Zman's Reply - long
From Stanford Wong's BJ21
Posted by Zman on 1 Jul 1998, 2:47 p.m., in response to Question for Zman on comps -- rolled off the page, posted by Chris on 29 Jun 1998, 5:41 p.m.
Chris,
To answer your questions let me break them down and answer them one at a time.
Q: How many hands do the casinos assume per hour?
A: I based my estimates on GPA's (game pace audits) that we performed on the dealers at the Aladdin. I estimated that a normal table of three to four players allowed a player to receive 70 hands an hour. The GPA takes into consideration the number of players and includes the dealer as well. We expected about 350 hph (hands per hour) from our dealers. Once I put in the shuffling machines I raised the average up to 75 hands. Normal GPA's for the double deck without machines ran about 375 to 400 hph. The difference between the DD and the six decks is due to the shuffling process. At the Aladdin we dealt deeper and shuffled faster then most casinos and actually had a very high hph. Since we only had two or three DD I used only one standard number, the 70 or 75 hands, for both game types.
Q: Does table crowding make a difference?
A: The system we used at the Aladdin was not very flexible. What I did was invent game types for the computer. We had "Blackjack" which was the norm at 70 (75) player hands per hour. I also had "Xblackjack" which represented only one or two players on a table. The number of player hph was set at 150. I then had "Sblackjack" which represented a full table. The player hph was set at 50.
Q: What is the assumed negative expectation for the player? Does it make any difference whether the customer is a BS player (-0.5%) or not?
A: I originally set the average h/a (house advantage) over the player at 1.5%. This was based on a study done by Peter Griffin back in the early 80's. Griffin estimated that the average player in LV played with a 1.67% disadvantage. His study was based upon the effect of how badly the average player varied for basic strategy and also took into consideration the average game type and average rules. I set the computer at 1.5% because I felt the average customer plays better today then in 1980 (based on the number of books, bs cards, and the effect of playing must of the games face up). After one year I compared the computer generated theoretical win with the actual win and the theoretical win was much higher. After looking at several factor (errors in game speed, playing time, rate card errors, etc.) I found that we were high on the average h/a. I then changed it to 1.2% and from then on the theoretical and actual came right in line. I also had to do something similar to the hph estimates when I wanted to indicate the ability of very good and very poor players. I used 0.5% for the basic strategy player/expert, which I indicated with an "E" (EXblackjack, EBlackjack and ESblackjack), and the beginning player I used 2.5% with the indicator letter of "B". From a theory standpoint this system looks good, however when you apply it in the pit it just doesn't work out real good. I had to get on my shift and pit managers to get the floor supervisors to make the correct adjustments on the rating cards. Most of the time the player was rated average speed and average playing ability. I had some real good BS players, along with so low level counters, and I set the default set to Eblackjack when ever their ratings were entered.
Q: Is the formula you provide an industry standard, or does it vary from place to place? I've heard about similar formulas with the percentage varying from 20% to 50%.
A: Their are several computer rating system being used in the industry. Some of them are very precise and some just come close. The standard for return on comp dollar spent is usually between 50% and 35%. It depends on how aggressive the casino is in their marketing of casino customers. Also, comp dollars are used to defer soft costs. Soft costs are in hotel rooms, food, beverage, and gift shop. Hard cost like airfares, show tickets off property, and cash back programs are treated different. Usually the casino will cut the comp dollars amount by half we deferring these costs. Example: If you had $400 in comp dollars and your in-house expenses come to $200, you could argue for $100 cash back for your airline ticket, not the entire $200.
(Note: All the games were set up in a similar manner. I did use double the ante bet in CSP as the average bet, and I had the floor supervisors divide the average bet in Pai Gow Poker in half to make up for the ties and the slowness of the game. I also used a more aggressive house way in PGP so I could eliminate some of the pushes.)
Q: I'd appreciate any insights you might have as a guideline when asking for comps.
A: My best answer here is to suggest you go out and buy Max Rubin's "Comp City". The book explains everything you want to know almost to the point of casino executive embarrassment. When I was at the Aladdin I had all my shift managers and manager relief's read his book.
I hope I answered your question for you. Good luck and good hunting
Zman