gordonm888
gordonm888
Joined: Feb 18, 2015
  • Threads: 38
  • Posts: 2481
September 30th, 2019 at 9:41:09 AM permalink
Quote: michael99000

If thatís what the dealer is doing, always showing a 10 if either of his cards is a 10, then that means he shows a 10 about 64% of the time.

Itís hard to believe people would keep playing in a game where they face a dealer 10 that often.



There would be no, or very little, advantage to the casino to show a Ten when the other card is a 7,8 or 9. If you were designing a cheating system you would probably only turn over a Ten when the other card is a 5 or 6. And maybe a 4.

If you combined that tactic with using a deck where you have removed 3/16 or 4/16 of the Ten cards -as suggested by the post above - then it would be harder for players to discern that either tactic was being used.

Edit: Someone should work out house edge for this case:

Rules as OP states, 6 Decks, S17, 3-2, DAS, DOA, Surrender, Resplit to 4 hands except Aces, Dealer peeks on Ace but not on Ten.
Remove 3/16 of Tens
When dealer is dealt a T-6 or T-5, he shows the Ten.
********************************************
Last edited by: gordonm888 on Sep 30, 2019
So many better men, a few of them friends, were dead. And a thousand thousand slimy things lived on, and so did I.
michael99000
michael99000
Joined: Jul 10, 2010
  • Threads: 9
  • Posts: 2106
September 30th, 2019 at 9:58:06 AM permalink
Quote: gordonm888

There would be no, or very little, advantage to the casino to show a Ten when the other card is a 7,8 or 9. .



Dealer has a 10 and a 7 for 17, I have 9. Anytime the dealer shows his 7 Iím doubling. If he always shows his 10 in that case, then Iím never getting the chance to double my 9 against him
gordonm888
gordonm888
Joined: Feb 18, 2015
  • Threads: 38
  • Posts: 2481
Thanks for this post from:
aruzin
September 30th, 2019 at 11:05:58 AM permalink
Quote: michael99000

Dealer has a 10 and a 7 for 17, I have 9. Anytime the dealer shows his 7 Iím doubling. If he always shows his 10 in that case, then Iím never getting the chance to double my 9 against him



I agree, though I would put that in the category of "very little edge." (And if some Tens are removed from the deck then the decision changes anyway, such that it is best to Hit 9 vs 7.) The big edge comes from when dealer has 10 and 6, or 10 and 5 and shows only the 10 - player will bust with frequency of 34% and split and double less often against the dealers 10-6 and 10-5.

If you were a crooked casino designing a crooked BJ game that gamblers would have a hard time detecting (on an Island where most of your customers are tourists who may never come back) what would you do?

*******************************
Rules as OP states, 6 Decks, S17, 3-2, DAS, DOA, Surrender, Resplit to 4 hands except Aces, Dealer peeks on Ace but not on Ten.
Cheat 1: Remove 3/16 of Tens
Cheat 2: When dealer is dealt a T-6 or T-5, he shows the Ten.
**************************************
Without the cheats, the house edge is < 0.4%
My "back of envelope" calculation is that the 2 cheats would result in a house edge of roughly 2.8% when player plays basic strategy.

If the two cheats were: Remove 4/16 of Tens, and always show the Ten when dealer has T-6, T-5, or T-4, I roughly estimate a house edge of 4%, given that player uses basic strategy.
So many better men, a few of them friends, were dead. And a thousand thousand slimy things lived on, and so did I.
aruzin
aruzin
Joined: Jan 31, 2014
  • Threads: 2
  • Posts: 26
October 1st, 2019 at 7:04:49 AM permalink
Quote: gordonm888


If the two cheats were: Remove 4/16 of Tens, and always show the Ten when dealer has T-6, T-5, or T-4, I roughly estimate a house edge of 4%, given that player uses basic strategy.




Very useful info, thanks. Now I have to try and see how many (minimum) hands one needs to play to approach the only 4% losses to casino... !?

