Playing It Smart: Blackjack can have a low edge, but you have to know how to lower it

House edge in blackjack arises because the dealer goes last. Players who bust by exceeding 21 lose, no matter what later befalls the dealer. The inherent advantage of this effect would make the game unworkable without provisions to curb it.

One way this is done is to impose restrictions on dealers' actions. First, they're limited to either hitting or standing. Second, they have to follow definite rules, and may not arbitrarily decide which to do. Specifically, dealers must hit totals of 16 and under, stand on 17 and above, and in the special case of "soft seventeen" formed with an ace evaluated as 11 stand or hit according to the house's stated policy.

A second class of factors that tame the house's edge comprises options available to players. They may hit or stand at will. They may also double down on their initial hands, raising their bets then taking a single additional card. And, they may split pairs, matching their original wagers then playing each half as the start of a new hand. Some casinos also offer "surrender," the chance to withdraw from a round by sacrificing half a bet. In all such instances, the probabilities associated with alternate decisions and results have been calculated. They are codified in "Basic Strategy," which gives the move for every combination of player's hand and dealer's upcard that minimizes overall edge.

Of options available to solid citizens, ability to stand on totals between 12 and 16 against appropriate dealer upcards is the most helpful. This alone cuts 3.2 percent from what otherwise would be the edge. These totals only win if the dealer busts, and no pushes are possible. Basic Strategy for such situations weighs the prospects of the dealer busting and any bettor still in the game winning, the chance of a player drawing a card and losing immediately, and the likelihood a hit will lead to 17 through 21 that ultimately beats, pushes, or loses to the dealer's total.

"Late surrender" applies after the dealer checks for a blackjack. A player with a 15 against 10-up or a 16 against nine-, ten-, or ace-up has under 50-50 chance of winning by hitting, and less by standing. So recovering half a bet is statistically superior to taking the shot. It shaves a mere 0.07 percent from the overall edge but should be a no-brainer when available.

Doubles, which can trim the edge by 1.6 percent, often create a quandary. One reason is that players must bet more to exploit them. And not everyone has the stomach or bankroll to do so. Another is that the chance of winning by doubling is the same as or less than that of the next best choice; the benefit comes from making a second bet which is more apt to win than lose.

Properly splitting pairs recoups 0.4 percent in edge. This tactic also perplexes some players. One issue is that more money must be placed at risk. In addition, while some splits obviously improve players' positions (e.g., against a six, starting with two sevens is intuitively stronger than standing on 14), others seem to pose a damned if you don't and damned if you do dilemma. A classic example is a pair of eights versus a 10. Many players argue that something like $10 on 16 is a loss waiting to happen, and $20 on each of two eights is twice as bad. True, twice the money is at stake. And the hands are underdogs either way. But the odds are less adverse with the eights than the 16. Regardless of what may transpire in any given round, the average loss versus a 10 with $10 on 16 exceeds that with $10 on each of two eights.

Another thing. It's the 3-to-2 payoff when players win with blackjacks, which is not mirrored by demands for more money when dealers receive these hands. This feature lowers the edge in the game by roughly 2.5 percent. Bettors needn't do anything to earn this bonus. But they should keep it in mind when they stumble on a blackjack variation with lots of bells and whistles, but that pays under 3-to-2 for a natural. You've heard what they say about looking too good to be true? Any other bait such a game seems to dangle rarely makes up for the 2.5 percent relinquished. As the beloved bard, Sumner A Ingmark, cleverly counseled:

Casinos survive 'cause they get the last laugh,
On know-it-all players too smart by a half,
Who cannot distinguish the wheat from the chaff.