be quite costly. You can practice your conventional strategy (based on the then myth of random cards), only to find that when you get into the casino something is seriously wrong.

Let me explain this another way. Imagine that you want to become the best free-throw shooter in the history of basketball. So, for years, you practice shooting hoops until you can do it with your eyes closed. The only problem is that you are practicing against an 8 foot high basket. In other words, you are practicing in an environment that does not exist in the real world. No amount or practice using an 8 foot hoop will prepare you for the 10 foot hoop that exists in the real world. Putting it bluntly, you are wasting your time.

As you may know, I take the position that cards in today's multi-deck shoe game (not to mention many single and double deck games) are the furtherest thing from random. If I was not firmly convinced of this, there would be no reason to discuss this issue in the first place. However, I don't intend to just leave it at that. Unless you understand how and why the cards are not random, this fact becomes useless. So, hold onto your hat while we take a journey INSIDE the game of Blackjack to find out what is REALLY happening and what you can do about it.

A History Lesson

It has been said that those who do not learn from their mistakes are doomed to repeat them. This is why history study is so important in life. It has also been said that history is written FOR the people it is about; in other words, to glorify the people/events being examined. While this can make for interesting fiction, it provides little in the way of useful information.

Most Blackjack authors today present a warped view of the game. They fail to recognize that the game has indeed changed during the last 15 years, promulgating strategies that may have worked years ago, but are of dubious value in todays game. While it is manifestly proveable that the cards in Blackjack have never TRULY been random, the decade of the 80's relegated the random game to the museum of ancient history. It's useful to briefly understand when and why this happened.

Less than a year after Resorts International opened its doors in Atlantic City, the question of whether card counters can be barred became an issue. In his books One Third of a Shoe and Million Dollar Blackjack, Ken Uston detailed the "counter convention" which began in late January of 1979 from which he and his teammates were eventually barred. This resulted in a lawsuit against Resort's which Uston initially lost. He took it to the appellate court and won. Part of the Resort's weak defense was that they would either have to discontinue Blackjack or go out of business if counters are allowed to play. They appealed the decision to the state supreme court and lost. September 15, 1982 marked the turning point in the game of Blackjack; not just in Atlantic City (due to the change in NJ state law), but eventually, the world over.

During the period 1981 to 1983, computer consulting firms were hired to analyze the game to determine what could be done to "protect" the casino from card counting players. You may well rememember such nuances as the Bart Carter shuffle and the 12-deck dual-shoe Blackjack II game, which was eventually discontinued. Since then, card shuffles and washes have been developed which all but obliterate any edge a conventional card counter may have. While most people dispute me on this, let me remind you that the majority of card counters (as widely reported in the Blackjack newsletters and journals, as well as the Internet newsgroups) will tell you that the Atlantic City games can't be beat [using conventional card counting techniques].

Before we examine evidence that points to the existence of card-clumping, let me point out a unique aspect of the Atlantic City game. Atlantic City, being on the ocean is unique amongst U.S. casino cities due to the salt in the air. This causes the cards to stick together somewhat; increasing as the hours wear on using the same cards. Couple this with summer humidity and the clumping is increased further.

During my first visit to A.C. in April 1979, I remember the dealers complaining about the cards sticking together. It took several years for the powers-that-be to realize the casino advantage to this phenomenon. Regular A/C players will tell you that game conditions in the winter are usually better (thanks to a lack of this condition), in addition to the tables being less crowded.

Conditions of a Random Game

It is important to understand that most (if not all) Blackjack strategies were devloped using (or derived from strategies assuming) randomly shuffled cards. Table #1 (below) compares situations that should occur in a [theoretical] random game versus today's typical clumped shoe games.

As you can see from the above statistics actual games found in the casino are CONSIDERABLY different from what is taught in most [publically available] Blackjack books. Likewise, most computer software products available for studying the game utilize random shuffle algorithms. Except for the Boris Blackjack simulator, software products claiming to simulate casino shuffles are grossly inaccurate. Practicing against poor software simulations can lull you into thinking you are a better player than you actually are. This can be quite costly.

Even with all the convincing evidence above, some people still feel that card-clumping is statistical aberration, a myth. Let me offer one more piece of information to convince you remaining skeptics. Distribution of ten-value cards is an excellent indicator of card randomness. In randomly shuffled cards, there should be one Ten for every 3.25 cards dealt. The odds of seeing 6 or more tens in a row in a game are quite small, yet, we see them every day; even more frequently in Atlantic City.

Table #2 (next page) details the odds of finding sequential tens in a randomly shuffled game. If the cards were indeed random we would RARELY see solid ten-clumps. And yet, we see them nearly every day; often every shoe. Tens are dealer-breaking cards. When tens are overly-clumped, two things happen:

1. They often get cut out of play, ending up behind the shuffle card;
   

2. They end up clumped together creating two-ten-pushes with the dealer;
   in which case they may as well have not been there at all.

20+ years ago, some unscrupulous casino operators used to have tens removed from the shoe (a procedure known as short-decking) in order to insure a dealer bias. In today's game, short-decking is unnecessary. Insufficient shuffling combined with poor playing of the hands guarantees the house a far greater edge over the players than ever imagined. Card counters are no longer a threat in today's shoe game. Card-clumping significantly reduces card-counting opportunities. True-counts quickly become false counts.

Now you know why I don't count cards in the shoe games anymore; and, virtually all I play are the shoe games. In fact, some of my favorite games are the 6-Deck Shuffle Master shuffle machines. Although Shuffle Masters allegedly produce random cards, this couldn't be further from the truth. Issue #3 of this Journal will introduce the Shuffle Master machines to give you even MORE things to think about.

Let me offer you one last point about this. It is my contention that if the cards were PERFECTLY random, even card-counting would not work - there would be nothing to count. Has it ever occurred to you that from the beginning, card-counting has relied on a degree of non-randomness? Think about it. Non-random shuffles and card-clumping are a reality folks. If you wish to believe the games are random, that is your choice. However, it could be an expensive one.

Table #2 - Odds of Sequential Tens in a Random-shuffled Game

Sequential 10's	Every N Cards	Every N Shoes (6-Deck)	Every N Hours (6-Deck)	Every N Days (6-Deck)	Every N Shoes (8-Deck)	Every N Hours (8-Deck)	Every N Days (8-Deck)
1	3.25	-----------	-----------	-----------	-----------	-----------	-----------
2	10.64	-----------	-----------	-----------	-----------	-----------	-----------
3	35	-----------	-----------	-----------	-----------	-----------	-----------
4	117	-----------	-----------	-----------	-----------	-----------	-----------
5	391	1.25	-----------	-----------	-----------	-----------	-----------
6	1319	4.2	-----------	-----------	3.2	-----------	-----------
7	4480	14	2.4	-----------	11	1.3	-----------
8	15327	49	8	-----------	37	4.6	-----------
9	52832	169	28	1.2	127	16	-----------
10	183431	588	98	4	441	55	2.3
11	641643	2057	343	14	1542	193	8
12	2260510	7245	1208	50	5433	679	28
13	8022548	25713	4286	178	19285	2411	100
14	28688633	91951	15325	638	68963	8620	359
15	103340000	331218	55203	2300	248413	31051	1293
16	375010000	1201955	200325	8347	901466	112683	4695

Note: This article was excerpted from Boris' Advanced Blackjack Journal
Volume I, No. 2 and is reprinted with permission of the Human Insights Group

Copyright 1995 - Human Insights Group - All Rights Reserved.

Back for more Articles          Back to the Home Page