I think this was discussed in another thread, but when stats are quoted about house edge, you think you should walk away with loosing just 4% of your money, but this is if you play (large) X number of hands - due to various STD and other factors, something always hard to grasp.

So if I play 100 hands I could double my money or loose 100%. But if I play say 100,000 hands or a million I should approach that 4%. I have seen some numbers in another thread, but what is your take of what that number is (approx.)... !?
DogHand
DogHand
Joined: Sep 24, 2011
  • Threads: 1
  • Posts: 377
Thanks for this post from:
aruzin
October 2nd, 2019 at 8:47:02 AM permalink
Quote: aruzin

Quote: gordonm888


If the two cheats were: Remove 4/16 of Tens, and always show the Ten when dealer has T-6, T-5, or T-4, I roughly estimate a house edge of 4%, given that player uses basic strategy.




Very useful info, thanks. Now I have to try and see how many (minimum) hands one needs to play to approach the only 4% losses to casino... !?

I think this was discussed in another thread, but when stats are quoted about house edge, you think you should walk away with loosing just 4% of your money, but this is if you play (large) X number of hands - due to various STD and other factors, something always hard to grasp.

So if I play 100 hands I could double my money or loose 100%. But if I play say 100,000 hands or a million I should approach that 4%. I have seen some numbers in another thread, but what is your take of what that number is (approx.)... !?



aruzin,

According to gordonm888, your Expected Value per hand is -4%, or -0.04. This means that if you play "n" hands, your total EV will be -0.04*(n).

For flat-betting with Basic Strategy, BJ has a Standard Deviation per hand of about 1.15. This means that if you play "n" hands, your total SD will be 1.15*(n)^0.5. Approximately 95% of the time, your actual result will be within 2SD of your EV.

Let's play with some numbers to illustrate.

If you play 100 hands, your EV is -0.04*100 = -4 units. Your SD is 1.15*100^0.5 = 1.15*10 = 11.5, so 2SD's is 23. This means that, with 95% certainty, your actual result will be between (-4-23) and (-4+23), so in the range of -27 to +19 units. If you are flat-betting $10 per hand, that's -$270 to +$190. Thus, you have a reasonable chance to be ahead after 100 hands.

If you play 10,000 hands, your EV is -0.04*10,000 = -400 units. Your SD is 1.15*10,000^0.5 = 1.15*100 = 115, so 2SD's is 230. This means that, with 95% certainty, your actual result will be between (-400-230) and (-400+230), so in the range of -630 to -170 units. If you are flat-betting $10 per hand, that's -$6,300 to -$1,700. Thus, you will almost certainly be losing after 10,000 hands.

Now you try! What will your 95% confidence range be if you play one million hands at $10 per hand? Hint: it won't be pretty ;-)

Hope this helps!

Dog Hand
aruzin
aruzin
Joined: Jan 31, 2014
  • Threads: 2
  • Posts: 26
October 2nd, 2019 at 10:04:41 AM permalink
Quote: DogHand


Now you try! What will your 95% confidence range be if you play one million hands at $10 per hand? Hint: it won't be pretty ;-)

Hope this helps!

Dog Hand



Thanks for the explanation, I think it is becoming clear where I was getting confused. I always thought about the 4% house edge or -EV in terms of my own bank roll and not the actual hands value played... !

Say I go to holiday here or Vegas and want to budget $1000 to play BJ in casino. I always thought if I play well and according to house edge I should loose -0.04 * 1000 = -$40.... and I only achieve that if I play enough hands in order to get close to that - because otherwise in the first night I could just loose -$500 but the next night might win, but if played enough I should loose approximately the -EV.... !!!

I see now that this is totally wrong, I need to take into account the total value of hands I play for -EV, so if I play 1000 hands of $10 each, then I am expected to loose -0.04 * 1000 * 10 = -$400, and if play 10000 hands, then -$4000 (that of course never happens because I have already lost my bank roll way before that !).

