0	−0.024608	0.016989	−8.00%*	5.52%*
1	−0.006024	0.004125	−7.83%	5.36%
2	−0.005665	0.003777	−7.36%	4.91%
3	−0.005166	0.003293	−6.72%	4.28%
4	−0.004326	0.002482	−5.62%	3.23%
5	−0.002671	0.000871	−3.47%	1.13%
6	0.000582	−0.002284	0.76%	−2.97%
7	0.005665	−0.00724	7.36%	−9.41%
8	0.013303	−0.014708	17.29%	−19.12%
9	0.01656	−0.017884	21.53%	−23.25%

*Reported in the literature as occurring 4× more often but figures are normalized because there are 4× as many 10 value cards in the deck.

In Version 2.1 of the BFC Rules Program, 2006 (e.g., referred to below as the Leslie Advantage) and as reported by two other researchers, Kilby (1996) and Whiting (1997), the statistical dealer advantages were shown and are compared below with the statistical advantage determined by Joseph with First Card Advantage knowledge.

FIRSTCARD ADVANTAGE COMPARISON

1^stCard	Leslie Adv.	Kilby Adv.	Whiting Adv.	Joseph Adv.

0	5.52%	5.52%	5.59%	10.77%
1	5.36%	5.36%	5.29%	9.63%
2	4.91%	4.91%	4.80%	8.68%
3	4.28%	4.289	4.14%	7.61%
4	3.23%	3.23%	2.87%	5.50%
5	1.13%	1.13%	1.20%	3.28%
6	0.76%	0.76%	0.68%	2.24%
7	7.36%	7.37%	7.22%	10.12%
8	17.29%	17.29%	17.21%	22.87%
9	21.53%	21.53%	21.49%	27.16%

As can be readily seen from these comparisons, First Card Advantage knowledge can significantly shift player advantage by appropriately placing wagers on the more advantaged position (Player versus Banker).

The system, method and apparatus described herein is intended for detecting a particular type of cheating in Baccarat termed “First Card Advantage (FCA).” If the player has the knowledge of the rank of the first card, and knowing this first card always goes into the player hand, the cheating player can bet on the player hand when the card value favors the player winning, or can bet on the banker hand if the first card is unfavorable to the player hand. The system uses a card rank reader, a position detector at each player station, a processor and software that correlates player hand selection and betting amounts with the rank of the first card dealt. The player position information may be correlated with player I.D. information either collected from a player tracking system, for example.

The basic content of the cheating method revolves around the fact that knowledge of the actual rank (suit is unimportant) of a first card out of a shoe and its intended location (banker/dealer hand or player hand) can assist players in predicting likely outcomes of the hand, or at least provide knowledge that is valuable in statistically assessing potential outcomes in that hand of baccarat.

With knowledge or pre-knowledge of first cards out of the shoe, the knowledgeable player would be likely make a Player wager of an increased amount on hands with a decided Player bet advantage, such as by way of a non-limiting example when the initial card is an 8 or 9. The knowledgeable player may make a Banker wager of an increased amount on hands with a decided banker bet advantage, such as when an initial card out of the shoe is a low ranking card, such as a 0, 1 or 2. Similarly, initial cards out of the shoe being a value of 7, 6, 5, 4 or 3 are wagered upon more moderately by a player.

The system of the invention therefore collects data and evaluates player wagering history, along with subsequent knowledge of the initial cards to determine if a pattern of betting indicates that a player was wagering with a high degree of knowledge of the count of a first card out of the dealer's shoe. This system would therefore be implemented as follows.

Card Reader

A component or system must be provided that can read at least the numerical count or value of a first card out of a dealer's shoe in the play of a round of the game of baccarat. The count (i.e. the numerical value of the card) may be taken in a wide variety of ways. For example, sensors or cameras may be placed in the shoe or shuffling device providing cards to the game. A card reading sensor may be positioned in the table and/or may be read by an overhead or field of view camera at the table. Such a reading may be done at various times. For example, cards may be read before the card is delivered. A shuffler that reads cards or a card reading shoe would read cards internally before cards are delivered. Cards may be read as the cards are removed from the shuffler or shoe, for example, by a reading element in the delivery end of the shoe, at the tip of the shoe, or adjacent the delivery end of the shoe. Cards may be read as they are moved away from the shoe or when a first card out of the shoe is placed face down in the appropriate first card out of the shoe position. Cards may be read on the table, for example at a player position when the card is face down, or when the card is revealed during play at the first card out of the shoe position.

