| |
Parker Pages: Blackjack Main
Re:Card-Steering.....some numbers...(long)
Posted By: Adam N. Subtractum In Response To: Re: --MORE INFO PLEASE!!!!!!!!-- (suicyco maniac)
Date: Wednesday, 1 January 2003, at 8:02 a.m.
Let's look at some numbers relevent to our situation. These are from _The World's Greatest Blackjack Book_, and are somewhat dated, as well as for single deck, but should prove sufficient.
-Player Advantage vs. Dealer Upcard-
Upcard,Bust%,BSadv%
2 = 35.30 = 09.8
3 = 37.56 = 13.4
4 = 40.28 = 18.0
5 = 42.89 = 23.2
6 = 42.08 = 23.9
7 = 25.99 = 14.3
8 = 23.86 = 05.4
9 = 23.34 = -4.3
T = 21.43 =-16.9
A = 11.65 =-36.0Now the first thing you should note here is that a 7 gives us a higher advantage than a 2 or 3 (46% & 6.7% higher, respectively) despite the fact that the dealer busts more frequently with the 2 & 3 (36% & 44.5% more, respectively). So, naturally, we will steer that card to the dealer as well.
Now as for the case of the 8's, a 5.4% edge is nothing to sneeze at, but there are important factors to take into consideration when deciding if we should steer this card to the dealer. These key factors are the accuracy of our cut, and the control we are able to have over the play of the hands leading up to our "steered" card. As this method is usually only used when head-up or with a team (to "lock-up" the table), the latter is usually not a problem assuming you have practiced at home and are somewhat experienced with the technique. But, its not like the good ol' days, when you could just cut 17, spread to 7 hands and sit back and watch the dealer bust, lol. Now its commonplace for the cut to have to be at least one deck from the ends of the pack, so intelligent play of the prior hands is necessary, as well as a count of cards played, to assure the dealer (or player) will receive the "steered" card. Near perfect accuracy can be achieved with proper practice.
The cut itself is the hard part, as any amount of inaccuracy will deflate our advantage proportionately. Take for example the following situations steering an 8 to the hole:ie.I-Cut accuracy-
Exact = 100%Advantage = 5.4% * 100% = 5.4%
In order to achieve our full edge, this example makes the unrealistic assumption that we can cut with 100% accuracy. Obviously not feasable in real world play.
ie.II-Cut accuracy-
Exact = 50%
offx1 = 30%
offx2 = 20%Advantage = 5.4% * 50% = 2.7%
This is a more realistic assumption of what kind of accuracy is attainable by the average player (assuming min 1 deck cut from ends, much higher accuracy will be attainable w/o this limitation) with only moderate practice. Note here we still have a healthy advantage of 2.7%
ie.III-Cut accuracy-
Exact = 20%
offx1 = 40%
offx2 = 20%
offx3 = 15%
offx4 = 5%Advantage = 5.4% * 20% = 1.08%
This is what could be expected from an inexperienced, sloppy cutter. Now subtract the house edge for say, an average six decker (remember, our numbers were for a break-even SD game), and we get: 1.08% - .40% = .68%, not so impressive anynore, huh? Remember, this is not considering any errors in play causing the card to miss the hole, so our advantage would suffer even more in that case.
So you see, depending on your accuracy the 8 may, or may not be worth sending to the dealer.Now, moving on to the 9's we see the card puts us at a 4.3% disadvantage when steered to the dealer. Conversely, when we receive a nine as our first card, we are at a .86% disadvantage (according to Cacarulo,6D/DOA/DAS/SPA1/SPL3/NS). So, what should we do? I believe the main query in your post was wether to send the dealer a good (for him) card, because we know what it is, and will be able to play against it accordingly. Intuitively, I don't think any number of plays could make up for the 4.3% disadvantage, let alone the 16.9% disadvantage of the Tens. but let's examine the situation more closely. Some numbers from Norm's Qfit site:
-Dealer hand probabilities-
Dealer's first card 9
One nine removed from deck/pack17 = 12.54%
18 = 08.71%
19 = 36.56%
20 = 12.42%
21 = 06.20%
Bst= 23.57%So the dealer will end up with a total of 19 or higher 55.18% of the time. Compare that to the normal frequency of 38.42% and you see what we're dealing with here. Now we will end up with a 19 only 32.9% (calculated from Wong's PBJ, Table D2) of the time, so we would have to hit more aggressively to increase that number. Some more numbers on that note:
-Player's probability of busting-
Hand=f%bust
17 = 69%
18 = 77%
19 = 85%Without getting into specifics it seems fairly obvious that no amount or type of play deviations will overcome such a large disadvantage (the dealer's 4.6% edge w/ a 9).
