MBSBJ
Joined: Jun 8, 2013
• Posts: 8
July 17th, 2014 at 10:36:27 PM permalink
The rules of said game are:

Single deck (Cards Shuffled after each round)
Blackjack pays 3-2
Dealer stands soft 17
Double any first two cards
Double after split
Re-splitting aces not allowed
Early surrender against 10 (European, Dealer takes one initial card)
Loses both bets in cases of split/double down against a dealer Blackjack

any information on such games available on the wizard's site? Would it be advantageous to keep a count with multiple boxes opened to tweak on a composition dependent strategy and what is the house edge (player edge possibly in this case possibly)
RS
Joined: Feb 11, 2014
• Posts: 8623
July 17th, 2014 at 10:43:48 PM permalink
It would be advantageous to learn the proper strategy and flat bet table max.
AxiomOfChoice
Joined: Sep 12, 2012
• Posts: 5761
July 17th, 2014 at 11:09:03 PM permalink
Quote: RS

It would be advantageous to learn the proper strategy and flat bet table max.

According to WoO, a 0.075% player advantage, assuming perfect composition dependent basic strategy.

This is essentially a break-even game. If you flat-bet \$1000 you will make \$75 per 100 hands. Kelly Criterion tells us that you need a bankroll of somewhere in the ballpark of \$1.5 million for bets of this size.

Realistically, this edge is too small to be worth anything. You can't make enough money relative to your bankroll (no one with a \$1.5 million BR is interested in \$75 per 100 hands)
MBSBJ
Joined: Jun 8, 2013
• Posts: 8
July 18th, 2014 at 12:17:21 AM permalink
Quote: AxiomOfChoice

According to WoO, a 0.075% player advantage, assuming perfect composition dependent basic strategy.

This is essentially a break-even game. If you flat-bet \$1000 you will make \$75 per 100 hands. Kelly Criterion tells us that you need a bankroll of somewhere in the ballpark of \$1.5 million for bets of this size.

Realistically, this edge is too small to be worth anything. You can't make enough money relative to your bankroll (no one with a \$1.5 million BR is interested in \$75 per 100 hands)

What if you're able to play for example 10 boxes, keeping a count and tweaking the composition dependent strategy
AxiomOfChoice
Joined: Sep 12, 2012
• Posts: 5761
July 18th, 2014 at 12:39:03 AM permalink
Quote: MBSBJ

What if you're able to play for example 10 boxes, keeping a count and tweaking the composition dependent strategy

Back of the envelope calculation:

I seem to recall someone saying that, with perfect play (doable if this is an online game) each card seen is worth 0.02%. The average number of cards seen per hand is 2.7. So the first hand has an edge of 0.075, and each subsequent hand has an edge of 0.054 more than the previous one. So your 10th hand would have an edge of 0.561%, which is definitely playable, though slim.

If you could play something like 9 hands of \$1 and the 10th box at \$1000, you would have an edge of \$5.60 per round. You'd still need a bankroll of something in the \$200k range. Of course you can scale this down -- smaller bets with smaller bankroll. I'm not sure how many rounds per hour you could play -- this level of win requires computer-perfect play.

With the 9x1 ; 1x1000 betting, each card you take on one of the first 9 hands is worth 20c of EV on your last hand. That means that in remotely close situations (and even a lot of not-so-close ones) you should take the strategy that will use the most cards -- ie, in any situation where the EV of hitting is within 20% of the EV of standing, you hit. This would raise your expectation slightly, but I doubt you will be able to get much more than \$6 per round. You could probably do better (with the same bankroll) by betting big on more hands -- with the same bankroll you could bet \$700 on each of the last 2 boxes which would raise your edge to over \$7.50 per round. I'm not sure where the sweet spot is here.

This is just an estimate. I have no idea how close to accurate it is. I would not bet serious money here without simulating it first.
RS
Joined: Feb 11, 2014
• Posts: 8623
July 18th, 2014 at 7:23:19 AM permalink
Quote: AxiomOfChoice

According to WoO, a 0.075% player advantage, assuming perfect composition dependent basic strategy.

This is essentially a break-even game. If you flat-bet \$1000 you will make \$75 per 100 hands. Kelly Criterion tells us that you need a bankroll of somewhere in the ballpark of \$1.5 million for bets of this size.

Realistically, this edge is too small to be worth anything. You can't make enough money relative to your bankroll (no one with a \$1.5 million BR is interested in \$75 per 100 hands)

I got 0.66% player edge with BJ info.

http://www.blackjackinfo.com/bjbse.php?numdecks=1+deck&soft17=s17&dbl=all&das=yes&surr=es&peek=no
MBSBJ
Joined: Jun 8, 2013
• Posts: 8
July 18th, 2014 at 9:21:15 AM permalink
Quote: RS

I got 0.66% player edge with BJ info.

Yeah i keep getting different player edge % from various sources

AxiomOfChoice
Joined: Sep 12, 2012
• Posts: 5761
July 18th, 2014 at 10:59:07 AM permalink
Quote: RS

I got 0.66% player edge with BJ info.

http://www.blackjackinfo.com/bjbse.php?numdecks=1+deck&soft17=s17&dbl=all&das=yes&surr=es&peek=no

Oh, my mistake. I missed the early surrender (WoO doesn't offer that as an option in the calculator).

This is a WAY stronger play, then. I think you can use the same method that I used to analyze it, although, again, these are just estimates. I would just take a day and write a simulator and be done with it.
AxiomOfChoice
Joined: Sep 12, 2012
• Posts: 5761
July 18th, 2014 at 11:02:44 AM permalink
Although, in fairness, that table assumes that ES is allowed against an A as well as a 10 -- I believe the stated rules were ES10. So the actual edge is not that high.
AceTwo
Joined: Mar 13, 2012