The reading may be done by any format of information gathering that is known, such as symbol reading, cameras, code reading (with codes on cards that are read by any method), dealer or observer input (e.g., the dealer visually reading a card and inputting data) and the like. The component is needed to provide information that is transmitted to at least a data storage base or other memory storage for real time or later time evaluation and analysis. The import of the data is to associate a specific count for a first card out of the shoe with a particular round of play for that first card out of the shoe. It is likely that the data of the first card count will be associated with or identified with a round of play identifier that may include at least some information such as Casino name, table number, date, time, hand number, dealer ID, and player ID. The first card count is also associated with a hand assignment such as the player hand or the banker hand.

Once the first card out of the shoe (for a particular round of play) has been identified and associated with a round of play and table and/or particular hand and/or player, that information must be transmitted at least to temporary storage, either by hardwired transmission or wireless transmission to the memory storage element. This transmission may be in real time, in packets for each round of play, or sent as a set of date at intervals from the table or from the pit. The transmission may be raw signals, translated signals, raw data or final data, as long as it is able to be stored with a content sufficient for the purposes of the practice of this technology. In this regard, proprietary technology of the assignee as described in Published US Patent Applications 20050288083; 20050012270; 20050026682; 20050051955; 20050062227; 20050073102; 20050082750; 20050282622; 20050288083; and 20050288085 are incorporated herein by reference to support the technical practice of the technology described herein.

The bets may also be sensed, valued, measured and information, data and signals there from, forwarded to the data storage, memory or active processor. The initial signals or data may be generated by any of the known sensing or measuring systems that have been contemplated by the gaming industry, much of which is disclosed in the applications and patents identified herein and which have been incorporated by reference. Among the systems that can be used are image capture systems, edge reading systems, magnetic strip reading systems, RFID chip reading systems, weighing of chips, code reading on chips (especially edge coding), color discrimination of chips, IR responsive reading of chips, and any other system that can provide data about the value of a wager at a specific position. Systems have been proposed and described that can actively follow the location of individual chips and associate those individual chips uniquely with individual players. Any such system can be used in the practice of the present technology. Even visual observation of wagering amounts and manual input by observers can be used in this feature of the technology, although automated wager reading systems are preferred.

FIG. 1 shows a casino table2 with the system implemented thereon. A shuffler4 is shown, along with aplayer hand position6,8 a

dealer hand position

10 and12, with the first card dealt being toposition6, a dealingshoe3, and three player wagering positions20,22 and24 with two wagering positions at each player wagering position, for wagers on the player's

hand

6,8 or the dealer's

hand

10 and12. The sensors, readers, cameras, etc. associated with the generation of signals and creation of data or information are shown sending signals todatabase30 and fromdatabase30 tomain processor32. Thedatabase30 andmain processor32 may be local or accessible via a network connection (not shown). It is to be remembered that this is only one of a number of embodiments described and enabled herein and theFIG. 1 should not be interpreted as limiting the generic scope of the technology described and enabled herein.

The identification of individual player positions, and associating wagers and amounts of wagers with player positions is also disclosed within the technology incorporated herein and known to the industry. This can be assisted by biometric information taken from the player, player ID cards, player ID's associated with purchased and identifiable/traceable chips, surveillance input from cameras or live personnel in the casino pits or any other method of providing the information so that it is associated with the identity of the players making the wagers, or at least the location/position of players making the wagers, which information may be later enhanced for specific and more complete player identification at a later time.

The information should be evaluated with at least some of the following sets of data as a basis for analysis of “first card knowledge” cheating:

- Player Position (PP), Round identification for Baccarat (RI), Wager Amount (WA), Hand Wagered on (HW), count of first card (CF), and location in hand of first card (LH).
  This information set of PP-RI-WA-HW-CF-LH is a composite or collective piece of data (CPD) that may be used for the analysis necessary for identification of a wagering pattern indicative of undesirable or even criminal activity. When this information is stored in the database, its content can be used for the comparison. Either a general and continuing review of the data by aprocessor32 may be used to identify the existence of a problem, or triggering events based on partial analysis may be used to initiate a review of a specific table or specific player.

