SD/EV for bottom carding, sequencing & holecarding

M

matt21

Well-Known Member
#1
#1
Some months ago I put together a spreadsheet model trying to calculate the EV and standard deviation associated with bottom carding. I think I arrived at reasonable answers, but it would be great to have some confirmation. Also I would be very interested in quantifying EV and SD for sequencing and holecarding.

Has anyone done any work on this, or could they point me in the right direction for appropriate reading?

Thanks in advance,
Matt21

PS I would be happy to share my completed modelling work re bottom carding.
 
K

k_c

Well-Known Member
#2
#2
matt21 said:
Some months ago I put together a spreadsheet model trying to calculate the EV and standard deviation associated with bottom carding. I think I arrived at reasonable answers, but it would be great to have some confirmation. Also I would be very interested in quantifying EV and SD for sequencing and holecarding.

Has anyone done any work on this, or could they point me in the right direction for appropriate reading?

Thanks in advance,
Matt21

PS I would be happy to share my completed modelling work re bottom carding.
Click to expand...
You might want to take a look at (Dead link: http://www.advantageplayer.com/blackjack/forums/bj-main/webbbs.cgi?read=26526) _this thread_.
 
M

matt21

Well-Known Member
#3
#3
k_c said:
You might want to take a look at (Dead link: http://www.advantageplayer.com/blackjack/forums/bj-main/webbbs.cgi?read=26526) _this thread_.
Click to expand...
thanks k_c.

Possibly to clarify with 'bottom card' I was meaning the situation where the player somehow gets to see the card that is on the bottom of the deck prior to the deck being cut, and the player then cuts the deck so that he can steer the card to the player/dealer as he wishes.

I calculated the SD and EV on the basis of the players 52-card cut being off by +/- 1 card, the player's bet size, as well as the EV from playing the rest of the shoe.

I am also interested in constructing models for calculating the EV/SD for various casino promotions. But rather than re-inventing the wheel I was trying to figure whether someone had done some of this work already, and might be willing to share some parts of it. :)
 
J

johndoe

Well-Known Member
#5
#5
moo321 said:
With holecarding, EV and SD aren't much of an issue. It's about finding the game and not getting caught.
Click to expand...
I disagree; depending on how much information you're getting (and how aggressive you are) the EV will vary a great deal, as will the SD. You can still very easily blow your bankroll by overbetting.

Both CAA and QFIT's book covers this pretty well, and CVDATA should be able to sim it.
 
M

matt21

Well-Known Member
#6
#6
johndoe said:
I disagree; depending on how much information you're getting (and how aggressive you are) the EV will vary a great deal, as will the SD. You can still very easily blow your bankroll by overbetting.

Both CAA and QFIT's book covers this pretty well, and CVDATA should be able to sim it.
Click to expand...
Johndoe, thanks for pointing out those resources - I think I will ahve some intellectual challenges with this exercises.
 
moo321

moo321

Well-Known Member
#7
#7
johndoe said:
I disagree; depending on how much information you're getting (and how aggressive you are) the EV will vary a great deal, as will the SD. You can still very easily blow your bankroll by overbetting.

Both CAA and QFIT's book covers this pretty well, and CVDATA should be able to sim it.
Click to expand...
Well, yeah, if you're getting only a paint/no paint read you should bet a lot less...
 
M

matt21

Well-Known Member
#9
#9
Well, I managed to reconstruct a probability model for exploiting situations where the dealer exposes the bottom card when offering the cards to the player(s) for cutting the deck. Somehow I had lost the original file!

I was now going to proceed to build a model for calcualting the SV/ED for ace tracking and ace sequencing.

This got me thinking - and I apologise if I am stating some real obvious things here - but I ain't much of an advanced player (so far!) - with both tracking/sequencing and bottom carding we are effectively trying to do the same thing - e.g. steer an ace to our hand, or sometimes a bad card to the dealer, the only difference between the two approaches being the manner in whcih we go about it - in one instance we exploit the bottom card being exposed, in another case it is remembering a sequence of cards and understanding the dealer's shuffle. Is that a fair point of view or am I misunderstanding something?

Presumably tracking/sequencing is better because we can potentially steer an ace in every single shoe, where as with bottom card, only 1/13th of the time can we expect the bottom card to be an ace?
(and it seems that with bottom carding fully 2/3rds of the EV comes from steering X's and T's, and only 1/3rd from steering 2-9's)

Any thoughts?
 
M

matt21

Well-Known Member
#10
#10
well it's a bit of a shame that not many people seem to be interested in this probability modelling stuff - maybe this highlights my insane passion for models, maths and theory? I just like to assess and evaluate the likely profitability and likely fluctuation associated with particular strategies before embarking on investing loads of hours practicing it all.

in any case, I think I have pulled together somewhat of a EV and SD model for bottom carding, tracking and sequencing now. :grin:

Good luck at the tables.

Matt21
 
K

k_c

Well-Known Member
#11
#11
matt21 said:
well it's a bit of a shame that not many people seem to be interested in this probability modelling stuff - maybe this highlights my insane passion for models, maths and theory? I just like to assess and evaluate the likely profitability and likely fluctuation associated with particular strategies before embarking on investing loads of hours practicing it all.

in any case, I think I have pulled together somewhat of a EV and SD model for bottom carding, tracking and sequencing now. :grin:

Good luck at the tables.

Matt21
Click to expand...
OK I'll bite. I try to use combinatorial analysis as much as possible to solve problems. To compute what happens when bottom card is cut into play would require setting some sort of probability on the proficiency of the person doing the cutting so that is the complicating factor

These are at least some of the considerations for bottom carding.

Code: 
1) If you have the cut then there are 2 choices - cut the bottom card out
of play or cut it into play. Cutting it out of play is easy to figure. Cutting
it into play depends upon the proficiency of the person doing the cutting.

2) If bottom card is a low card
    a) If it is cut out of play then player's EV=EV(full shoe)+EOR(bottom card)
       (EOR stands for effect of removal). Low card out of play=+EV, OK choice
    b) If card is cut into play then you'd probably want to make it dealer's
       up card.
       If you are 100% successful in making it dealer's up card then EV is
       known by combinatorial analysis.
    c) If you are less than 100% successful in cutting it to dealer's up card
        i) Dealer's up card will be conditional probability of not being the
           bottom card when unsuccessful, which will increase house edge.
        ii) Depending upon where player is sitting his probability of having the
            bottom card in his hand needs to be considered.
        iii) EV depends upon proficiency of cutter

3) If bottom card is a high card
    a) If it is cut out of play then player's EV=EV(full shoe)+EOR(bottom card)
       (EOR = effect of removal). High card out of play=-EV, poor choice
    b) If card is cut into play then you'd probably want it your hand.
       If you are 100% successful in steering it to your hand then EV is known
       by combinatorial analysis.
    c) If you are less than 100% successful in steering it to your hand
        i) Player's hand EV will need to be conditioned on the bottom card not
           being in his hand when unsuccessful. This will worsen player's EV.
        ii) Depending upon where player is sitting the probability of dealer
            having the bottom card as up card needs to be considered.
        iii) EV depends upon proficiency of cutter
 
M

matt21

Well-Known Member
#12
#12
thanks for your reply k_c. It's great to have someone else share their view - particularly when their perspective of the question/issue is quite different. I think I was quite narrow-minded in the type of model that I developed, following the path that I used with previous models.

Your points made me aware of some other angles for this situation.

Many thanks!
 
You must log in or register to reply here.
Top