Basic Strat ending up with player Edge

I've tried to recreate a basic strategy chart / EVs for my local set of rules. I'm starting with the simplifying assumption of infinite decks (In reality, my local casino has 8 decks in a CSM.

I'm starting with Michael Shackleford's video here

He links to a spreadsheet which I used as my starting off point. However, I did have to make modifications given that this is ENHC, so the dealer could in fact have blackjack as opposed to the American peeks for blackjack. So I added a row of potential dealer ending values for a given starter that includes a BJ.

Additionally, I need to figure out how to model OBBO, I've simply added back a unit of EV when doubling facing a dealer 10 or A x the probability of a blackjack, so either 1/13 or 4/13 as you essentially get a rebate if you double and the dealer has a blackjack. Same thing for splits. This is probably the weakest adaption I did. I think I need to subtract that rebate by the chance that the player busts after splitting.

I tried to do things right but I end up with a player edge of 0.2325% instead of a house edge and I can't quite figure out where I'm going wrong.

Here is my modified sheet:

Google Sheet

Any one out there able to point out what I'm doing wrong?

Edit: Hmm. I'm having difficulty posting links.

Here is the youtube video: youtube.com/watch?v=jCF-Btu5ZCk
Here is the spreadsheet: docs.google.com/spreadsheets/d/1RfzlPrOHxiMvlbS1Y0zXXgRfajnhUpD1/edit?usp=sharing&ouid=113666895806815773578&rtpof=true&sd=true

The most likely problem is on the surrendering rules of ENHC. In a hole game, players are allowed to surrender to any dealer upcard after dealer peek, but in an ENHC, players are not allowed to surrender to a dealer upcard of Ace.

Quote: monodactyl

Here is my modified sheet:

/spreadsheets/d/1RfzlPrOHxiMvlbS1Y0zXXgRfajnhUpD1/edit?usp=sharing&ouid=113666895806815773578&rtpof=true&sd=true

Any one out there able to point out what I'm doing wrong?
link to original post

Was not able to connect with your spreadsheet. Can you post it as a link? check the formatting codes below this screen on the bottom right.

Try
https://docs.google.com/spreadsheets/d/1RfzlPrOHxiMvlbS1Y0zXXgRfajnhUpD1/htmlview

You're right about not surrendering against an Ace. I initially made that mistake. However, I removed surrender as an option against Ace upcard for where surrender was the best option (player 12 - 17). Right now there are no surrenders against a dealer Ace.

I’m interested. How about you restate the rules you use for the EV calculation? Also, there are usually 6 decks in an automatic shuffler, but you say 8.

Actually, for simplicity I'm using infinite deck on the spreadsheet - I'm not sure how to move on to actual composition dependent spreadsheet yet, I probably first want to figure out why I'm getting a positive EV in this simplified version. As to the other rules, it's:

ENHC, OBBO
S17
DAS
Split 3 times to make 4 hands
No resplitting or drawing to split Aces.
Surrender against any card except a dealer A.

There are two versions of OBBO here. One is true OBBO, the other is OBBO EXCEPT when doubling an 11. If the dealer ends up with blackjack, both original and doubled bets are taken.

I see at least two problems here. In the Wizard’s Excel spreadsheet, pair-splitting is restricted to one time only to make 2 hands. So, how do you generate this to splitting 3 times? In an ENHC game I played, players are not allowed to double 11 vs. 10. I forgot, but they probably didn’t allow surrender 16 vs. 10 either.

Yeah. There's a lot of distance between the real rules of my casino and the sheet, but being able to model 3x splitting would probably improve my EV even further. I'm mainly trying to figure out the bigger problem of why in this supposedly more rigid against player set of rules (infinite deck, limited resplitting) it's still showing up as positive EV.

I'm 100% sure that doubling vs 11 vs 10 is allowed. That's where the OBBO exception rule (Doubles when player has 11 have both bets lost when dealer gets blackjack).

Surrender against 10 is allowed, against A isn't.

A lot of the rule can actually be modeled by: i, the exception being the OBBO rule. My workaround with bjstrat is just back 1 unit x 4/13 against A and 1/13 against a dealer 10 for the times I'd get my rebate for the added bet.

But yeah, I'm more concerned at why my spreadsheet is yielding positive.

bjstrat is yielding house edge of -0.2634% and doesn't include the favorable OBBO nature of the my actual game.

I’m very interested in your pair-splitting part. In the Excel spreadsheet you attached, do you split pairs one time only? Or up to three times?