For example, if a player at a table shows erratic variations in wagering patterns, such as 1×, 3× and 10× at other times, the variations may be sufficient to trigger a review. If the first card dealt was a card of significant count (e.g., 0 or 1, with the player wagering on the Bank hand, or 8 and 9 and the player wagering on the Player hand), more detailed screening and review of the wagering at that position may be undertaken. More likely, a continual review of all wagering events at each player position may be investigated. Team playing analysis, later described may also be undertaken.

The basis of the analysis is as follows. Successful use of first-card out advantage (FCA) is most successful when the selection of the wager type (Player versus Bank) is based upon that knowledge. This can be tracked in a number of different ways. It is first understood that the system provides data in the form of the value of the first card out in each hand. Analysis may be based only upon sets of count cards in the first card, that is a set may be 9 only, 0 only, 8 and 9, 9 and 0, 8, 9 and 0, 9, 0 and 1, 8, 9, 0 and 1, and the like. A processor accessing this information may analyze the information associated with data relating to only wagers within the selected set. The analysis will operate on the basis of the following type of operation of software:



	Look at Round #
	Is first card within set?

	Ignore if not in set
	Proceed to next step if in set

Was wager placed on strategic hand?

	Ignore if not made on strategic hand
	Proceed to next step if made on strategic hand

What was statistical variance of this wager versus wager on first

cards outside set?

	If statistical variation insignificant, ignore
	If statistical variation significant, alarm and further
	analysis of wagers of this player/position.

This is the evaluation of the wagering play on a single hand after at least a few hands have been played at a table by a single player. The need for data on multiple hands is apparent as a single player may sit at a table and make a single large and fortunate wager, without any FCA knowledge. It is the statistical and collective event situation that the casino is interested in, which could be indicative of illegal and undesirable activity.

For example, look at the following wagering pattern of two players at a single table and statistical pattern implying FCA can be seen at one position. B represents a banker hand wager and P represents a player hand wager. F represents the count of the first card out.



	PLAYER 1		PLAYER 2

Wager				Wager
Amount	Hand Played	F	Hand Played	Amount

$10	B	8 P	P	$100
$10	B	6 P	P	$10
$20	P	0 B	P	$100
$20	B	5 B	B	$10
$10	P	9 P	P	$100
$20	B	7 B	B	$30
$30	P	1 B	P	$100
$30	P	4 P	B	$10
$10	B	8 B	B	$100
$20	B	2 B	P	$20

A review of the betting patterns, by visual observation and objective analysis shows thatPlayer1 made some good bets and some bad bets, whilePlayer2 made wagers that were extraordinarily optimal wagers on each hand, with wagers required on each hand. Not only were the wagers optimal on hand selection, but also the variation in the amounts of the wagers was uniformly designed to take advantage of the assumed knowledge of the first card out and the location of that first card. Not only were the wagers generally properly placed, but the variation in the amount of a wager on the proper hands maximized the opportunity for profit. This data pattern for a single player would be readily evaluated by software looking for such statistical patterns of best position wagers coupled with best variation of wagers on hands.

It is to be expected that players, even with foreknowledge of the first card out would vary their wagers to avoid such ready detection, even betting against the favored position, usually with a reduced value wager, to fend off automated analysis of wagering patterns. For example, the analysis software and system may review groups of wagers (e.g., sets of ten wagers) and look at patterns over different ranges, as part of the automated program. The system may evaluate sets of wagers, (i.e., as large as 100 or more) and emphasize only statistically anomalous wagers. For example, if in 100 wagers, 25 wagers were statistically greater than others (the degree being set in the software, such as 3×,4×,5×, etc. with regard to the lowest wager or the average wager), and a statistically significant proportion of those wagers (e.g., all 25, at least 24, at least 23, at least 22, at least 21, or at least 20, for example) were optimally wagered, this would again be indicative of FCA knowledge. It is unlikely that wagers on a 50/50 event would be wagered at an optimal rate exceeding 80/20 or more over an extended period of time, especially when there is a multiplied wager amount associated with that optimal wager. The software would identify this anomalous pattern and an alert or flag would result from the data, most likely going to security, to the pit, to record keeping, to the host facility, or the like. The player would be identified and the potential source of the FCA would be sought.

