OBO different from Peek

[NotEnoughHeat]

I think I brought up something similar when I first joined the forum but never quite received a satisfactory response.

From what I understand, OBO should be equivalent to hole card games since the important issue is whether you lose your original bet against dealer blackjack (which from what I understand is the case in almost all hole card games regardless of peeking).

However, in CVCX the sim results do not seem to agree.

I ran six sims: hole card, no peek; hole card, peek ace; hole card, peek aces & 10s; 2x OBO; 2x dealer takes all (DTA).

Each sim was 4 billion rounds which the exact same setting. 1-2x20 spread.

The following are the SCOREs:

HC: 43.89
HC-PA10: 43.78
HC-PA: 43.50
OBO2: 38.94
OBO1: 38.83
DTA2: 32.99
DTA1: 32.88

As expected OBO beats DTA. I simmed DTA because I had a suspicion that OBO was still being counted as DTA.
The disparity in results looks large enough to rule out variance between sims.

What's going on here?

[Three]

One thing that happens is you eat cards on rounds that the dealer has BJ on no hold card with any rules. That amounts to fewer rounds per shoe. I am tempted to say reduced pen but you do see the extra cards when the dealer has BJ. Having never played no hole card OBO it is hard to comment on procedures but I did play no peek OBO which amounted to the same thing. The dealer should get BJ about 1 out of every 20 rounds. If you simmed heads-up there's about 8 cards used per round if you went to 1 hand at all those rounds used 5.4 cards/round. Assuming you simmed 6 deck/1.5 decks cut off there are 234 cards before the cut card. That's only about 30 rounds per shoe. So extra cards are eaten on 1 or maybe 2 dealer BJ rounds per shoe. It might cost you a round of play on average per shoe. That doesn't seem like enough to explain the difference since if all assumptions hold true your difference in EV per round should be less than $0.5/round assuming you left the 100 rounds per hour assumption most use.

[davethebuilder]

NHC OBO means that the dealer takes Original Bets Only so the player can safely split and double without fear of losing more than their original bet. For calculations, it is the same as the dealer peeking for BJ.

[Phoebe]

What he said.
From template, try to change both "Dealer peeks" to NHC/OBO. Look at results.

[NotEnoughHeat]

Dave. While I appreciate you taking the time to respond, that didn't add anything. You might want to look at the problem in the OP again.

Philippe. The results are as posted above.

[Phoebe]

Philippe. The results are as posted above.

Yes, I know. It was a sample protocol for Dave.
And anyone who will want check the issue.

[davethebuilder]

I was just trying to clarify the playing procedure of NHC OBO for T3 as he stated he has never played under those rules. As for the OP's issue, assuming the sims were set up correctly, I do not understand why the results differ so much so I suggest you talk to Norm.

[NotEnoughHeat]

I was just trying to clarify the playing procedure of NHC OBO for T3 as he stated he has never played under those rules. As for the OP's issue, assuming the sims were set up correctly, I do not understand why the results differ so much so I suggest you talk to Norm.

Fair enough. Occasionally even I would clear things up that everyone in the thread already understood but could come off as cryptic to newcomers.

[Three]

I ran six sims: hole card, no peek; hole card, peek ace; hole card, peek aces & 10s; 2x OBO; 2x dealer takes all (DTA).

Each sim was 4 billion rounds which the exact same setting. 1-2x20 spread.

The following are the SCOREs:

HC: 43.89
HC-PA10: 43.78
HC-PA: 43.50
OBO2: 38.94
OBO1: 38.83
DTA2: 32.99
DTA1: 32.88

Okay, realizing you were talking SCORE and not EV I decided to see if I could choose something that has the same EV for peek and then cut off 8 more cards to simulate a lost round from playing a round each shoe when the dealer has BJ so it wouldn't eat 8 cards. Now this is not exactly the same because you get to see that extra cards eaten so they are not like cards behind the cut card but you will get 1 round less some of the time.

Using Hiopt2/ASC with H17,DAS rules, 6 deck/69 cut off. The c-SCORE for 2x1 unit to 2x20 units is 43.53. To get the c-SCORE to of 39.84 you need to have 4 more cards cut off, cut off 73 cards. This would all make sense if you didn't see 4 extra cards used in a round the dealer had BJ and didn't check for BJ for whatever reason but you do see those cards. The effect is like losing a round without affecting penetration I think that is a slightly different matter. The thing to consider is you lose the last round of the shoe which will have much higher average EV than the average EV per round. Factoring that in it would come close to explaining things.

[NotEnoughHeat]

Factoring that in it would come close to explaining things.

Thanks for the post. I had considered the card eating but perhaps didn't give it as much weight as I should have. I'm still surprised that there is such a difference here, but considering the spread it could make sense that differences are more noticeable.

[NotEnoughHeat]

Now that I think of it. T3s explanation would make sense explaining the difference between Peek and NHC but if it was the case then there would be a noticeable difference between Peek-Ace-Ten and Hole-Card-No-Peek. The latter would be more similar to OBO but the sim shows it as being closer to the Peek sims.

[Phoebe]

Try version 5.0.132

[Norm]

If you specify Full Indices wit CVData/CVCX, it doesn't matter. If you specify Ill18 or Sweet 16, CV removes the indices not associated with Ill18 or Sweet 16, and replaces them with the basic strategy plays in the included Complete Basic Strategy. If you have No Hole Card specified, it uses NH basic strategy. However, the included strategy is not the correct strategy for OBO. So, it was playing an OBO game but not using the correct strategy for some plays. Changing the included BS to OBO strategy would provide the correct results. But, I’ve changed it so it uses the normal BS when you are simulating OBO.