Abstract

I. Introduction
Blackjack holds a special place in the eye of the casino patron and the casino owner. It is easily the most popular table game in America. 1 Casinos take in about $1 billion from blackjack players every year in Nevada alone, 2 yet blackjack is a game that can be beaten. This presents a problem for casinos. Casinos are profitable because, in the long run, the games they offer cannot be beaten, yet skilled blackjack players who can count cards can beat blackjack. By carefully tracking each card that passes through the game and raising their bets when the count is in their favor, 3 card counters can shift the edge from two percent in favor of the house 4 to approximately one percent in their favor. 5 Advantage players, such as card counters, are a serious threat to casinos' profits, and casinos have developed countermeasures to reduce their effectiveness. The legality of certain countermeasures is not always clear.
In 1982, the Supreme Court of New Jersey held that no New Jersey casino could exclude a player from its blackjack tables just because the player was counting cards. 6 Excluding a player from the table was the ultimate countermeasure against card counters. Without this countermeasure available, New Jersey casinos were forced to turn to other methods, such as increasing the decks in a shoe, inserting multiple cut cards, or continuous shuffling machines. 7 The problem with these countermeasures is they are often met with patron resistance. 8 Casinos want to keep the house edge without signaling to players that they hold that edge. Casinos seek a countermeasure that can be stealthily utilized and only when the casino desires. Preferential shuffling appears to be the solution.
Preferential shuffling occurs at the intersection of two controversial features of many casinos: high‐tech surveillance devices and at‐will shuffling. Commonplace features of every modern casino are advanced surveillance systems that track every card that passes through the game. The house knows the count in real time, when a shoe is hot, and can instantly identify suspected card counters. The question left for the casino is what to do with this information. After all, card counting is legal. In states other than New Jersey, the casino may ask the player to leave, but that means the casino loses a customer. 9 The ideal solution in any state is to beat the card counter at the table, and that is where countermeasures such as at‐will shuffling come into play.
At‐will shuffling occurs when the dealer decides to shuffle mid‐game, as a deterrent against card counters or perhaps simply to enhance the randomization of the game. When at‐will shuffling is used in conjunction with the dealer's knowledge of the count, shuffling becomes more than a mere randomization event. By shuffling when, and only when, the count is favorable to the card counter—or any player—the casino alters the odds of the game in its favor. This is referred to as preferential shuffling.
The legality of preferential shuffling has been addressed by few courts, and those courts concluded that preferential shuffling is not cheating, but a permissible countermeasure. As this article will discuss, preferential shuffling is not simply a countermeasure. It alters the odds of the game, without notice, to the detriment of all players at the blackjack table, card counter or not. 10 Casinos can use preferential shuffling to stealthily change the odds of blackjack without the knowledge of patrons. Courts, including the Third Circuit, have failed to realize this.
Part II of this article discusses the highly advanced technology that allows casinos to count cards in real time. Part III discusses the history and purpose of shuffling at‐will. Part IV discusses the effects of preferential shuffling on the game and the legality of preferential shuffling, particularly whether it falls under state cheating statutes in Nevada and New Jersey. 11 This section also discusses problems with detecting shuffling at‐will, and possible future side effects.