First-card Out Pre-knowledge or advantage (FCA) can occur from a number of different causes. A player may be in collusion with a dealer (who may have access to card identification by control of the shuffling), the player may have marked cards, the dealer may be purposely exposing cards before cards are dealt from the shoe, the cards may be pre-stacked with the order of cards known to the player or dealer in collusion with a player, and the like.

As noted, the analysis may be done on a real time basis, that is, information from hands is immediately fed to a processor and the comparison of wagering variations tied to the count of first cards is made, or the analyses may be made over statistically significant time periods and over statistically meaningful numbers of hands. The analysis may be made by a processor at the table (with the appropriate software or hardware therein), in the pit, in a security area or at a central processor area.

One methodology of practicing the technology described herein is a method of detecting irregular play in the game of baccarat. The method may comprise each player electing to make a wager on one of the player hand, banker hand or a tie hand. The system provides information on and detects each player election. The system collects information on and detects at least a wager presence or preferably a wager presence and wager amount made by a player on one of a Player hand and a Banker hand. The system records the wager amount and associates that wager amount with a specific player or player position. The table ordinarily deals a first card in a round to the Player hand. Alternatively, the first card could be dealt to the Banker hand, without changing the underlying game or the First card Advantage issues discussed herein. The system detects and stores a rank of the first card dealt in the round. The system then compares player hand elections and wager amounts to stored data relating to player hand elections, wager amounts and first card out counts. Based on the stored data, the system determines whether the player may have or appears to have obtained unfair advantage by knowing the rank of the first card dealt and wagering in a manner to optimize that advantage.

The step of detecting each player's election may be accomplished by the detection of a wagering chip, especially one having a known value, on a location on the table that is assigned for receiving a wager or bet on the player hand or the banker hand. The player wagering location and wagering amount may be simultaneously or sequentially detected and transmitted. The player election, the sensing of the count of a first card and/or the wager amount may be effected or detected prior to a card being dealt or prior to a card being exposed. Data is stored to enable a comparison of statistical events, such as the statistical rate of a player wagering on a player hand or banker hand and the relationship of those wagers to the count of the first card out of the dealer's shoe delivered to the player's hand and/or the banker's hand. The stored data should correlate or enable correlation of at least information relating to a player's hand election with a rank of a first card dealt and the location in the player's hand or the dealer's hand of that first card dealt.

The methods may be practiced on apparatus and systems for detecting unfair player advantage in a game of baccarat that may, for example, include: a card count (which may be provided by reading rank) reader; a sensor for sensing at least a presence of a wager and preferably a value for the wager; a player hand election identifier; a player position identifier; and a processor programmed to analyze, compare, evaluate, categorize, and/or correlate player elections and at least one wager made with that election with stored data, wherein the processor performs an analysis to identify player wagering patterns or events that are indicative of prior knowledge of a first card dealt. The card count or rank, as previously noted, may be read at any stage in the play of the game by any device that can provide signals or data that can ultimately be stored as indicative of the card count or rank, in a shuffler, before insertion into a shoe, in the shoe, as it leaves the shoe, after it has left the shoe and before being exposed and after being exposed. Similarly and as previously described, the chip and wager value may be read by any available technology at any appropriate time during the gaming event in any manner that provides a signal or data that can ultimately be stored as identifying the amount of a wager in a gaming event. A preferred wager-reading system would comprise a gaming table with a first RFID antenna embedded in a player wager location and a second RFID antenna embedded in a banker wager location. Optionally, a third RFID antenna could be provided to record wagers on ties.

The location of the various reading, sensing, transmitting, storage, memory, and processing components may vary according to the design and choices of the manufacturer and user. For example, the memory for storing information relating to wager amounts may be in the shuffler, the dealing shoe, the game controller, a local processor or a central processor. The processor and/or memory may be located within a card-reading shoe, within a local computer, within a network computer, within a card-reading shoe, within a local computer, and/or within a network computer, and the memory may be accessible via a network connection, a hard wired line, wireless line, through nodes or any other communication line or link. The information relating to card count or rank, wagers and player hand election may be delivered via a network connection to a network database that can be accessed by a processor.