Negative ev

[Archvaldor]

I think there's something off with your numbers. Even if a SD game with Ro7 pen existed, which I doubt, it's unlikely you'd achieve a SCORE of $90 with a 1-6 spread.
I ran a simulation using Hi-Lo for SD, H17, DOA, DAS, SPA1, SPL3, BJ 6:5, and Ro7. The SCORE I got with a 1-6 spread was only 5.20.

It should be noted in practice the "Rule Of" critieria tends to be greatly relaxed in heads-up games with increments or even multiple increments of hands dealt over and above the technical procedure.
So you might get an extra couple of hands dealt before the shuffle. This will make a difference.. (I'm not asking you to redo the simulations, you've done enough already, but I thought this was worth mentioning).

Attachment 5230

This is single deck, H17, DAS, no surrender, no resplit aces, split to 4 hands, Rule of 7, 3 players, counter sitting in seat 3, 1 to 6 spread. You're looking at a winrate of almost $100.

I don't think these numbers accurately represent actual game conditions. The dealer will violate RO7 if they are worried about potentially running out of cards. That's when most of your big advantages will occur.


At the start of this thread I wrote "All things being equal there is no reason to play 6:5.

If there are procedural aberrations such as dealer errors, card exposure or they deal down to the last card or something you might have an opportunity. "

Nothing I've seen in the subsequent discussion has suggested otherwise.

[moo321]

Rule of 7 pen is pretty good. It's not nonexistent, but it's good. It doesn't have to involve manipulating the parameter of Rule of X pen, though. It still works with a cut card.


Obviously, if you have a very deep cut card, like 14 cards, you'll have a huge advantage with a very small spread. But even at 18 cards, which is 2/3 penetration, you can get a winrate of almost $50 with only a 1 to 6 spread. That's H17, DAS, no surrender, using Hi Lo and sitting at 3rd base with 4 players.

6 to 5.jpg

Even with a garbage game like 26 cards cut off, you can get ok results if you use a larger spread. 1-20 spread will still get you a winrate of about $20. Not great, but not worse than most of the 8 deck H17 shoes that tons of people play right now, with the exact same spread.

[G Man]

I am on my way out at this moment, but you seem to get really higher advantages at similar counts compared to my data.

[Archvaldor]

Rule of 7 pen is pretty good. It's not nonexistent, but it's good. It doesn't have to involve manipulating the parameter of Rule of X pen, though. It still works with a cut card.


Obviously, if you have a very deep cut card, like 14 cards, you'll have a huge advantage with a very small spread. But even at 18 cards, which is 2/3 penetration, you can get a winrate of almost $50 with only a 1 to 6 spread. That's H17, DAS, no surrender, using Hi Lo and sitting at 3rd base with 4 players.


Even with a garbage game like 26 cards cut off, you can get ok results if you use a larger spread. 1-20 spread will still get you a winrate of about $20. Not great, but not worse than most of the 8 deck H17 shoes that tons of people play right now, with the exact same spread.

There's clearly a role for a specialist hunting down these games specifically, perhaps using creative approaches. I do feel some other respondents in the thread have been a little too negative about the possibilities here which are interesting.

[DSchles]

I am on my way out at this moment, but you seem to get really higher advantages at similar counts compared to my data.

I agree. Something is wrong. The screen shot looks more like a regular 3:2 game than 6:5. The penalty is 2.325% in SD. Where is that showing up at counts 0 and below or +1?

Don

[Archvaldor]

I agree. Something is wrong. The screen shot looks more like a regular 3:2 game than 6:5. The penalty is 2.325% in SD. Where is that showing up at counts 0 and below or +1?

Don

He's talking about a deeply dealt single-deck game. Most of the money is bet at deep penetration. Strategy gain is going to be much higher than we would expect with a conventional shoe game and will compensate at negative counts.

It used to be the case that you could obtain a small advantage from flat betting 3:2 games in SD when the penetration was excellent, which would have resulted in approximately the same reduction in house edge. That was without bet variation when more money on average would be bet on initial hands in the deal.