II. Casino Surveillance
Since before blackjack was Nevada's most profitable game, casino owners were wary of card counters and used their security systems to keep a watchful eye. 12 To keep up with card counters, the casino must be able to count cards itself. Some of the first devices for casino gaming analysis enabled casinos to detect card counters. Programs like BJ Tracker allowed casino operators to track cards on the table and helped operators keep track of the count and identify possible card counters. 13
Then came “smart shoes” and “smart tables,” devices that track the cards and chips as they appear on the table in real time. An integrated surveillance system can track the cards, calculate the count at each blackjack table in real time, and then relay that information to casino employees. One casino surveillance company boasts: “In real‐time, the system can count cards and overlay text of the true count on the corresponding video. Alarms are established to notify operators when a shoe is ‘hot.’” 14 Casinos appear to value the ability not only to count cards, but to instantly notify operators when there is a shoe that is favorable to the player. With data coming to the operator in real time, operators have more than secondary data with which to analyze the actions players have already taken, they have data upon which they may act affirmatively to affect the game as it happens. Critics concluded that operators could easily manipulate the game with this information. 15
Casinos counting cards is nothing new, 16 but modern casino surveillance systems make acquiring this information easier and more affordable than ever. What casinos do with that information is a more closely guarded secret. At least one patron brought suit against a Nevada casino for allegedly using a MindPlay surveillance system to count cards and then preferentially shuffle. 17
No one can dispute that a casino has a significant interest in obtaining a large amount of detailed data concerning the games that are played at their tables. They want to guard the integrity of the game from players who might use technology of their own to cheat. 18 However, the ability to count cards in real time and instantly notify a pit boss or blackjack dealer when a deck favors the player may exceed the scope of protecting its interests. It creates a clear incentive to instruct the dealer to shuffle when the count favors the player.
Perhaps the most concerning aspect of these practices is how easy and unnoticeable card counting is for the casino. The average casino patron is likely unaware that the blackjack table or shoe holding the cards could be tracking every card that is dealt and alerting the pit boss when the patron is more likely to win the next hand. These practices, if known to the public, would certainly trigger backlash from the casino community. 19
III. Shuffling at‐will
At‐will shuffling refers to a casino's decision to shuffle the deck after any round of play. In its most basic sense, at‐will shuffling enhances the randomization of the game. After all, the shuffle is the event that randomizes the play of the game. 20 Shuffling at‐will is permitted in every state with blackjack, sometimes even expressly. 21 According to Fred Gushin, a former commissioner of the New Jersey Casino Control Commission (CCC), New Jersey law expressly provides shuffling at‐will as a method “which would counter the possible threat of card counters.” 22 This is because shuffling mid‐game can ruin a card counter's opportunity to raise his bet when the count is in his favor.
New Jersey law expressly permits shuffling at‐will. “[A]fter any round of play as may be determined by the casino licensee … the dealer shall shuffle the cards so that they are randomly intermixed.” 23 New Jersey regulations consider the effects of shuffling at‐will, particularly in the context of a blackjack player counting cards.
The economic impact of this proposed amendment would vary depending on when in fact the cards were shuffled. For example, if the cards were always shuffled after the first round of play regardless of the point count, then the casino advantage against the basic strategy player and average player would probably remain the same with the advantage enjoyed by the card counter being decreased. If the cards, however, were only shuffled in positive point count situations and not in negative point count situations, the casino advantage against all types of players would increase. 24
The Third Circuit in Doug Grant cited the CCC's above analysis as evidence that New Jersey's shuffle at‐will statute was intended to be a countermeasure for casinos against card counters. 25 However, the CCC's statement acknowledges that shuffling at‐will when the count favors the player alters the random chance of the game. The CCC does not go so far as to take a position on a casino's ability to use information about the count to decide when to shuffle at‐will. It simply acknowledges the varying “economic impact” that such practices could have.
The CCC has considered the use of shuffling at‐will when the casino is aware of a card counter. Again, Mr. Gushin, a commissioner of the CCC, stated regarding shuffling at‐will:
We looked at this request and were cognizant of the card counter issues related with respect to this petition, but we also believe there are other procedural rules that this Commission has adopted which would counter the possible threat of card counters here. I refer specifically to shuffling at‐will, which could be utilized at any time, should Tropicana feel that it has a problem with respect to card counters.
26
The CCC stated that shuffling at‐will can prevent card counters from being successful (by altering the count when it is in the card counter's favor), but, again, the CCC stopped short of considering a casino that may choose to shuffle at‐will only when the count favors the player. The CCC acknowledges that shuffling in positive point situations and not in negative point situations would alter the game in favor of the casino, but the CCC also references shuffling at‐will as a potential one‐off weapon to use against card counters.
