The legitimate percentage of chances in favor of the bank in this game is enormous. Out of thirty nine chances, the bank runs eighteen of losing and has twenty-one of winning, or three additional chances in its favor, which is equivalent to fully 5½ per cent. in favor of the bank in all cases, even where a bet is placed upon either of the zeros or the eagle. In the latter case, the bet on either zero or on the eagle is paid 35 to 1, the same as on any single number.
Here the bank has thirty-five chances out of thirty-nine of winning, and only one of losing, or four more chances in its favor than the payments warrant, thus yielding the same 5½ per cent.
It follows that the odds against the players in the various chances may be expressed as follows:
| Upon a single number, | 37 | to 1 |
| Upon any twelve numbers, | 13 | to 6 |
| Upon two numbers, | 18 | to 1 |
| Upon three numbers, | 11⅔ | to 1 |
| Upon four numbers, | 17 | to 2 |
| Upon six numbers, | 16 | to 3 |
| Upon odd or even, red or black, | 10 | to 9 |
In the case of a bet on the first or last eighteen numbers, the odds are ten to nine, the same as on odd or even, or red or black.
When, however, a stake is laid on all the numbers, and the bank only pays the winner thirty-five times his stake, it clears four; thus, supposing thirty-nine dollars to be a stake, and that the ball is thrown twice in a minute, the gain of the bank, without incurring the slightest risk, would be eight dollars per minute, or $480 per hour. Although, in whatever way a player may bet, the chances are always in favor of the bank, still the latter’s risk varies in proportion to the number of chances which are not filled up. To illustrate, if only ten numbers are filled, and the ball were to enter one of them, the bank would, in that case, lose thirty-four dollars, and only win eight; whereas, when all the numbers are filled, it wins four without risking a cent.
From what has been said, as to the chances in favor of the bank, it would seem to be hardly necessary to use any additional means of swindling, inasmuch as the percentage in its favor is so large that the game is very seldom beaten, even if “played on the square.” An old gambler once remarked in my presence, that the percentage of the game was forty per cent. worse than stealing. However, despite this fact, the gambler is not satisfied, and has succeeded in devising schemes, whereby he may win every bet made against him, if he sees fit.
The first method of cheating which I will describe, is as follows: The roulette is manufactured for the purpose, the machinery being entirely concealed from view. The gambler who manages the game can cause the ball (A) to fall in a red or black number, as he may think proper. After throwing the ball he watches it closely, and if it should fall in the red, when he wished it to go into the black, while still revolving, its course can be quickly changed to the desired color. This is accomplished by means of a lever attached to the circular wheel, and connecting with one of the legs of the roulette. This leg has the same appearance as others, but is a trifle shorter, not quite touching the table on which the roulette rests. The gambler has only to touch this leg while the wheel is revolving, and in a second the ball is changed from one color to another, as he may prefer. In fact, so quickly can the ball be changed, that it is difficult to detect the motion after one has been shown how it is managed, unless the wheel is turned slowly. This is one of the most ingenious contrivances in use.
There is yet another kind of roulette, which is made in the following manner: One-half of the small pieces of metal which form the pockets for the ball are made a trifle longer than the others, lettered on the diagram E E E. After the stakes have been placed, if the proprietor wishes the ball to fall in a red color, it is necessary for him merely to throw the ball around to the right hand, and if he wishes it to fall in the black, he casts the ball toward the left. The players may observe that he throws the ball in a different direction on different occasions, but the action appears to be so trivial that it excites no suspicion.
Another fraudulent contrivance used in playing this game consists in the gambler’s having two centers to a wheel, apparently identical, one of which, however, is “square” and the other “faked.” This device is known to the members of “the profession” as the “double center.” The “square” wheel is used at first, and, at an opportune moment, the “fake” is substituted, after which the sharper has everything his own way. This wheel is operated on very much the same principle as the “needle wheel,” for the construction of which the reader is referred to the chapter containing a description of that device. A system of levers radiating from the centre of the apparatus is operated by a rod terminating at the edge of the table. By bringing to bear the requisite pressure, these levers cause fine needle points (lettered C C C on the diagram) to rise through the cloth, one coming up in front of each alternate compartment on the rim, thus obstructing the entry of the ball and causing its course to be so changed that it shall fall into one of the next adjacent divisions, as in the case of the “needle wheel” above referred to.