I don't think the number supplied is itself outlandish offhand though some external confirmation would be helpful.

[Cacarulo]

I’ve been running some sims, and honestly, I haven’t been able to get a SCORE of 90 for the third base player (Ro7).
However, there is definitely an advantage, but I need a bit more time to see exactly what it is.

Sincerely,
Cac

[DSchles]

He's talking about a deeply dealt single-deck game. Most of the money is bet at deep penetration. Strategy gain is going to be much higher than we would expect with a conventional shoe game and will compensate at negative counts.

It used to be the case that you could obtain a small advantage from flat betting 3:2 games in SD when the penetration was excellent, which would have resulted in approximately the same reduction in house edge. That was without bet variation when more money on average would be bet on initial hands in the deal.

I don't think the number supplied is itself outlandish offhand though some external confirmation would be helpful.

I used his penetration in CVCX for a SD 3:2 game with his rules. Subtracting 2.325% from the <=0 and +1 counts yields edges much worse than what his sim showed.

Don

[DSchles]

Rule of 7 pen is pretty good. It's not nonexistent, but it's good. It doesn't have to involve manipulating the parameter of Rule of X pen, though. It still works with a cut card.


Obviously, if you have a very deep cut card, like 14 cards, you'll have a huge advantage with a very small spread. But even at 18 cards, which is 2/3 penetration, you can get a winrate of almost $50 with only a 1 to 6 spread. That's H17, DAS, no surrender, using Hi Lo and sitting at 3rd base with 4 players.

6 to 5.jpg

Even with a garbage game like 26 cards cut off, you can get ok results if you use a larger spread. 1-20 spread will still get you a win rate of about $20. Not great, but not worse than most of the 8 deck H17 shoes that tons of people play right now, with the exact same spread.

So: SD H17, DAS. 1-6 spread. Four players, which is what your screenshot purports to show. SCORE was 11.49 (compare to your 48.82!). Hourly win was $23. Edge at 0 was -1.37%. At <= -1. it was -3.07%. Edge at +1 was -0.66%.

As I said, something is wrong.

Don

[DSchles]

Tried to print screen shot for the above comments, but got gibberish. What needs to be done?

Don

[Archvaldor]

I used his penetration in CVCX for a SD 3:2 game with his rules. Subtracting 2.325% from the <=0 and +1 counts yields edges much worse than what his sim showed.

Don

Thanks very much, I appreciate it.

A point of theory: would we not expect a lower relative edge in negative counts in a conventional 3:2 game vs 6:5, as the blackjack bonus is worth more initially and consequently declines more in value?
(This may not be germane to the discussion about the moo sims but I feel it would be useful if nothing else to convert sim data from one game meaningfully to the other).

[DSchles]

A point of theory: would we not expect a lower relative edge in negative counts in a conventional 3:2 game vs 6:5, as the blackjack bonus is worth more initially and consequently declines more in value?

Not sure. You're still getting only 6:5 when you do get a natural, instead of 3:2.

Would have to create the sims both ways and then compare.

Don

[Cacarulo]

Hi everyone,

After running several simulations, here are the conclusions I’ve reached. Let’s consider the game proposed by moo321:
SD, H17, DOA, DAS, SPA1, SPL3, NS, BJ 6:5, and RO7 pen.
For my analysis, I used Hi-Lo, R22 indices (floored), full-deck division, and deck estimation to the nearest 1/4 deck. 20 billion rounds for each sim.

First off, while this game with that level of penetration (if offered by a casino) is quite playable, it’s highly unlikely to achieve a SCORE of 90 with a 1-6 spread.
Here are the SCOREs I obtained (with the counter seated at third base):

a) 1 player, 6 rounds (heads-up): 5.95
b) 2 players, 5 rounds: 33.13
c) 3 players, 4 rounds: 34.78
d) 4 players, 3 rounds: 22.23
e) 5 players, 2 rounds: 0.66

Clearly, the most promising options are b) and c).
Even with only 22 indices, I don’t believe using all of them would get you a SCORE of 90.

Sincerely,
Cac