Insurance is the most misunderstood option at the blackjack table. It sounds like protection, but mathematically it is a separate side bet on the dealer having blackjack — and for players who are not counting cards, it loses money over time. Here is the arithmetic, in plain terms.
What is insurance in blackjack?
Insurance is an optional side bet offered when the dealer’s face-up card is an Ace. It costs half your original wager and pays 2:1 if the dealer’s hole card gives them blackjack. It is not protection for your hand — it is a wager on one specific dealer outcome, settled before the hand continues.
How insurance works, step by step
- You bet C$20. The dealer deals your two cards and shows an Ace.
- The dealer offers insurance. You may stake up to half your bet — C$10.
- The dealer checks the hole card.
- If the dealer has blackjack: insurance pays 2:1, so C$10 returns C$30 (C$20 profit). Your main C$20 bet loses. Net result: you break even.
- If the dealer does not have blackjack: you lose the C$10 insurance immediately, and the hand plays on with your original C$20 still at stake.
That break-even outcome is why insurance feels safe. The problem is how often case 5 happens compared with case 4.
How much does blackjack insurance pay?
Insurance pays 2:1. Because it costs half your main bet, a winning insurance bet exactly offsets the loss of a full main bet — hence the “break-even” impression. Note that if you hold a blackjack yourself and the dealer shows an Ace, you may instead be offered “even money”, which is the same bet in a different wrapper.
What are the odds of blackjack insurance?
For insurance to be a fair bet at 2:1, the dealer would need blackjack one third (33.3%) of the time. In reality the dealer completes blackjack roughly 31% of the time, because only ten-value cards (10, J, Q, K) do it — 16 of every 52 cards. That gap between 31% and 33.3% is the house edge, and it is why insurance is a long-term loser for most players.
Expected value by situation
The figures below assume the dealer shows an Ace and you have seen only your own two cards. Break-even requires a 33.3% chance of a ten-value hole card.
| Situation | Ten-value cards left | Chance dealer has blackjack | Expected value per C$1 insured |
|---|---|---|---|
| Break-even point (reference) | — | 33.3% | C$0.00 |
| Six-deck shoe, no tens in your hand | 96 of 309 unseen | ≈31.1% | ≈ −C$0.07 |
| Single deck, no tens in your hand | 16 of 49 unseen | ≈32.7% | ≈ −C$0.02 |
| Single deck, you hold two ten-value cards (a 20) | 14 of 49 unseen | ≈28.6% | ≈ −C$0.14 |
The pattern is consistent: insurance is always negative for a basic-strategy player, and it gets worse the more ten-value cards you can see in your own hand. Insuring a 20 is the single worst version of the bet — precisely the hand players most want to “protect”.
Should you ever take insurance?
As a basic-strategy player, no — decline it every time. The only players for whom insurance becomes profitable are card counters who know the remaining deck is rich in ten-value cards; when the true count is high enough that ten-values exceed one third of the unseen cards, the bet flips positive. That is a narrow, advantage-play condition, not general advice. Also decline “even money” on your own blackjack for the same reason: taking a guaranteed 1:1 gives up the higher expected return of letting the hand stand.
Insurance vs other blackjack decisions
Insurance is often confused with two genuinely useful options. Splitting divides a pair into two hands with a second equal bet — correct with aces and eights. Doubling down doubles your bet for exactly one more card — correct on strong totals such as 11 against a weak dealer card. Both are strategy plays that improve expected value in the right spots; insurance is a side bet that does not.
Wildz runs Evolution live tables alongside more than 2,000 slots, and pays 10–20% cashback weekly at only 1x wagering.
Where to play and learn
Practise the decision in low-stakes or demo games before it costs you anything — see our online blackjack hub. If you prefer games without strategy decisions, compare roulette or bingo instead.
Play responsibly
Understanding a bet’s expected value is useful, but no strategy removes the house edge. Set deposit and loss limits, and treat blackjack as entertainment rather than income. Free support: ResponsibleGambling.org. Minimum age 19+ in most Canadian provinces, 18+ in Alberta, Manitoba and Quebec.
Frequently asked questions
Is insurance in blackjack ever worth it?
Only for card counters when the remaining deck is unusually rich in ten-value cards. For everyone else it carries a negative expectation on every hand.
What does insurance pay in blackjack?
2:1. Since it costs half your main bet, winning it offsets the loss of that bet exactly — which is why it appears to break even.
Is “even money” the same as insurance?
Effectively yes. Taking even money on your own blackjack against a dealer Ace is mathematically the same as fully insuring it, and it is also negative expectation.
Should I insure a 20?
No — it is the worst case. Holding two ten-value cards removes two of the cards the dealer needs, pushing the expected value further against you.