Unlike New Jersey, Nevada statutes do not contain specific provisions for each game or any specific provision on shuffling at‐will. Dealing procedures in Nevada are established by the casinos, not by regulators. 27 However, the chairman of the Nevada Gaming Control Board has indicated that preferential shuffling, with or without the aid of a surveillance system, is likely impermissible under Regulation 5. 28
IV. Legality of Preferential Shuffling
Before examining the legality of preferential shuffling, it is important to understand exactly how preferential shuffling affects the game. By shuffling every time the count favors the house, the player is exposed to less favorable cards over the course of a game and has a decreased chance of winning. Consider this simplified example of the effects of preferential shuffling:
The deck has only four cards, two red and two black. The dealer turns a card for the player. If it is red, the player wins. If it is black, the player loses. This is a perfectly fair game because the player's chances of winning and losing are 50/50. When the dealer shuffles the cards, there are six equally likely orderings of the deck, as shown by the six rows in Table 1. However, a dealer who uses preferential shuffling will shuffle the cards early whenever there are more red cards remaining in the deck than black cards. Therefore, the player will only be exposed to the highlighted cards in Table 1. The player's odds of receiving a red card have dropped from 50% to just under 36%.
29
Preferential Shuffling Example
Reprinted from
In a game of blackjack, rather than red cards, tens and aces are cards favorable to the player. When many tens and aces appear early in the game, the casino will continue to deal so that the less favorable cards are eventually dealt. This is a fair game in which, theoretically, the entire deck is used. However, when few tens or aces are dealt early in the game, the dealer may choose to shuffle at‐will, changing a deck that should have a count favorable to the player to a deck that is fully randomized again. Over the long term, the player's chance of winning is decreased just like the player in the example above. 30
Rather than simply track tens and aces, the casino can count cards. Unlike a card counter who likely uses an imperfect system, 31 a casino with an advanced surveillance system can count cards perfectly. It knows the exact moment the remaining cards become favorable to the players. It can then put procedures in place to reshuffle the deck whenever the remaining cards reach a certain threshold that is favorable to the player. The casino can methodically and precisely inhibit a player's access to hands that are favorable to him.
This presents potential problems regarding the way the casino is conducting the game. Changing the player's odds of receiving favorable cards could fall within state cheating statutes. 32 Also, leaving this power in the sole discretion of the casino allows the casino to give similarly situated patrons different odds. 33 Perhaps most concerning is that all of these issues apply to all casino patrons, regardless of whether they are card counters. This is inconsistent with the Nevada and New Jersey gaming commissions' conception of preferential shuffling as merely a tool to use against card counters. 34
In Doug Grant, the Third Circuit states explicitly that preferential shuffling is legal under New Jersey's law on shuffling at‐will and does not run afoul of New Jersey cheating statutes. Under New Jersey law it is unlawful to knowingly operate games “which have in any manner been … operated in a manner, the result of which tends to deceive the public or tends to alter … the normal chance of the game which could determine or alter the result of the game.” 35 The Third Circuit reasoned that preferential shuffling did not alter the random selection of the game because the CCC specifically considered the effects shuffling at‐will could have in positive and negative point count situations. 36
However, as discussed above, the CCC simply acknowledged that a casino's decision to shuffle mid‐game can change the odds for a player if performed when the count is high. Such an observation does not amount to an endorsement of such techniques with the intent to alter the random chance of the game. New Jersey's shuffling at‐will statute permits shuffling “after any round of play as may be determined by the casino licensee.” 37 This appears to be a broad grant of authority. The statute does not indicate that the dealer may or may not have any specific intent in mind when shuffling, such as the intent to reduce the players' odds of winning.
The casino also may not operate the game in a manner, “the result of which tends to deceive the public.” 38 The casino's ability to implement this technique without any notice to the player indicates that it may be operated in a manner that deceives the public as to the player's odds in the game. To return to the previous example, the player expects to have a 7.7% chance of receiving an ace. However, if the casino is using preferential shuffling, the player's chances of receiving an ace are actually less than 7.7%.
