Bet Spread for Shoe Games

Pelerus

Well-Known Member
#1
My question concerns the best bet spreads for a multi-decker with wonging out at a true count of -1. It is my understanding that when playing all counts, the bet spread at a shoe game must be a minimum of 1-12 and preferably 1-16 or, some have even said, as high as 1-20. Given a large enough bankroll, this strategy is not a problem, but with even moderate bankroll constraints a spread of 1-20 at a $10 or $15 game becomes unthinkable.

An alternate strategy, backcounting until a positive count has been reached before playing, has its own share of problems, including being ineffective in crowded conditions, drawing scrutiny from casino staff, and often demanding very long periods of waiting before a positive count is reached.

The middle ground between these two tactics is playing at a zero count but not at any negative counts, in order that one may play off the top. Using this strategy, the bet spread need not be as high as for play all. However, just how high does it need to be? Will 1-4 suffice, as it does for single deck games, or does wonging out not produce that much advantage, and the spread must be higher?

N.B. The playing conditions at my casino of choice, Foxwoods, are as follows: S17, DAS, DOA, S4, late surrender, penetration uncertain (somewhere between 75% and 85%). Under those rules, the off the top house advantage is 0.357%. The question, however, may also be generalized to any multi-deck game.
 

BJStanko

Well-Known Member
#2
The only safe way for you to know how much you have to spread your bets and how much to bet according to count and your bankroll is to consult one of Blackjack Sim. Softwares. If you play often $150 for software is nothing because it will save you enormous amount of time when analyzing game and game conditions.

For game conditions you are playing in I think you would have 0.72% advantage if you spread 1-4 and leave when your advantage drops below
-0.5%.


-1.0% no bet
-0.5% bet 1 unit
0.0% bet 1 unit
0.5% bet 2 unit
1.0% bet 4 unit
1.5% bet 4 unit
2.0% bet 4 unit
2.5% bet 4 unit

This is just aproximation!!
 
#3
Revision

BJStanko said:
-1.0% no bet
-0.5% bet 1 unit
0.0% bet 1 unit
0.5% bet 2 unit
1.0% bet 4 unit
1.5% bet 4 unit
2.0% bet 4 unit
2.5% bet 4 unit
This is just aproximation!!
I would at least do this -

-1.0% no bet
-0.5% bet 1 unit
0.0% bet 1 unit
0.5% bet 2 unit
1.0% bet 4 unit
1.5%+ bet 4+4 unit

- zg
 

Pelerus

Well-Known Member
#4
A question regarding both BJStanko's and zengrifter's responses: how did you determine how many units to bet at each count?

If I were to devise a betting schedule for a 1 to 4 spread, I would first take into account that the max bet occurs at a true count of +5. (That is what I have heard should be the point at which the betting ramp reaches its peak, but perhaps this is not the case in all games.) Given that one is betting
$10 * 4 = $40 at +5
and an advantage of (5-1)/2 + (0.5-0.357) = 2.143,
the proper Kelly bankroll is $40/(2.143*0.0076) = $2456.
After quantifying the bankroll, one can determine the bet at a count of
+4 as $2456*1.643*0.0076 = $31,
+3 as $2456*1.143*0.0076 = $21,
+2 as $2456*0.643*0.0076 = $12
+1 and 0 as less than $10, but bets of $10 will be placed at those points.
Therefore, the betting schedule according to Kelly is:
-1 no bet
0 bet 1 unit
1 bet 1 unit
2 bet 1 unit
3 bet 2 units
4 bet 3 units
5 bet 4 units

Also, a last question for zengrifter: is your terminology of 4+4 units just the same as 8 units, or is there additional meaning?
 
#6
Pelerus said:
Therefore, the betting schedule according to Kelly is:
-1 no bet
0 bet 1 unit
1 bet 1 unit
2 bet 1 unit
3 bet 2 units
4 bet 3 units
5 bet 4 units
Kelly is only part of the consideration - additionally we must factor in distributed frequencies of rising count levels- and overall risk-flux.

Therefore, the bet 'ramp' that you propose can be 'quickened' at the lower-end by not betting untill the TC is +1 and then ramping it faster, and then choking it off at +5. zg

0 bet 0 unit
1-2 bet 1 unit
2-3 bet 2 unit
3-4 bet 2+2 unit
4-5+ bet 3+3 units

Initially you did not define the bank. zg
 

Pelerus

Well-Known Member
#7
Thank you for the replies. Zengrifter, if I understand you correctly then, one can 'adjust' Kelly by betting more than Kelly at certain points if one also bets less than Kelly (or less than one would have bet) at other points. In the example you gave, you bet more than Kelly at a count of +2, for example, but you bet less than my proposed bet at a count of 0.

I also want to give a specific bankroll and my considerations for handling it with an 8 deck shoe. I will be playing with an $8000 bankroll and the rules specified above and wonging out at -1. I am willing to bet somewhere between full Kelly and one half Kelly, with the former implying a 1-13 spread ($8000*2.143*0.0076 = $130) and the latter a 1-6.5 spread. With these conditions, what do you recommend as the best possible betting schedule, given that due to risk aversion I would like to bet as close to one half Kelly as possible.

(N.B. If you give a bet schedule with multiple hands being played, could you also include one with only single hands played, due to constraints in crowded conditions.)
 
Last edited:
Top