And now finally we get to the Tens. As seen in the previous example with the 9's, it clearly wouldn't be worthwhile to try and overcome the even larger edge the dealer holds with a Ten (16.9%). So we need to steer the card to our own hand(s), where according to Cacarulo, we have a 14.34% advantage (conveniently the same number for SD & 6D). Now let's begin to note the power of this technique. Let's use the same cutting accuracies as our previous examples.
ie.1a-Cut accuracy-
Exact = 50%
offx1 = 30%
offx2 = 20%Advantage = 14.34% * 50% = 7.17%
Note that in this situation, when we are steering the card to our own hand, it may be more optimal to spread to more than one hand depending on our accuracy, or the "signature" of our cut, if you will, to increase frequency of catching the ten. But this is a matter for another post...
ie.2b-Cut accuracy-
Exact = 20%
offx1 = 40%
offx2 = 20%
offx3 = 15%
offx4 = 5%Advantage = 14.34% * 20% = 2.868%
We still have an edge of over 2.4% (2.8% - .4%) with atrocious accuracy, and this doesn't include playing multiple hands, which would obviously be beneficial given the "signature" of this cut.
As you can see, this can be a very powerful technique, even when used this sloppily.Since this post turned out to be a little more extensive than originally planned, I might as well throw in some additional commentary for those new to the technique of card-steering.
Though steering can be more accurate than other methods of location play, such as card sequencing, you can only perform the maneuver once per shoe, obviously, where with sequencing, you can key-up as many cards as you are mentally capable of.
Though many casinos are lax with the viewing of the back card before the cut, many casinos are quite paranoid and will not allow this. -TIP: Using a partner can prove beneficial in situations like these. With players at the table at different positions you have a much greater chance of seeing the card* (plus the aforementioned benefits of "locking-up" the table). If the card is not visible from any spot at the table, a team member can arrive at just the right time**, approach the table at the necessary angle, and relay the card value to the Big Player.(*/** Team members would have to have other responsibilities as well to make there presence $worthwhile$. For example in the first scenario the other player(s) at the table would be shuffle-tracking and/or sequencing and/or keeping an insurance count. In the second, the "roaming" signaler would be Backcounting other shoes and/or looking for fresh shuffles when not signaling us.)
The major drawback to the technique of card-steering is that it needs to be employed without other (non-team) players at the table. Other players' play of hands will throw off your accuracy by too much to make it possible to steer the card where you want it. Obviously, finding head-up games is difficult, especially at the lower levels. The solution to this is to play higher limits, where crowding is significantly less, and No Mid-Shoe entry is more prevelent.
There are obvious drawbacks to this, most importantly the increased scrutiny and heat from the pit, as well as the eye in the sky.ANS
p.s.: We didn't get into the Aces case, but the strategy would be the same as with the tens. An important note however, is that the commonly used equation for Kelly bet size (BR*ev)/variance) is not log optimal in this case (or any case w/ extremely high edge). The number would be closer to 30% of BR, rather than the commonly used 40%, but this is whoooole 'nother matter :)
p.s.s.: Otherwise, I trust you have a good idea of how to calculate your bets for this strategy, if not, or if you'd like to optimize, feel free to ask me any more questions on this matter.
| |
Parker Pages: Blackjack Main is maintained by Viktor Nacht with WebBBS 5.00.