Although the CSM gets a lot of criticism from players the truth is that it actually lowers the house edge. To prove this I ran almost 75 billion hands through a simulation program. The following table displays the results according to the rules and whether a cut card or CSM was used. The other rules were: dealer stands on soft 17 (except single deck), double after a split allowed, split to up to three hands, no surrender, no unusual rule variations.

The reason for the difference in results at all is difficult to explain. The way the basic strategy player plays causes more large than small cards to be played. When a specified number of hands is played per shoe the result is the player friendly penetration into the deck. However the cut card removes that effect and levels the playing field. The following two tables show the distribution of ranks actually played in both a cut card and CSM games.

Note how the distribution is weighted towards large cards in the CSM game as opposed to the even distribution in the cut card game. The Chi-Squared statistic is a measurement of how far the results deviate from expected. The numbers in the lower right cell of each table show a Chi-Squared statistic of less than 3 in the cut card game and 1250 in the CSM game.

However it must also be stressed that the CSM allows the dealer to deal continuously, increasing the number of hands dealt per hour by about 20%. For the basic strategy player this will result in a greater expected loss on an hourly basis. In actual casinos I have only seen these machines used in shoe games however the effect is the same on the Internet where it is common to see games shuffled after every hand in all numbers of decks. The effect of shuffling after every hand and the use of a CSM are the same.