The dealer turns an ace, says the word insurance, and the table waits. Somebody with a hard 20 slides out a half bet, because it would be painful to hold a good hand and lose it to a natural. That instinct has cost more money than any misplayed 16, and the reason is that the offer is not what the player thinks it is.
Insurance is not a hedge. It is a second, unrelated wager on one question: is the dealer's hole card a ten, jack, queen or king? Your own hand is not a party to that bet, because the payout does not depend on it.
Price the offer
Six decks, 312 cards, dealer showing an ace. The hole card is one of the 311 cards you have not seen, and 96 of those are ten-valued. So the chance the bet wins is 96/311, or 30.87%.
Insurance pays 2 to 1. Expected value on one unit staked is therefore 2 x (96/311) minus (215/311), which is (192 - 215)/311, which is minus 23/311, or minus 7.4%.
For the bet to break even, the hole card would have to be ten-valued a third of the time. Six decks put it at 30.87%. That gap is the price, and it is fixed by composition rather than by anything the operator chose. A single deck is a little kinder, at 16/51 winning chances and an expectation of minus 3/51, or minus 5.88%, and every extra deck pushes the number further from break-even. Nothing in the rules panel moves it.
Your good hand makes it worse
Insurance gets more expensive precisely when the player feels most like taking it.
|
Your hand against the ace |
Chance hole card is a ten |
Expectation per unit insured |
|---|---|---|
|
Unknown or irrelevant |
96/311 = 30.87% |
minus 7.40% |
|
Two ten-value cards (a 20) |
94/309 = 30.42% |
minus 8.74% |
|
A natural (ace plus ten) |
95/309 = 30.74% |
minus 7.77% |
Holding a 20 built from two ten-cards means two of the cards insurance needs are already face up in front of you. The bet you are being offered is worse than the generic version by more than a point of expectation, and it is worse because your hand is strong. Insuring your best hands is the single most expensive habit available at the table.
What that costs in a game with almost no edge
Now put it against a main game engineered to be cheap. The published figures for Duel Blackjack are a 99.46% base return and a stated 99.78% effective return once a rakeback of up to 60% is credited per wager, which leaves a house edge of 0.22% on the base bet.
A dealer ace shows up on 4/52 of hands, which is 7.69%. Insuring every one of them for the standard half bet costs 0.0769 x 0.5 x 0.074 = 0.00285 units per hand dealt, or 0.28% of the base bet.
Read those two numbers next to each other. The game costs 0.22% per hand. The insurance habit costs 0.28% per hand. A player who takes insurance whenever it is offered has more than doubled the price of their session, and the extra cost comes from a bet they think of as defensive. All the engineering that got the main game down to a fifth of a percent is undone, with change left over, by a reflex that takes half a second.
Even money is insurance in a better coat
The same bet reappears when the player holds a natural against an ace and is offered even money. Take it and you are paid one unit, guaranteed, and the hand is over.
Check what you gave up. Without the offer, the natural pushes when the hole card is a ten, which is 95/309, and otherwise pays 3 to 2, so the expectation is 1.5 x (214/309) = 1.039 units. Even money pays 1.000. The offer costs 0.039 units every time, close to 4% of the bet.
It is also literally insurance. Stake half a unit: if the hole card is a ten your hand pushes and insurance pays one, netting exactly one unit; if it is not, your natural collects 1.5 and insurance loses 0.5, netting exactly one unit again. Identical outcomes, different vocabulary. Calling it even money removes the sensation of placing a side bet, and that is the only work the phrase does.
The error is in the framing, not the sums
Nobody taking insurance has miscalculated. They have not calculated. What they have done is treat two independent wagers as one position, because losing a 20 to a dealer natural feels like a compound disaster and paying a small premium feels like buying a smaller one. Certainty gets confused with value. Equity in a strong hand cannot be moved onto a side bet that loses money on its own terms, and a wager that cuts variance while lowering expected return is a cost, not a protection.
Which is the quiet problem with low-edge products generally. They attract exactly the sort of player who reads a strategy chart, compares return figures and computes their rebate rate, and that player will still hand back multiples of everything they saved through insurance, side bets and one comfortable click when a dealer shows an ace. The main game is where the arithmetic is public. The leaks are everywhere else.
