ES: Hold (Win) Percentage and EV, Table Game and Slot Machine, Drop and Handle

[ES]

Why the statements that "The casino hold (win) percentage for slot machines is lower than the hold for blackjack, craps and roulette. Therefore, slots are a better bet than table games." is RUBBISH.

Hold Percentage for a table game = Casino Win/Drop.
Drop is NOT the amount of money lost by the players. It is the amount of cash markers in the drop box exchanged by players for chips. Suppos that a player buys in for $200 at a $10 table and loses $30. The hold percentage for his play is 30/200 = 15%. Did he likely play each chip he received for cash once, win, lose or push, and then leave? Most likely not. Most such players recycle winnings, playing for a predetermined time limit, e.g. until a show starts or they have to get on the bus out of Atlantic City, or, until they reach a goal or lose their limit, thus making bets totalling much more than the $200 buy in. Total of bets depends on many things including time or loss limits, speed of the game, when a player gets tired or bored, how much he is enjoying himself, etc. Prior to RFID and without much obrsevation from the eye in the sky, it was impossible to calculate the total amount of money bet by players. The drop is, however, easily calculated: Count the cash and markers in the drop box! Thus, the formula Hold Percentage = Win/Drop became a standard measure of casino win rates for table games

Slot machine Hold percentage = Casino Win / Handle.
The handle is the total number of coins counted by machine's in-meter. Again typical play is not to bet each coin once, win or lose. Suppose that a player Player gets 1000 dollar tokens, plays a machine with 90% payback percentage, experiences exactly that result over 2 cycles of play. After first cycle player has lost $100, leaving himself with $900 and the machine's in-meter reads 1,000; after second cycle player has lost another $90, for a total loss of $190 quarters and the machine's in-meter reads 1,900. Casino win percentage = Casino win / Handle = 190/1,900 = 10%, which is the EV for the machine. This is true after each cycle of play.

Do players experience a slot machine's EV on each cycle? No, if they did, then no one would ever win and slots would no longer be make money for the casinos. Casinos hate it if any player has positive EV, but they need enough winners through randomness, but not too many, to attract business. What is true, however, is that the general slot-playing public's aggregate experience differs by less and less from the EV as the amount of play gets very large over time by the law of large numbers.

If the casino used the drop, rather than the handle, in the denominator for slot machines, then the hold percentage is 190/1,000 = 19%. Slot machines have long had in-meters, but tables had long had no method of precisely measuring the total amount bet by players. This difference between tables and slots works to the detriment of players who are easily misled into thinking that slots are the best bet, in addition to having the lowest minimum bet and least amount of intimidation.

Calculation of the hold percentage or win percentage is susceptible to inaccuracies. Slot players, especially at low denomination machine, play pocket change in addition to rolls of coins obtained in exchange for paper money. Such pocket change would not be counted if the Win/Drop formula, where the Drop is the amount of paper money exchanged for coins and tokens, is used instead of the Win/Handle formula. The hold percentage at a blackjack table is artificially decreased in the following circumstance: A big, but less knowledgeable player buys in for $5,000 at a $100 table. He bets $100 on the the first hand and loses with 20 when the third base player makes an unconventional play. In disgust, he takes his $4,900 in chips to a craps table and lose, without having bought in at such table. The hold percentage at that table is artificially increased.
Calculation of EV is math. Calculation of the hold percentage is casino accounting.

Much of the above was explained to me by Don about 20 years ago. I hope this helps.

[Francis Salmon]

I entirely agree that the only meaningful way to compare casino games is win divided by total action. For precision I would like to add that for blackjack the action should include only the initial wagers and ignore splits, doubles and insurance.
In your post you mention RFID. Without knowing what you mean exactly with this I guess it has to do with player rating.
Now my question: Do casinos actually use RFID for calculating their EV at Blackjack or are they still sticking with the old hold-system?
I would be very much interested in the actual Blackjack-EV of casinos. Not only in order to compare with other games but rather to compare the Blackjack-EV between different continents, countries and casinos.

Francis Salmon

[ES]

I don't know if RFID Chips are used to measure the win percentage or to rate individual players. Drop and win numbers are reported individually for each casino for each game in Atlantic City. They are available from the NJ Casino Control Commission. It would seem that the win percentage is proportional to EV, i.e. if one casino has all 8 deck games w/o resplitting of pairs and another casino has some or all 6 deck games with resplitting--these are the only fifferences in rules and playing conditions currently available in Atlantic City, except for penetration, which does not affect non-counters--then the casino with more favorable rules and a lower EV for the casino would have a lower win percentage. This is not necessarily true because the longer that players play in a negative EV game the more the casino wins and length of play can depend on factors not related to EV, e.g. how the casino treats players. A casino with a lower hold percentage can still be more profitable than one with a higher hold if it gets more action. Casino gambling is an other peoples money business.

[Don Schlesinger]

> I would be very much interested in the actual
> Blackjack-EV of casinos. Not only in order to compare
> with other games but rather to compare the
> Blackjack-EV between different continents, countries
> and casinos.

Petr Griffin did exhaustive studies of this some time ago. Actual e.v. for casino blackjack is about 2%. Those who think that "stupid" players play worse than this are probably incorrect. Believe it or not, it is difficult to play more ignorantly than -2% at a decent-rules BJ game.

Don

[Don Schlesinger]

I read through this spectacularly well-written, most informative piece, and was about to post, to commend you for such a fine explanation, when I got to the last line and really had to laugh. :-)

> Much of the above was explained to me by Don about 20
> years ago. I hope this helps.

Just curious: As I obviously didn't explain any of this over the Internet, from where do we know each other?

And, by the way, since I'm sure you weren't quoting verbatim, you did, indeed, do a wonderful job of explaining.

Don

[ES]

Maybe it was 15 years ago. We don't know each other. I probably wrote to you c/o Blackjack Forum and you were kind enough to respond. I probably still have your letter. I didn't quote verbatim. I most likely added more material. I am probably good at explaining beginner to intermediate topics, but you and Norm have a much better understanding of advanced material. I do not find all of BJA to be easy reading, very informative, however.

[ES]

Play casino war, three card poker, Carribean stud, let it ride, etc.

[Bettie]

> Play casino war, three card poker, Carribean stud, let
> it ride, etc.

Caribbean Draw. It was only available in certain states, Nevada not being one of them, and may still be available in those states. S.O. found a strategy for it many years ago and we would occasionally play while waiting for a Spanish 21 table to open up (back in the day when you could still find S17 tables). Caribbean Draw was slow, but I liked it. Kind of a "poker-for-dummies" game. And yes, I do realize that I just called myself a dummy.

I still remember the first time I saw Casino War; we were staying at the Bellagio and walked by the only table, with just one player, at the end of a row of tables. S.O. stopped and said he wanted to watch until there was a "war." We didn't have to wait but 3 hands later, and we were laughing the whole time, making cracks about the "National Lampoon's Vegas Vacation" movie. You know, the scene where Clark and his brother-in-law go into the casino (the exterior is the Klondike on the south Strip) and play "guess the number" and "rock-paper-scissors," along with "casino war." Hysterical!

Bettie