The casino also has a duty to preserve the “normal random chance of the game which could determine or alter the result of the game.” 39 Read in conjunction with the shuffling at‐will statute, the casino may shuffle at any time, but may not do so in a manner that may alter the odds of the game or deceive the public.
When confronted with this argument, the Third Circuit concluded that preferential shuffling does not violate the New Jersey cheating statute because the CCC determines what “random selection” is, and because New Jersey law permits shuffling at‐will, “random selection” must include the results of these practices. 40
The Third Circuit avoids the conflict between true random selection and preferential shuffling by replacing New Jersey's shuffling at‐will statute with a preferential shuffling statute. The court concluded that because the statute considered the effects shuffling at‐will has on the odds of a game, the provision must necessarily permit a casino to utilize shuffling at‐will to intentionally alter the odds of the game.
This conclusion also seems inconsistent with the New Jersey Supreme Court's analysis in Uston v. Resorts Int'l Hotel, Inc. There, the Court found the provisions under the Casino Control Act to be extensive and all‐encompassing. “All gaming must be conducted according to rules” which set forth limitations necessary to assure “fair odds to and maximum participation by casino patrons.” 41 The CCC's provisions on blackjack are especially detailed. “It is no exaggeration to state that the Commission's regulation of blackjack is more extensive than the entire administrative regulation of many industries.” 42 The CCC did not explicitly permit casinos to shuffle at‐will when the casinos detect card counters, and it certainly did not permit casinos to shuffle when the count favors the player. Doing so is a violation of N.J. Stat. Ann. § 5:12–115(a)(2).
Read in harmony, New Jersey's shuffling at‐will statute and cheating statute permit a casino to shuffle at‐will only when shuffling is truly random. The dealer (or whoever orders the dealer to shuffle) cannot know the count. The Third Circuit in Doug Grant gives too much deference to the CCC's broad authority over shuffling procedures and its ability to define “random chance.” 43 The CCC has merely noted that shuffling at‐will can have adverse effects on players, or casinos, if used during positive or negative point count situations. 44 Casinos still have a duty to preserve players' expectations that they have the same chances of drawing an ace as they do a two.
The first New Jersey case to address preferential shuffling, Campione, reasoned that the practice is fair because it affects all players at the table evenly. 45 Therefore, even when the casino's intent is to hinder the card counter, all players are affected evenly and no player is disadvantaged. This supposedly fair result is limited by the casino's ability to catch the card counter. Suppose two card counters are playing at two separate, otherwise identical, blackjack tables. One counter utilizes methods of disguising its card counting abilities, while the other is detected. All players at the detected card table are subject to additional shuffles when the count favors them, while the other table is not. One table has better odds than the other, even though the casino advertises the tables as having the same rules and payouts. Here, preferential shuffling's effect on all players at a particular table changes preferential shuffling from a preventative measure to a tool that alters the odds of the game from table to table, unbeknownst to the players.
Now, imagine a scenario in which a card counter is incorrectly identified. The odds at one table are then silently altered and without any supposed justification articulated by the CCC. 46 The problem, again, lies in the casino's discretion. The shuffling at‐will statute has no standard set for a casino's motive in deciding to shuffle mid‐game. Due to that absence, Doug Grant and Campione concluded that a casino must be able to shuffle whenever it wants. That may be never, only when a card counter is suspected, or every time the count is in favor of the player. Once one turns to the New Jersey cheating statute, the limitation to shuffling at‐will becomes obvious. The casino may only shuffle if it does so to truly further the random chance of the game. Therefore, the decision to shuffle at‐will must be truly random as well.
Opponents may assert that this is impossible. The casino would have to track the count, determine when the count favors neither player nor casino, and then shuffle. This would render the provision pointless. Rather than track the count at all, the casino should simply shuffle at random.