But what do you guys really consider your "n" to be, on one night in casino, on your two weeks holiday, per year, over the course of your life time ... !!!?

I was also thinking how to calculate what is a good amount to take to play for one night for say a $10 table (considering good/bad streaks can happen) - I have seen many who walk in with their credit card and look so miserable when leaving, loose a lot more than they wanted/could afford. I always go with a fixed amount I plan to loose in my pocket and no more, it is mainly for entertainment and some fun, my moto is to loose X amount on a night or double it to 2X and walk out - it's not ideal but works for me, sometimes... ! lol
Last edited by: aruzin on Oct 2, 2019
dddkkk1
dddkkk1
Joined: Oct 11, 2019
  • Threads: 3
  • Posts: 12
Thanks for this post from:
aruzin
October 19th, 2019 at 8:34:00 PM permalink
Quote: aruzin

... what if they peek or somehow get extra info about the cards and act accordingly, now that is simply nasty....


Yes it is terrible if they are cheating, below is probably the worst hand the dealer can do this to you on, but I haven't checked doubles or splits.

For this game the cost of hitting a 16 (10+6) against a dealer 16 (10+6) is 59.34%, because if the player stands then the EV is +22.08% and if the player hits then the EV is -37.26%.
Also, if the dealer only cheated on the hand stated above, "player 10+6" vs "dealer 10+6", it would add 0.064% to the house edge for a basic strategy player.
The working out I used for the figure above is as follows:
"Chance of Player 10+6" x "Chance of Dealer 10+6" x (-59.34%)
= ((96/312 x 24/311) x 2) x (95/310 x 23/309) x (-59.34%) = -0.064%

Note: I know the true chance of a dealer 10 + 6 is ((95/310 x 23/309) x 2) in the above scenario, but the chance of a non-cheating dealer getting a 10-up hand in the above scenario is 50%, so that is why i did the working out like that.

Lastly, it seems other replies have already worked out the EV for a "cheating dealer" for this game already so I probably won't analyze this any further.
Last edited by: dddkkk1 on Oct 19, 2019
aruzin
aruzin
Joined: Jan 31, 2014
  • Threads: 2
  • Posts: 26
November 4th, 2019 at 10:05:49 AM permalink
In our dodgy casino I noticed there is still one dealer who shows up different card as dealer card (sometime her first and sometime second), no idea why, but I stop play when she comes, other 3 or 4 dealers don't, so I play with them.

But from a few weeks ago they started a new rule, they peek the whole card for BJ when showing 10s. This was done in another casino but not here, but they started it now. They used to just peek under A... !

I was wondering why they started this, I am sure it benefits the odds of the casino, they never do anything to benefit players... ! lol

So this is a 6 deck game, pays 3/2 on BJ, accepts insurance and pays 2/1 for A and stays on all 17s. Can surrender any hand (not against A) and split any hand and as many times (for AA only can split once).

In old days if you doubled down (say your 11, or could for any hand) against dealer 10 and he flipped a BJ you only lost half the wager - so not much of an advantage to anyone now that they are peeking at hole card under 10s, because this scenario never happens.... (we lose the same amount when dealer has BJ when showing 10s).

But you can only surrender now before dealer peeks on 10s, most guys are not paying attention and/or are not quick, so maybe that is where the casino get an advantage.... !? Any idea how much advantage that adds to house edge/EV and do you need to change basic strategy because of this new peek... ?

Also I guess another advantage to casino is that it speeds up the game, once he peeks on 10s and has a BJ everybody loses and no need to waste time and play the hands anymore (unless if you have BJ too, which will be no loss). This also could lead to other issues, as dealer behavior can change once he knows what he has, in a corrupt environment he can help a player or otherwise, I tried to observe them afterwards for any changes or look at their eyes when they peek and see if anything can be read ...! (this was only available for As peek before, now more common).

(Also dealers cut the deck to 50% or lower, so most card countings are pretty useless on their table.)

  • Jump to: