cellardoor
Joined: Jan 4, 2011
• Posts: 70
February 4th, 2011 at 12:31:28 PM permalink
Quote: mkl654321

Yes, that would be logical behavior. And that logical behavior would last right up until the moment when there were three or four 7s and/or 11s on the comeout, and the logical bettor only won the table minimum each time. Isn't that part of the attraction of craps--that the comeout roll favors the Pass Line bettor? You can have a whole string of winners in a very short time.

I don't know about you, but I don't play the pass line specifically to profit from the comeout roll, it's just a nice bonus when a 7 or 11 hits. Wouldn't it be in the bettors favor, in the unlimited odds scenario, to keep the minimum on the pass line and just throw some money on a 'any 7' or '11' to make up for the missed opportunity?
thecesspit
Joined: Apr 19, 2010
• Posts: 5936
February 4th, 2011 at 1:43:27 PM permalink
Quote: mkl654321

Percentage points ARE "actual" gains. And that was my point--that by offering unlimited odds bets (which cost them nothing), they might attract more Pass Line bets. They don't care about the total dollar value bet--they care about total expected value. If you think about it, offering huge Pass Line odds bets is like a very compelling promotion that doesn't cost the casino a single dime. Of course, they take on additional variance, but that isn't all that big a deal since no odds bet pays better than 2-1, and the casinos have more than adequate cash reserves. It's unlikely that they would be in any danger.

My point is 1.41% EV BY itself does not tell you the profit of a table. They care about the total expected value IN dollars.

The variance of the pass line bet is ~1.

The variance of the odds bet is 1.55

A x5 odds bet there has a variance of 7 times the variance of a pass line bet, and x100... 150 time the pass line bet.

The variance on the odds starts to swamp that of the pass line bet... meaning that while the casino has adequate cash reserves it could take much much longer for it to settle towards the long term value (of \$0).

Say 30 passes per hour, 24 hours per day = 720 decisions on the pass line bet (I'm approximating away a whole set of things here like the fact the pass line resolves more often than the odds bet)

\$1 on the line, \$5 on the odds (x5 odds)

EV on pass line = 720 * 1.41% = \$10.52
Standard Deviation = sqrt(720) * 1 = \$26.83

EV on the odds = 720 * 0 = \$0.0
Standard Deviation = sqrt(720) * sqrt(1.55) * 5 = \$167.

As these are closely related the casino will made \$10.52 +/- \$387.66 (covering 2 standard deviations).

Over 365 days :

EV on pass line = 262,800 * 1.41% = \$3705.48
Standard deviation = sqrt(262,800) * 1 = \$512

EV on the odds = \$0
Standard deviation = sqrt(262,800) * sqrt(1.55) * 5 = \$3191

=> casino makes \$3705 +/- \$3703

=> There a chance (small though it is) with 5 times odds that the table would not post a profit from play with 5 times odds over the year.

As the odds goes up, there's more of a chance over the years play that the table is not profitable. I'm sure the minimum threshold for a table is more than "doesn't make loss" but is \$x per day, then you can see that the casino will be risking that the variance will hurt it's bottom line, even in a game it has the edge in.

(I've approximated a lot here, and may have made some glaring error).

Variance is what CAN make some punters winners over a period of time. The same variance can mean a casino loses over a period time, despite having the cocked dice. They tend to trade of variance for EV... if you want more variance, you pay more of a cost.
"Then you can admire the real gambler, who has neither eaten, slept, thought nor lived, he has so smarted under the scourge of his martingale, so suffered on the rack of his desire for a coup at trente-et-quarante" - Honore de Balzac, 1829
goatcabin
Joined: Feb 13, 2010
• Posts: 664
February 4th, 2011 at 4:31:19 PM permalink
Quote: thecesspit

My point is 1.41% EV BY itself does not tell you the profit of a table. They care about the total expected value IN dollars.

The variance of the pass line bet is ~1.

The variance of the odds bet is 1.55
A x5 odds bet there has a variance of 7 times the variance of a pass line bet, and x100... 150 time the pass line bet.

The variance on the odds starts to swamp that of the pass line bet... meaning that while the casino has adequate cash reserves it could take much much longer for it to settle towards the long term value (of \$0).

Say 30 passes per hour, 24 hours per day = 720 decisions on the pass line bet (I'm approximating away a whole set of things here like the fact the pass line resolves more often than the odds bet)

\$1 on the line, \$5 on the odds (x5 odds)

EV on pass line = 720 * 1.41% = \$10.52
Standard Deviation = sqrt(720) * 1 = \$26.83

EV on the odds = 720 * 0 = \$0.0
Standard Deviation = sqrt(720) * sqrt(1.55) * 5 = \$167.

As these are closely related the casino will made \$10.52 +/- \$387.66 (covering 2 standard deviations).

Over 365 days :

EV on pass line = 262,800 * 1.41% = \$3705.48
Standard deviation = sqrt(262,800) * 1 = \$512

EV on the odds = \$0
Standard deviation = sqrt(262,800) * sqrt(1.55) * 5 = \$3191

=> casino makes \$3705 +/- \$3703

=> There a chance (small though it is) with 5 times odds that the table would not post a profit from play with 5 times odds over the year.

