3 card blackjack variant?

My local casino has a game called "3 card blackjack" that does NOT match the standard game normally called 3 card blackjack (that is a hybrid of 3C poker and blackjack). Instead the basic thrust is that each side can only hit once, or stand.

I'm curious if anyone else has seen this game, and also if there is analysis available for it?

Full rules at: http://www.coushattacasinoresort.com/uploads/2014/12/table-3-card-blackjack.pdf

Transcription of rules:
Player may hit one card, or else stand
Only split Aces, and no re-splitting
No doubling
Blackjack pays even money
Dealer will stand on >=16, or hit one card on <=15

Also a bonus bet called "Poker Play" that pays on a payout table based on the player's 2 card or 3 card hand. But if the hand busts, it will not pay the bonus either.

Weird. Not an analyst by any means, but my hunch is stand on any high hand, hit any low hand, hit a medium hand if your ones digit is less than the dealer's exposed card, stand if it's higher. Also, can you hit split aces?

I have to imagine the Straight, Flush, and Straight Flush will only pay if you take a hit, and not on the 2 card hand? Because 50 to 1 for any suited connectors, 10 to 1 for consecutive cards, and 6 to 1 for two suited cards seems too good to be true. But the poorly worded rules say " the initial 2 cards, or 3 cards if a hit is taken, are playing for the Poker Play pay table," which would suggest that two suited cards WOULD be a flush, for example. Forget about the Blackjack, if that Pay Table was valid, you should play for the Poker Hands! Or you could at least get away with it if the dealer was as confused the way those rules are written!

Quote: cyrus

...
I'm curious if anyone else has seen this game, and also if there is analysis available for it?
...
Player may hit one card, or else stand
Only split Aces, and no re-splitting
No doubling
Blackjack pays even money
Dealer will stand on >=16, or hit one card on <=15...

An infinite-deck model of the main game without the side bet gives 0.628% as the house edge (assuming the dealer stands on soft 16 as well as hard 16 and split aces receive only one card each). [If split ace hands may be hit, the house edge is only 0.462%.]

Here's the basic strategy for the main game (if the player doesn't bet the poker side bet or if play couldn't help the side bet win)
Soft standing: stand on all soft 17's except vs dealer's up cards of 8, 9, and 10 (then stand on soft 18)
Hard standing: stand on 14 vs 2; 15 vs 3-5; 16 vs 6; 17 vs 7-10; and 16 vs A
Always split aces

How many decks are in the shoe, and is a continuous shuffle machine used?

I also wonder how much the poker side bet is going to screw with your strategy. And I assume the rack card is misleading, I'm thinking you have to hit to win the poker bet.

I'd think you could win the pair with two cards. The other hands probably need three cards.

Quote: Venthus

I'd think you could win the pair with two cards. The other hands probably need three cards.

That may be true. If 2 cards were only needed to win most of the other payouts, then it would be a lot of standing going on, I think.

Quote: ChesterDog

An infinite-deck model of the main game without the side bet gives 0.628% as the house edge (assuming the dealer stands on soft 16 as well as hard 16 and split aces receive only one card each). [If split ace hands may be hit, the house edge is only 0.462%.]

Here's the basic strategy for the main game (if the player doesn't bet the poker side bet or if play couldn't help the side bet win)
Soft standing: stand on all soft 17's except vs dealer's up cards of 8, 9, and 10 (then stand on soft 18)
Hard standing: stand on 14 vs 2; 15 vs 3-5; 16 vs 6; 17 vs 7-10; and 16 vs A
Always split aces

How many decks are in the shoe, and is a continuous shuffle machine used?

Thanks for running those numbers. I'd be intrigued to see in CVCX (or other software) what a Wonging counting game would do to this. If the dealer stands on 16 and only draws 1 card on low hands, Wonging in with rather large counts seems quite advantageous. Then again no double and BJ = even money seems quite disadvantageous..

Quote: tringlomane

That may be true. If 2 cards were only needed to win most of the other payouts, then it would be a lot of standing going on, I think.

Yeah. A two-card blackjack hand will be suited 25% of the time. No way that can play 6:1.

If it does, pack you're bags. We're taking a road trip to Louisiana.

Quote: Romes

...Thanks for running those numbers. I'd be intrigued to see in CVCX (or other software) what a Wonging counting game would do to this. If the dealer stands on 16 and only draws 1 card on low hands, Wonging in with rather large counts seems quite advantageous. Then again no double and BJ = even money seems quite disadvantageous..

Below is a comparison of the "effects-of-removal" of cards in regular blackjack and the main bet of this 3 card blackjack variant (with no hitting of split aces.) The regular blackjack effects-of-removal are from this Wizard page. You can see that the effects of card removal are mostly much less for the new game compared to regular blackjack, which would mean that card counting would be much less effective for the new game.

A count for the main bet of this new game might use these tags: ace = -5, two = -1, six=+1, nine=+1, ten=+1, others=0. It's interesting that the removal of tens is good in the new game and the removal of twos is bad.

Card	Blackjack	3-card
10	-0.0049	0.0009
9	-0.0020	0.0007
8	-0.0001	0.0003
7	0.0029	0.0003
6	0.0046	0.0010
5	0.0080	-0.0003
4	0.0062	0.0001
3	0.0046	-0.0004
2	0.0039	-0.0010
1	-0.0058	-0.0041

Too bad the EoRs are garbage.

Otherwise you could spread min to max counting and nobody would be the wiser.

Quote: ChesterDog

Below is a comparison of the "effects-of-removal" of cards in regular blackjack and the main bet of this 3 card blackjack variant (with no hitting of split aces.) The regular blackjack effects-of-removal are from this Wizard page. You can see that the effects of card removal are mostly much less for the new game compared to regular blackjack, which would mean that card counting would be much less effective for the new game.

A count for the main bet of this new game might use these tags: ace = -5, two = -1, six=+1, nine=+1, ten=+1, others=0. It's interesting that the removal of tens is good in the new game and the removal of twos is bad.

Card	Blackjack	3-card
10	-0.0049	0.0009
9	-0.0020	0.0007
8	-0.0001	0.0003
7	0.0029	0.0003
6	0.0046	0.0010
5	0.0080	-0.0003
4	0.0062	0.0001
3	0.0046	-0.0004
2	0.0039	-0.0010
1	-0.0058	-0.0041

Well, that is quite interesting indeed. It would seem this is a much safer game for the house with all of the effects dampened, let alone the intriguing differences (as you mentioned, such as ten and deuce). I wonder why the ace is still so important, even though blackjacks pay 1-1... Hmm, I've got something to think about =p. Thanks again for numbers/comparison!