The question “Does craps or blackjack offer better odds” was posed to me at breakfast this morning with Ryan.
I said “Blackjack”.
Ryan replied, “I’ll make you a $10 bet that Craps has better odds than Blackjack”.
I took that bet.
The problem with this bet is a common one in business… it’s a very poorly defined contract. We did not write it down. We didn’t define a number of key terms. We didn’t say which situation this should apply to.
Here is some of the issues:
- Real life or hypothetical? Hypothetically, Craps could approach nearly 49.999% odds of winning for the player if the player only played the pass line bet than went 1,000 time higher on the Odds bet. In real life, the very best deal we could find was a casino in Vegas offering 100 to 1 on the odds bet. Hypothetically, a perfect Blackjack player could count cards and have somewhere around a 51.5% chance of winning. Which leads to the next point…
- Is this just a one time bet or over the course of an entire day of playing the game? I was thinking it was over the course of time… no one just plays one hand of Blackjack, you sit down and play for awhile. Ryan thought it was just one individual bet inside of one single toss of the dice for Craps.
- I also simply have never played Craps and had to have him explain the structure to me. Ryan seemed more knowledgeable in Craps, but not an expert. He did not seem to know much about Blackjack, and I know just enough to be mildly dangerous.
- In a casino or at home? I think we both meant in a casino, but this was not defined.
- Bad players vs good players vs optimal players vs superhuman players? Most people who play both craps and poker are not good players and lose significant money at both. This article from UNLV analyzes this a bit… a craps player playing pass / come gives the house a 1.4% advantage, but a player playing craps with pass/come with double odds only gives the house a 0.6% advantage. The average blackjack player gives the house 2% odds, a good blackjack player only gives the house a 0.5% advantage, and a card counting player actually has a positive 1% advantage for himself. Ryan tried to counter the card counting player by saying theoretically a superhuman craps player could control how to roll the dice like a bowler bowls, but there is no real life examples of this.
I argued vociferously that most people can be trained to do a basic card counting system, therefore blackjack is the best game in a casino for the optimal player. I also said that an average player will do better with blackjack by just following the dealer’s advice and only give the house a 1% advantage, whereas the average craps player usually does lots of weird bets that make the game more fun but give the house a 10-20% advantage.
Ryan argued equally strongly that card counting is not allowed by casinos and we should only look at a single portion of a single bet (the 50-50 ratio of the Odds bet in craps) vs a single hand of blackjack.
I think the real lesson here is the vital importance of only making agreements that both parties have a much stronger understanding of than what Ryan and I had here. It was fun though to debate and learn more about each game though.