Discount Gambling

At Barona Casino near San Diego, they’re spreading a new blackjack table game with an electronic twist. The game uses a human dealer, regular cards, and a modified blackjack table (Shuffle Master i-Table). However, players don’t use chips to bet, but instead use a drag & drop touch-screen. There’s no printing on the mat, and the table has a clean look, uncluttered by chips. This game is offered in a separate “Chipless Blackjack” pit of 4-5 tables.

The decks are shuffled by hand, and loaded into the shoe normally. The shoe reads the cards as they’re dealt, so it knows what everyone’s totals are. Your display prompts you for your actions. The game is a little faster than normal blackjack, since the dealer doesn’t have to mess with any chips, except during buy-ins and cash-outs.

This electronic format offers an interesting odds side bet on your starting two-card hand. The computer offers you particular odds for winning, based on your total vs. the dealer up card, as an optional side bet. For example, say you’re dealt 6,4 (=10), and the dealer is showing a 4. It offers you something like $2.69 on an additional $5 side bet (equal to the original bet) to win the hand. If you lose the hand, you lose the side bet. If your hand pushes, the side bet also pushes. Note that the true odds for winning this double should pay about $2.95, so the vig here is about $.25 😦

I worked out the true odds for a $5 bet, a six deck shoe, and dealer hit on soft-17. (The offer is made after the dealer peeks for blackjack.) The table is presented below.

I played for a few hours, and checked how close the offered side bet was to the true odds. For a $5 bet, the payouts were almost always lower than true odds by about $.25 -$.50 (no surprise here). For the payouts above $10, this might be acceptable. However, this vig really hurts the low payouts. For example, a player 20 against a dealer 10 should win about $.32 for a $5 bet. Their offered odds on this hand were often lower than $.10. That’s really bad, since you’re not even getting 1/2 the fair odds. The only odds that seemed to be true odds were for 14 vs. a dealer 6 and 7. The worst odds seemed to occur for double and split opportunities, and anything against a dealer 10 or A. (I think this is because in some situations, like 7,7 vs. an 8 upcard, the odds are quite different for splitting vs. hitting; to avoid any player edge, they offer you the lower of the two.)

In my session, almost no one took odds. Most people correctly assume the house isn’t offering true odds, and they don’t have a chart for comparison. True odds aren’t very intuitive, since people don’t think of blackjack hands in these terms.

So, as usual, you have to pay some vig to get odds in a blackjack side bet. The vig for $5 bets seems to average between $.25 and $.50 for all payouts. I’m not sure, but I think the vig scales up with the bet size, so its no better for $10, $25, or $100 bets.

Positive Count Shoe

I tried a simple experiment to see if a positive count shoe would swing the odds enough to make the side bet +EV. I removed a 2, 3, 4, 5, and 6 from the deck (+5), then re-calculated the odds. Overall, this made the basic blackjack game +EV, swinging it from about -0.5% to +0.5%. Then I checked a few odds points that would benefit the most from this positive count. I saw that a (T,4) vs. a dealer 6 moved from $6.37 down to $6.20. This doesn’t overcome the vig, and the offered odds are still probably below this. Similarly, for the same +5 shoe, a (5,6) vs. a dealer 6 improves only from $2.34 to $2.25. So, a negative shoe count doesn’t help the odds bet much, and will never overcome the vig.

Negative Count Shoe

Experiments with a -5 count shoe (five Tens removed) show that the higher odds bets improve more in absolute payouts than for improvements from positive count shoes. For example, a 14 vs. a dealer 10 upcard improves from $14.78 down to $14.36. Intuitively, the odds improve because fewer Tens in the shoe mean a higher chance of the dealer not having 20, and a better chance for the player to draw to a hand. Similarly, a 16 vs. a dealer 9 improves from $16.59 down to $16.18. These are significant improvements, but I believe that the vig on the higher odds bet are steeper for a reason: to account for this variation. So overall, it looks like counting isn’t going to ever make the odds bets +EV 😦