Special BJ rule - re-doubling/tripling

[henkel]

Hello all BJ gurus,

I have found a casino with a special re-doubling /tripling rule in Prague. I guess, that in combination with all other rules it should be slightly + EV game for the player even without counting. Would be someone so kind to help with correct BS for this game ? I know that this topic was already discussed here in the past, but I did not find any clear solution. BJ softs usually do not support calculations with this special rule - just "normal" redouble. /Spanisch 21/

The BJ rules :

a) 6 deck, S17, BJ(3 to 2), ENHC, DAS, DOA, ES10

b) Non- Ace split to maximum 4 hands , Ace split to maximum 2 hands(only 1 card to split ACE)

c) Redouble is allowed after double, for example, original bet is 1 unit, after double, there are total 2 units for that hand, player allowed to REDOUBLE this hand by placing another 1 unit to draw a card, all total bet is now 3 units(NOT 4 units !)

Thanks in advance for your help or any suggestions.

henkel

 

[Taff]

Is this a shoe game.??

 

[henkel]

yes of course - 6D with quite poor pen - about 3,5 - 4 deck

 

[Taff]

As seems to be the case with Prague. Only advantage would seem to be counting and only then at neg counts when you hit a Ace on your 11 v 5/6. Cant see anywhere where you'd have an advantage at high counts but with that rubbish pen its a non starter anyway.

 

[henkel]

the pen is really rubbish so counting is not very valuable... but I guess that even by pure correct BS playing the game should be about 0,15-0,2 + EV for the player . Some more doubling on 9/2, 8/5,6, 7/5,6 etc... and also some more doubles on soft hands should be in place - but what are really the exact correct deviations ?? thats what I want to find out - but no success for now...

 

[Taff]

but I guess that even by pure correct BS playing the game should be about 0,15-0,2 + EV for the player .

How'd you work that out??. I'm trying to find the exact figure for you.!!

 

[Taff]

I'm looking at this off the top of my head but I would apreciate anyone who could find a way of simming it. Again I would want to know info on the remainder of the deck and I think that's how your store is cutting your legs off by almost half shoe-ing you. A lot of the research on this seems to revolve around SP.21 where it gives you about .4 but the re doubles are not done in the same way so I think the edge would be considerably less.

7 vs 6
8 vs 3-6
9 vs 2-6
T vs 2-9
11vs 2-9

A2 vs 2-6
A3 vs 2-6
A4 vs 3-6
A5 vs 3-6
A6 vs 3-6
A7 vs 2-6

 

[London Colin]

A lot of the research on this seems to revolve around SP.21 where it gives you about .4 but the re doubles are not done in the same way so I think the edge would be considerably less.

Maybe not. The re-doubling strategy would be different. The less it costs to re-double, the more it can be used defensively, to lose less rather than win more, by taking an extra card that would otherwise be denied to you.

 

[Taff]

True but it's all theory unless it can be simmed and I'm stumped and looking through archive posts on the subject so too have others been.

 

[johndoe]

This is somewhat similar to a promo that was run at Mohe_gan Sun years back, where you could triple down, rather than double down. It was roughly 0.5%-1.0% off the top, if I remember, and this should be even better. Pen shouldn't be that relevant, since counting is very much secondary. WOO did a full analysis back then, with optimal strategy. See if you can find it. I might have it in my archives.

Warning: The increased doubling/tripling significantly increases variance, so be careful.

The big difference is getting that extra card, which changes the strategy quite a bit. But it's worth exploring carefully, it could be a good game.

(How many of you were at that event? It was like an AP convention! I sat there for the full 24 hours, and got to be good pals with the dude sitting next to me all day.)

 

[Taff]

Pen shouldn't be that relevant, since counting is very much secondary.

I wouldn't want to do that. On the first round, sure you're throwing both B.S and deck knowledge out the window to see what you get. But you're going to be faced with a lot of 3 hand stiff totals as a result. I would want to be armed with that knowledge before deciding whether to re-double or not. Thinking marginals like 13v2 and 14v4 etc. Overall I dont think it adds a lot as I assume you flat bet but I just dont like losing.

 

[gronbog]

would apreciate anyone who could find a way of simming it

I'm simming it now. I'm first generating the proper basic strategy and then will need to run a sim to get the house/player edge. Both are being done by simulation (I don't have my own CA). My sim is slower than CV, so please be patient.

 

[London Colin]

I was toying with the idea of hacking Eric Farmer's CA. I think it should be fairly straightforward to do.

 

[gronbog]

Independent results using different methodologies are always a good thing.

 

[London Colin]

Independent results using different methodologies are always a good thing.

Indeed. It's the prospect of being able to compare results that has spurred me on to dig a little deeper than just toying with the idea.

Unfortunately, it's not proving quite as straightforward as I had hoped. I've found myself going off on a tangent, trying to add support for early surrender in as generic a way as possible. I had already added ENHC, and it's the potential combinations of peek/no-peek and late/early surrender that are giving me some headaches.

@henkel,
Regarding this specific game, am I right in assuming that ES10 means that early surrender against the 10 is the only situation in which surrender is available? (So there's no late surrender against 2-9, for instance?)

And more generally, can anyone say if it make sense to cater for the possibility of late surrender against ace|10 in a ENHC game? (I seem to recall seeing something about a procedure for this in which a 'lammer' is used to indicate that surrender has been chosen, and then the hand is later settled once it is known whether the dealer has a BJ.)

 

[henkel]

Thanks to all members contributing to the redouble/triple topic.

It would be really great if Gronbog generate the BS and HE . To be honest, the BS is much more important to me than the exact HE - if the game is just slightly positive EV for the player, than it makes sense to play there with correct BS even with this poor pen. Its the best casino in Prague - location, services, free drinks and food - so good place for relaxing play, not a big cash cow..

To London Colin question regarding ES10:

There is no dealer HC in the game /its the same in almost all casinos in Europe/ so there is nothing to check (picking) . So if Surrender is allowed /very often rule/ than it is just an Early one. ES10 means that you are allowed Surrender against all cards except an Ace. If dealer has an A player is offered for insurance/even money. Player could potencially Surrender any non A situation - but following the BS its an advantage just in 4 cases - 14,15,16 against 10, and 16 against 9 if no counting.With counting and some more reasonable pen there is one most usual situations on the top - 15/9.

 

[Taff]

Unfortunately, it's not proving quite as straightforward as I had hoped. I've found myself going off on a tangen

I cant caculate the edge to incorporate ES10 into ENHC game. Gronbog.??

 

[London Colin]

I cant caculate the edge to incorporate ES10 into ENHC game. Gronbog.??

The figures for the additional edge are available in BJA3-
ES 0.629
ESA 0.386
ES10 0.241
LS 0.073

I presume ESA + ES10 should equal ES, and the slight discrepancy is just a rounding error.

As with my earlier question, I'm not sure whether, in this terminology, ES* includes LS (e.g., whether the 0.629 includes 0.073 from surrendering 16v9.)

 

[xengrifter]

the BS is much more important to me than the exact HE - if the game is just slightly positive EV for the player, than it makes sense to play there with correct BS even with this poor pen.

My intuition tells me you'd be better off counting the muck in a CSM.
Someone prove me wrong?

 

[Taff]

I presume ESA + ES10 should equal ES, and the slight discrepancy is just a rounding error.

That would seem to make sense as ES10 is anything up to and including 10 but feel free to point out D.S has made an error.