Shuffling at‐will can still be used to deter counters, just not in a manner that alters the card counter's odds. For instance, if a casino detects a card counter, the casino may shuffle at any point but may not do so with reference to the count. This standard is complicated if the point of detection is always associated with a point in the game in which the count is high (e.g., when the card counter substantially increases his bet). If the casino finds this to be the case, the dealer must wait until the shoe is cleared before utilizing shuffling at‐will. This result may appear harsh to the casino, but remember that prior to any point in which the dealer knows the count, the dealer may shuffle at any time. 47 This creates the best incentives for both parties. The dealer and player are incentivized to not know the count to avoid discretionary shuffling.
The intent and knowledge of the casino is of course very difficult to apprehend. Is the casino retaliating against a card counter? Is the casino shuffling because the count is high or simply shuffling randomly? These are difficult for the average casino goer to know. Only perhaps the best card counters could pick up on such behavior. Because most patrons are unaware of how preferential shuffling might affect the game they are playing, the shuffle at‐will provision must be clear that casinos that wish to utilize this countermeasure must do so in a manner that does not alter the random chance of the game.
For example, casinos could have a wall of separation between the employee who knows the count and the dealer or other employee who decides when to shuffle. The casino's surveillance team could use smart tables and smart shoes to count cards and more easily spot card counters. Surveillance could notify the dealer or floorman when a card counter is present. The dealer can then increase the frequency of shuffling or simply shuffle at random without risk of such shuffles being motivated by knowledge of a hot shoe.
V. Conclusion
Critics of preferential shuffling contend that this issue is just one example of a larger problem: a lack of representation for casino patrons in the world of casino regulation. 48 However, as discussed in the beginning of this article, the casino has an interest in creating a fair gaming environment for the player as well, or at least the appearance of one. 49 That is why gaming regulators should remove the cloud of uncertainty over blackjack shuffling. Shuffling at‐will should be permissible only when its use is truly random. Only then will it truly enhance the randomization of the game, deter card counters, and give all players a fair shot.
Footnotes
2
David G. Schwartz, Nevada Gaming Revenues: Long‐Term Trends 3 (Las Vegas: Center for Gaming Research, University Libraries, University of Nevada Las Vegas, 2017).
3
The “count” refers to the number of cards left in the deck that are favorable to the player relative to the cards that are unfavorable to the player. A high “count” means the player's odds of winning have increased.
4
Bolton, supra note 1, at 4.
6
Uston v. Resorts Int'l Hotel, Inc., 89 N.J. 163, 166, 445 A.2d 370, 371 (1982).
7
A continuous shuffle machine (CMS) shuffles the cards throughout a game of blackjack, making card counting impossible. Casinos have also made small changes to the rules of blackjack that increase the house's odds. For instance, the dealer hitting on a soft 17 increases the house edge by about .2%. Preventing a player from doubling after a split increases the house edge by about .15%. There are other limitations on a player's ability to surrender or resplit cards which gives the house an increased edge, but by far the biggest change is paying six to five on a player's blackjack as opposed to the traditional three to two payout. This lower payout on blackjacks increases the house edge by about 1.5%.
8
9
See Uston, 89 N.J. 163.
10
See Robert A. Loeb, Casino Countermeasures, Preferential Shuffling and Casino Cheating, in
11
Nevada and New Jersey were chosen because of their high levels of casino activity and highly developed system of gaming laws.
12
13
David W. Schnell‐Davis, High‐Tech Casino Advantage Play: Legislative Approaches to the Threat of Predictive Devices, 3
15
See
16
In July 1994, Nelson Rose stated, “It seems the eye in the sky has acquired a brain. The watchers have learned to count cards.” I. Nelson Rose, Card Counting by Casinos, in
17
Schnell‐Davis, supra note 13, at 332 (citing Allen v. Nev. State Gaming Control Bd., No. A493817 (D. Nev. Oct. 18, 2004)).
18
See Sheriff v. Anderson, 746 P.2d 643 (Nev. 1987) (involving a player using “computer shoes”); Van v. Grand Casinos of Mississippi, Inc., 767 So. 2d 1014 (Miss. 2000) (regarding accusations of casino dealers using shuffling techniques favorable to dealers' relatives).
