When to leave a shoe using an unbalanced count

[BackCounter]

Hello; I had wanted to know when to leave a shoe when using an unbalanced count such as UBZII. A rule using a fixed Wong-out number clearly will not do because we expect the running count to be lower early in the shoe. The BJA chapter 13 method gives an idealized answer but is very demanding; it is easier to calculate the expected player advantage for the rest of a shoe given the shoe position and count, and use that as a “point of departure” (!) for making the decision. In other words, if the count two decks into the shoe is only X, should I stay or should I go?

Here it is; 6 decks, H17, DAS, 4.5/6 penetration; sim was run using bet camouflage, 1-10 spread, 14 indices. Baseline player advantage is about 0.46%. Averaged over the whole shoe, the current-hand player advantage turns positive at a count of -6. The graph shows that two decks into the shoe, if the count is -18, the expected advantage for the rest of the shoe is zero and there is no reason to keep playing. If the count is -15, the expected advantage is 0.5% and we might as well stay. But three decks in, we don’t have the same advantage unless the count is -10.

advantageUbzRunning.png

The above is sort of a hybrid between using a running count and a true count. If the running count is converted to a true count, the graphs look like this:

advantageUbzTrue2.png

The same thing using Hi-Lo and the 1-12 “Rabbit” bet ramp:

advantageHiLoTrue2.png

Having calculated these, the departure points are like indices.

In case anyone is interested . . . .

[Cacarulo]

I never liked the categorization of "unbalanced counts" to refer to those where the RC is not converted to TC.
Both balanced and unbalanced counts can be converted to TC. This applies to KO (TKO) as well as UBZII (TUBZII).
With this clarification, it is best to convert the RC to TC and leave the table when it reaches "-5" or lower.
For this conversion to work, we must start the count with an IRC of -4 times the number of decks (-24).
A TC of "-5" in TKO or in TUBZII is equivalent to a TC of "-1" in a balanced system.

Hope this helps.

Sincerely,
Cac

[BackCounter]

Hello; that got me thinking; had not paid any attention to true-counting unbalanced counts before. I re-ran the sim using true counts from the start rather than just converting them for the graph; the departure counts went down a little. Plus, the win rate went up a little. The advantage goes positive at TC -2.

advantageTUbzTrue.png

[Cacarulo]

Hello; that got me thinking; had not paid any attention to true-counting unbalanced counts before. I re-ran the sim using true counts from the start rather than just converting them for the graph; the departure counts went down a little. Plus, the win rate went up a little. The advantage goes positive at TC -2.

advantageTUbzTrue.png

The first advantage should appear at -3 (which is equivalent to a TC = +1 in a balanced system).
Not only do you have to start the IRC at -24, but you also need to adjust your indices to use the same IRC.

Sincerely,
Cac

[Lester]

Leaving a shoe?
Yesterday left a table at -8rc and then left another table at -12rc. After that, the tables were packed, no spaces available. Drove an hour and 20 minutes to get there. Left at the 1 hour mark unable to reenter a table.