Card Counting the Bad Beat Baccarat Side Bet: Natural 9 beats Natural 8

Well, you knew it was going to happen sooner or later, that I would analyze another AP opportunity. A reader made a comment on one of my YouTube videos, saying

“I’m a pit boss, been in the industry about 20 years. Had some success counting blackjack. Recently I’ve been thinking about computer perfect blackjack play at online casinos. Another exploit that came to my attention recently is the casino I work at was offering bad beat baccarat. A young Asian couple were counting to beat nat 9 over nat 8 (pays 50:1). I’ve been looking for the system they used but haven’t had much luck. I’ve seen a few systems for beating similar bets, but not specifically nat 9over nat 8. Anyone have help to offer?”

I quickly located the side bet over at the Wizardofodds website, the so-called “Bad Beat” baccarat side bet — here is a link:

The wager in question is the one that pays 50-to-1 whenever a natural 9 beats a natural 8. I wrote a quick computer program to verify Mike’s computation of the house edge, and I agree with his result.  With 8 decks, the house edge is 8.4585%.  While this is a decent edge for the house, it is clearly within range if there are obvious key cards.

So yes, there are obvious key cards.  You would have to believe Pluto is still  a planet to not immediately see that the AP loves a shoe rich in 8’s and 9’s against this side bet. I wrote a program to compute the EOR’s and quickly found this nearly optimal counting system:

  • 2, 3, 4, 5, 6, 7, T, J, Q, K = +1
  • 8,9 = -5
  • A = 0

Well, nearly optimal because the Ace roughly has the same EOR as all the other cards 2, 3, 4, … but we do need a balanced count to make life easier. So I randomly assigned A = 0.

Here are the results of a simulation of 200M (two hundred million) eight deck shoes, with the cut card placed at 14 cards:

  • Trigger true count = 6
  • Percent of hands with edge: 14.31%
  • Average edge: 10.54%
  • Units won per shoe: 1.23
  • Units won per 100 hands: 1.51

This side bet is definitely beatable. A player could make a decent living winning 1.51 units per 100 hands, though there is considerable variance owing to the rarity of the winning hand. The Bad Beat “Natural 9 beats Natural 8” side bet wins at almost twice the rate of the Dragon 7 side bet.  In other words, it’s a very solid opportunity for the AP.

Addendum, May 3, 2022

A comment on my YouTube channel after I posted this article mentioned that the same wager is available on 9 beating 7 in their casino, and wondered if the same counting system will work. That is, Player natural 9, Banker = 7 or Player = 7, Banker natural 9 both pay 50-to-1, otherwise the wager loses.

The house edge for this variation is 8.103214% — about the same as the previous version.

The count system for this side bet is very similar to the previous:

  • 2, 3, 4, 5, 6, 8, T, J, Q, K = +1
  • 7, 9 = -5
  • A = 0

A simulation of 100M shoes, with the cut card placed at 14 cards, yields the following:

  • Trigger true count = 6
  • Percent of hands with edge: 14.33%
  • Average edge: 10.85%
  • Units won per shoe: 1.26
  • Units won per 100 hands 1.56

If your casino offers either of these side bets, definitely watch out for advantage play!

[(U) —]