sunfighter: london disatvantage

[sunfighter]

What is the disadvantage at a 4d game,no split 4s,5s,10s, no soft doubles,double only 9,10,11
with no insurance allowed?Is it close to -0.75?

[Don Schlesinger]

> What is the disadvantage at a 4d game,no
> split 4s,5s,10s, no soft doubles,double only
> 9,10,11
> with no insurance allowed?Is it close to
> -0.75?

No, nowhere near that bad, if we're talking about a s17, das game, which you forgot to specify. I'd say around -0.46 or so. I can't give you a precise answer, because I don't have the positive effect (usually +0.142) of das when soft doubling and not splitting 4s vs. 5 and 6 are not permitted, but the effect is minuscule.

Note that not splitting 5s or 10s and no insurance have no effect, because the plays aren't basic strategy anyway.

Don

[Magician]

> Note that not splitting 5s or 10s and no
> insurance have no effect, because the plays
> aren't basic strategy anyway.

How much does eliminating 3 out of the top 5 index plays affect a counter's win rate (ballpark)?

[Don Schlesinger]

> How much does eliminating 3 out of the top 5
> index plays affect a counter's win rate
> (ballpark)?

I can't answer that without the exact rules, spread, etc. Even then, someone will have to run a sim.

No insurance (except even money on naturals) hurts quite a bit. Splitting tens hurts less, but can be valuable, as well.

Don

[sunfighter]

> No, nowhere near that bad, if we're talking
> about a s17, das game, which you forgot to
> specify. I'd say around -0.46 or so. I can't
> give you a precise answer, because I don't
> have the positive effect (usually +0.142) of
> das when soft doubling and not splitting 4s
> vs. 5 and 6 are not permitted, but the
> effect is minuscule.

> Note that not splitting 5s or 10s and no
> insurance have no effect, because the plays
> aren't basic strategy anyway.

> Don
What is your advantage when you play in U.k a four deck game with das,S17,no soft doubles,double only on 9,10,11,but no insurance and only even money when you have a bj and the dealer has an ace, you count with hi-low and the true count is +2?

[Don Schlesinger]

> What is your advantage when you play in U.k
> a four deck game with das,S17,no soft
> doubles,double only on 9,10,11,but no
> insurance and only even money when you have
> a bj and the dealer has an ace, you count
> with hi-low and the true count is +2?

I don't know. You'll have to sim it. In any event, for the shoe game, Hi-Lo insurance isn't right until +3, so if you want to know the counter's edge, playing the I18, at +2, for the rules you stated, I'd add about 1% to the starting disadvantage. So maybe your edge is around 0.5-0.6%.

Don

[MikeH]

> What is your advantage when you play in U.k
> a four deck game with das,S17,no soft
> doubles,double only on 9,10,11,but no
> insurance and only even money when you have
> a bj and the dealer has an ace, you count
> with hi-low and the true count is +2?

Over the course of the last week Ive been writing my own Blackjack simulator in VB. I know lots of excellent products exist to do this, but its the kind of project I really enjoy (yes Im a nerd/geek). Anyway, Ive learned an awful lot about blackjack (distribution of things, how they change with the rules etc) from doing this. So, its been edifying.

Anyway, Ive had a bash at simulating the rules you outline. To be clear, I assumed:

4 decks
S17
No splitting 4,5,10
1 card on split aces
No resplitting (been too lazy to code that up yet, perhaps one of the elders can advise what extra benefit including resplitting might provide)
DAS
D 9/10/11
No soft doubling
No insurance
No surrender
No hole card (lose all to dealer natural)

I played through 700,000 shoes with one player sitting at the table playing basic strategy (according to Wong '94 for no hole card, DAS). Thats 20.1 million hands. I assumed the dealer played through to 75% penetration so I could record the count, although the player didnt vary bet size or strategy with the count (again, that still needs to be coded up).

The edge on the game I found to be 0.569% (not sure what the standard error on that is), which sounds pretty plausible to me. I think no hole card whacks the edge up over the standard US games because it stops you doubling with 11s and 10s vs dealer A and T and limits splitting.

When the True Count was 2 (strictly greater than 2 but less than 3) I found the player had an edge of 0.696%. This figure is less reliable than the main edge calculation as I (obviously) have fewer observations with which to estimate the edge.

I found the TC was above 2 around 18.0% of the time and above 5 around 4.3% of the time.

I think I know the London casino that you are thinking about - I recently subscribed to Trackjack. Thats the reason why Ive been simulating this game - personal self interest, not just public spirit.

While I have given my coding a very thorough going over, its always possible that I have messed up in some non-obvious way. So it would be great if someone could comment on whether my numbers seem plausible or not, or run a further simulation using recognised software as a cross check.

Anyway, I hope this is useful. Maybe run into you at the tables in London sometime?

Mike

[ilya2121]

4d edge is -0.51%. Aces maybe resplitted unlimited times. Lack of insuranse costs to CC from 25% up to 32% of EV, depends on penetration around 1 deck.
Anyway the game is pretty good, the only problem is heat. London casinos became much more paranoid during last 2 years.

[MikeH]