As the odds goes up, there's more of a chance over the years play that the table is not profitable. I'm sure the minimum threshold for a table is more than "doesn't make loss" but is \$x per day, then you can see that the casino will be risking that the variance will hurt it's bottom line, even in a game it has the edge in.

(I've approximated a lot here, and may have made some glaring error)

Not bad. For \$1 pass, 5X odds, the ev is -\$.01414, the SD \$5.891, so for 262,800, ev is -\$3716, SD \$3020. The aggregate players need to be lucky only to the extent of 1.23 SD to break even with the casino. That's a probability of just about .11. If the odds multiple were 100X, SD would be \$51,670 for 262,800 bets, same ev. That's about a .47 probability that the variance on the odds would wipe out the passline ev, but just for those players that bet the minimum and took 100X odds.

Quote: thecesspit

Variance is what CAN make some punters winners over a period of time. The same variance can mean a casino loses over a period time, despite having the cocked dice. They tend to trade of variance for EV... if you want more variance, you pay more of a cost.

That's what's so special about the odds bets -- the variance is free if you're willing to risk a lot of money.
Cheers,
Alan Shank
Woodland, Ca
Cheers, Alan Shank "How's that for a squabble, Pugh?" Peter Boyle as Mister Moon in "Yellowbeard"
thecesspit
Joined: Apr 19, 2010
• Posts: 5936
February 4th, 2011 at 4:55:27 PM permalink
Quote: goatcabin

That's what's so special about the odds bets -- the variance is free if you're willing to risk a lot of money.

I saw some of your workings when looking to do this (elsewhere), but went back to rough first principals. Thanks, glad it was in the ball park.

I get that it's potentially good for a punter putting up a fair amount... which is rather my point... if it's good for the punter, it may not be as good for the casino who sees each and every table as profit centre and wants a nice stable income stream, not something that varies wildly... hence they may want to keep the odds down, despite it being "free".

As ever, EV is not the whole story, as much as people may want to propose that it is all there is to care about. Not only "what do I expect to have in my wallet" but also "how close am I likely to be to that expectation".
"Then you can admire the real gambler, who has neither eaten, slept, thought nor lived, he has so smarted under the scourge of his martingale, so suffered on the rack of his desire for a coup at trente-et-quarante" - Honore de Balzac, 1829
pacomartin
Joined: Jan 14, 2010
• Posts: 7895
February 4th, 2011 at 5:06:34 PM permalink
Quote: Nareed

With those numbers, the large space required by a craps table, and given how many people it takes to deal craps, I wonder they keep the game at all.

It's not the most profitable game in the casino, but a craps table usually brings in 3 to 4 times as much as a blackjack table. The problem is if you go through the casino and eliminate the less profitable games, you tend to destroy the atmosphere of the place. In the outlying areas, sometimes they us "tubs" for craps tables. The game is much slower, but a single person can run the game.

They hardly seem worth it in Mesquite, but it is part of the atmosphere.

AREA \$K /day/table Craps tables \$K per day
Total Strip \$3.33 201 \$669.1
Total Non Strip \$1.75 130 \$227.2
Downtown \$2.09 37 \$77.3
Boulder \$1.89 26 \$49.1
Balance of C. County \$1.76 36 \$63.2
Laughlin \$1.31 16 \$21.0
Mesquite \$1.15 4 \$4.6
North Las Vegas \$1.09 11 \$12.0
goatcabin
Joined: Feb 13, 2010
• Posts: 664
February 4th, 2011 at 6:25:02 PM permalink
Quote: pacomartin

It's not the most profitable game in the casino, but a craps table usually brings in 3 to 4 times as much as a blackjack table. The problem is if you go through the casino and eliminate the less profitable games, you tend to destroy the atmosphere of the place. In the outlying areas, sometimes they us "tubs" for craps tables. The game is much slower, but a single person can run the game.

They hardly seem worth it in Mesquite, but it is part of the atmosphere.

AREA \$K /day/table Craps tables \$K per day
Total Strip \$3.33 201 \$669.1
Total Non Strip \$1.75 130 \$227.2
Downtown \$2.09 37 \$77.3
Boulder \$1.89 26 \$49.1
Balance of C. County \$1.76 36 \$63.2
Laughlin \$1.31 16 \$21.0
Mesquite \$1.15 4 \$4.6
North Las Vegas \$1.09 11 \$12.0

Where do you get all that information?
Cheers,
Alan Shank
Woodland, CA
Cheers, Alan Shank "How's that for a squabble, Pugh?" Peter Boyle as Mister Moon in "Yellowbeard"
pacomartin
Joined: Jan 14, 2010
• Posts: 7895
February 4th, 2011 at 7:38:39 PM permalink
Quote: goatcabin

[Where do you get all that information?
Cheers,
Alan Shank
Woodland, CA

Pretty detailed information on Gaming Commission Website. They give sums for the last month, the last quarter, and the last year. I divide by 365 to get current up to date per day and per table informaion.

Data can be sketchy in places. Number of tables does not distinguish between tubs and full size craps tables (I don't think). HOWEVER, THERE ARE NOT THAT MANY TUBS.

Since the recession table games have outearned slots on the strip. That may be related to more competition rather than the recession , per se.

Table games in PA make an average of \$2K per day and up to \$2.6K per day in ones near NJ border. No distinction between craps and BJ tables. Of course there RE 5,527 TABLES IN NV andonly 609 in PA.