19
As has been the case with continuous shuffle machines and other countermeasures. See Jones, supra note 8.
20
I. Nelson Rose, Dealers Who Count Cards, in
21
See
22
Campione v. Adamar of New Jersey, Inc., 274 N.J. Super. 63, 80, 643 A.2d 42, 51 (Law. Div. 1993), rev'd, 302 N.J. Super. 99, 694 A.2d 1045 (App. Div. 1997), aff'd as modified, 155 N.J. 245, 714 A.2d 299 (1998) (quoting Statement of Commissioner Gushin).
23
24
Doug Grant, Inc. v. Greate Bay Casino Corp., 232 F.3d 173, 185 (3d Cir. 2000) (citing 14 N.J. Reg. 470 (May 17, 1982)).
25
Id.
26
Campione, 274 N.J. Super. at 80 (quoting Statement of Commissioner Gushin).
27
Rose, supra note 20, at 48.
28
Schnell‐Davis, supra note 13, at 332 (citing an interview with Mark Lipparelli).
29
See
30
For a slightly more complex example, consider the following: The odds of an Ace from the top of a fresh deck are one in thirteen, or 7.7%. Now, imagine if 13 cards were played in a game of blackjack and either one ace or no aces appear. The dealer responds to this information by electing to shuffle at‐will. If two or more aces appear, the dealer continues to deal. This is a basic preferential shuffling strategy. The dealer “randomizes” the deck when the players' odds of receiving an ace are greater than 7.7%. But the game continues when the aces appear early in the game, and the player is left with less than a 7.7% chance of receiving an ace for the rest of the game. The long‐term odds of receiving an ace in the above scenario will be less than 7.7%, the normal odds of receiving an ace. See I. Nelson Rose, Preferential Shuffling, in
31
A common card counting system is Hi‐Lo in which the counter assigns a 0 to 7, 8 and 9, +1 to lower cards, and −1 to tens and aces.
32
“In both New Jersey and Nevada, cheating is defined as doing acts which tend to alter the normal random chances or results of the game.” I. Nelson Rose, Preferential Shuffling, in
33
“Strangers might find the deck shuffled every time it turns positive. Friends could be treated quite differently.” I. Nelson Rose, Card Counting by Casinos, in
34
Campione v. Adamar of New Jersey, Inc., 274 N.J. Super. 63, 80, 643 A.2d 42, 51 (Law. Div. 1993), rev'd, 302 N.J. Super. 99, 694 A.2d 1045 (App. Div. 1997), aff'd as modified, 155 N.J. 245, 714 A.2d 299 (1998).
35
36
Doug Grant, Inc. v. Greate Bay Casino Corp., 232 F.3d 173, 185 (3d Cir. 2000) (citing 14 N.J. Reg. 470 (May 17, 1982)).
37
38
Id.
39
40
Doug Grant, Inc., 232 F.3d at 186 (“[I]t is absurd to allege that practices approved by the Commission as being consistent with its rules constitute ‘cheating’ under section 115 of the Act.”).
41
Uston v. Resorts Int'l Hotel, Inc., 89 N.J. 163, 168, 445 A.2d 370, 372 (1982).
42
Id. at 169.
43
Doug Grant, Inc., 232 F.3d at 185–86.
44
Id. at 185 (citing 14 N.J. Reg. 470 (May 17, 1982)).
45
Campione v. Adamar of New Jersey, Inc., 274 N.J. Super. 63, 79, 643 A.2d 42, 50 (Law. Div. 1993), rev'd, 302 N.J. Super. 99, 694 A.2d 1045 (App. Div. 1997), aff'd as modified, 155 N.J. 245, 714 A.2d 299 (1998).
46
Id. at 80 (quoting Statement of Commissioner Gushin).
47
Also recall that the house has approximately a 2% edge over a player who plays the game perfectly without counting cards.
48
Loeb, supra note 10, at 108 (describing the casino as a place “where legal rules give way to house rules”).
49
