Double Down on Soft 12

[Cacarulo]

For 6 decks HiLo these are my approximations for around mid-shoe (155 cards.) I just use best strategy.:

S17
AA v 3 (dbl v hit no ace removed)
RC=23 dbl: 13.514%, hit: 13.238% diff: .276%
RC=22 hit: 12.910%, dbl: 13.054% diff: -.144%
interpolate: RC=22.34, TC=7.5

AA v 3 (dbl v hit 1 ace removed)
RC=22 dbl: 13.572%, hit: 13.241% diff: .331%
RC=21 dbl: 12.972%, hit: 13.059% diff: -.087%
interpolate: RC=21.21, TC=7.1

H17
A-A v 3 (dbl v hit no ace removed)
RC=22 dbl: 13.304%, hit: 13.186% diff: .118%
RC=21 dbl: 12.699%, hit: 13.001% diff: -.302%
interpolate: RC=21.72. TC=7.3

A-A v 3 (dbl v hit 1 ace removed)
RC=21 dbl: 13.347%, hit: 13.185% diff: .162%
RC=20 dbl: 12.747%. hit: 13.003% diff: -.256%
interpolate: RC=20.61, TC=6.9

k_c

Hi k_c,

Your approximations are very good despite not taking pen into account.
These are my numbers:

S17
AAv3    = +7.4
AA(A)v3 = +7.0

H17
AAv3    = +7.2
AA(A)v3 = +6.8

Sincerely,
Cac

[Gronbog]

Yes, the indices of 14v3, 13v3 and 12v3 do not vary with the extra ace removed. Therefore the index of AA(A)v3 still gives me +7.

OK, but what I'm trying to clarify is whether the plays associated with those indices were made while calculating the index for A,A(A)v3. In particular, for H17, did the calculation assume standing on 12v3 at TC>=+1 after hitting A,A(A)v3 ?

[Cacarulo]

OK, but what I'm trying to clarify is whether the plays associated with those indices were made while calculating the index for A,A(A)v3. In particular, for H17, did the calculation assume standing on 12v3 at TC>=+1 after hitting A,A(A)v3 ?

Yes.

Cac

[k_c]

Hi k_c,

Your approximations are very good despite not taking pen into account.
These are my numbers:

S17
AAv3    = +7.4
AA(A)v3 = +7.0

H17
AAv3    = +7.2
AA(A)v3 = +6.8

Sincerely,
Cac

The only simple method I'd have to use TC based on game pen is to use the TC at pen/2. Additionally since I use RC after player hand is dealt but before up card the TC is for pen/2 - 1. I do this because that's how my CA works. So for 6 decks dealt 4.5/6 = 234/312 I might approximate TC at 234/2 - 1 = 116 cards.

I think this might be better for more decks. If I use this method to compute generic insurance assuming deal to last card (HiLo) to get an example for comparison -

single deck, TC=1.39 with 25 cards remaining
8 decks, TC=3.10 with 207 cards remaining

Above insurance TC is computed without specifically removing an ace, then incorporating ace into RC before computing TC.

It may not be perfect but it is relatively simple. I would be worried that if I tried to weight TCs across differing cards remaining I might be in danger of falling off the edge of the earth.

k_c

[Jack Jackson]

I was looking at a new online blackjack game, and their basic strategy said to double down on a soft 12 against a 5 or 6. However, it also said to split Aces. I suppose it would be possible to have a soft 12 if you had a lot of Aces in your hand. Anybody ever hear of such a thing?

More than likely, and noting the fact that you mentioned that this was a "new" online site they simply got confused between S12 and S13 esp. since it says to DD against a 5 or 6...