The Best Bet at Craps is Usually Honored in the Breach

Guess what's the least popular bet at craps, in terms of how many players toss how much on it? Here are two hints. Together, they'll either be a give-away or make the question more enigmatic. First, the wager in question vies with Pass and Don't Pass for allowing the casino the least edge in the game. Second, it occupies the most space on the felt of any bet.

It's the Come bet. Come is a distant second to Place and Buy for money on the layout after a shooter establishes a point. Yet, players who leave their come bets flat -- take no odds -- give the casino an edge equivalent to only $0.014 per dollar at risk. With odds taken, the house's fraction of the action on is smaller still. Compare this $0.014 or less to $0.015 placing the six or eight, $0.040 placing the five or nine, $0.067 placing the four or 10, and from $0.026 to $0.050 buying the four or 10.

Come bets get short shrift for a sheaf of reasons. Some can be justified psychologically -- players should be comfortable with their wagers. But none hold up under statistical scrutiny.

In a few instances, players simply don't understand Come bets. Shame on them for joining a dangerous game without mastering it!

Another objection to Come bets is that money moved to a number this way is automatically taken down after a win. In contrast, Place and Buy bets stay put if the shooter repeats. This argument is meaningless in that a new come bet can be made after a payout. And, who's to say what the next roll will bring?

Occasionally, solid citizens dislike inability to specify the number to which the Come bet will move. Great for the oracles who can reliably predict the future. For those of us not gifted with omniscience, the issue is moot.

Sometimes, the complaint about Come bets is that they're the only propositions in the casino that can't be recalled or pressed at will. These are, indeed, restrictions. But they only apply to the flat portions of the wagers. Odds on Come bets can be raised within the limits or removed at any time. And, anyway, everyone remembers when they "knew" something was going to happen and were right, but forgets when they had the same feeling and were wrong.

The most common, and simultaneously silliest, argument against Come as opposed to Place or Buy bets runs to the effect that "a Come bet has to hit twice before you win anything, but you make money on a Place or Buy the first time it rolls." To appreciate why this contention is fatuous, consider the chances that Come and the various Place or Buy bets will actually win and lose.

Come bets win on the initial rolls if a seven or 11 pops (22.22 percent probability); they lose at this stage on a two, three, or 12 (11.11 percent probability). The chance of a Come bet moving to the six or eight, then winning, is 12.63 percent; going to the six or eight, then losing, has a likelihood of 15.15 percent. Similarly for chances of reaching and resolving the other numbers. On the five or nine, overall win and loss probabilities are 8.89 and 13.33 percent respectively. And on the four or 10, they're 5.56 and 11.11 percent. Adding it all up, come bets have prospects of 49.29 percent to win and 50.71 percent to lose.

In contrast, a Place bet on the six or eight has chances just over 45.45 percent to win and 54.54 percent to lose. On the five or nine, the figures are 40 percent to win and 60 percent to lose. And on the four or 10, the probabilities are roughly 33.33 percent to win and 66.67 percent chance to lose.

So Come bets are always more apt to win than Place or Buy bets. And, when the last "equals" on the calculator is finally pushed, the probability differences more than compensate for the payoffs on Places and Buys being higher multiples than the even money on the flat parts of Comes alone. The odds are pure gravy. So, why are Come bets most often honored in the breach? Maybe the craps crowd missed this musing of the wily wordster, Sumner A Ingmark:

If risk grows faster than reward,
A bet is wisest when ignored.