Why insurance feels right and is wrong
Insurance is a well-designed bet — designed, that is, for the casino. It arrives at the exact moment you feel most exposed: you have money on the table and the dealer is showing an Ace. The word itself borrows credibility from a product most people consider prudent. And the immediate outcome, when the dealer does have blackjack, is that you lose nothing, which the brain files as a success.
What the brain does not file is the far more frequent outcome: roughly seven times in ten the dealer does not have blackjack, the insurance stake disappears silently, and the hand continues as though nothing happened. Because that loss is small and undramatic while the win is vivid and memorable, players systematically overestimate how often insurance pays. The arithmetic below is the correction.
Why the break-even point is exactly one third
Insurance pays 2:1. For any bet paying 2:1, you need to win once for every two losses simply to stand still — that is a win rate of one in three, or 33.3%.
You can verify it with a stake of C$1 across three hypothetical rounds where you win exactly once: you gain C$2 on the winner and lose C$1 twice, netting zero. Any win rate above 33.3% is profitable, anything below is a loss. The dealer’s actual rate of completing blackjack from an Ace is about 31%, and that 2.3-point gap is the entire house edge on the bet.
Why the dealer only makes blackjack about 31% of the time
With an Ace showing, the dealer needs a ten-value card underneath — a 10, Jack, Queen or King. There are 16 such cards in a 52-card deck, which is 30.8% of the pack. Once your own two cards and the dealer’s Ace are visible, the exact figure shifts slightly depending on what those cards were, but it hovers around 31% and never reaches the 33.3% the payout requires.
This is the whole story. The bet is priced as though ten-value cards made up a third of the deck, and they make up slightly less than a third. Everything else — deck count, your hand, the count — only moves the number a little around that centre.
The counter-intuitive part: your good hands make it worse
The instinct is to insure when you have something worth protecting, and that instinct is precisely backwards. Every ten-value card in your own hand is one fewer available to the dealer, which pushes the probability of dealer blackjack down and the house edge on insurance up.
So a 20 — the hand players most want to protect — is the worst possible hand to insure. As the table earlier in this article shows, insuring a 20 in a single deck carries roughly seven times the disadvantage of insuring a hand with no ten-value cards. If you take one thing from this page: never insure a 20.
Even money is the same bet in a suit
If you hold a blackjack and the dealer shows an Ace, you may be offered “even money” — a guaranteed 1:1 payout instead of playing the hand out. It sounds like a different, safer offer. It is arithmetically identical to fully insuring your blackjack.
Playing it out, you win 1.5× your stake when the dealer does not have blackjack (about 69% of the time) and push when they do (about 31%). Expected return: roughly 1.035× your stake. Taking even money guarantees exactly 1.0×. You are paying about 3.5% of your stake for the comfort of certainty. Decline it for the same reason you decline insurance.
When insurance genuinely becomes correct
There is one legitimate exception, and it is narrow. Card counters track the ratio of high to low cards remaining. When the deck becomes rich enough in ten-value cards that they exceed one third of the unseen cards, the probability of dealer blackjack crosses the 33.3% break-even and insurance flips positive.
In practice this is signalled by a high true count, and insurance is one of the more profitable individual decisions available to a counter precisely because it is a clean, high-edge bet when the condition is met. Two caveats matter for ordinary readers: this requires accurate counting through the shoe, not a gut feeling that “lots of small cards have come out”; and continuous shuffling machines, which many online and land-based tables now use, remove the condition entirely. In online blackjack against an RNG that reshuffles every hand, the count never develops and insurance is always negative.
What to do instead at the table
The decisions that actually improve your expected return are unglamorous and well documented.
- Learn basic strategy properly for the specific rule set you are playing. It reduces the house edge to well under 1% in good games.
- Check the blackjack payout. A table paying 6:5 instead of 3:2 costs you roughly 1.4% — far more than any decision you will make during a hand. Walk away from 6:5.
- Prefer fewer decks and rules such as dealer stands on soft 17, double after split, and late surrender where available.
- Skip the side bets. Perfect Pairs, 21+3 and their relatives carry house edges several times that of the main game.
- Decline insurance and even money every time, unless you are counting in a game where the count exists.
Online versus live-dealer blackjack
Against a software RNG the shoe is effectively reshuffled continuously, so no count develops and insurance carries its full negative expectation on every hand. Live-dealer blackjack uses physical cards, but most studios shuffle well before the shoe is deep enough for counting to be practical, and camera angles plus betting-time limits make it impractical anyway. For all realistic online play, treat insurance as a bet to decline without thinking about it.
More frequently asked questions
Does insurance protect my original bet?
No. It is a separate wager on the dealer’s hole card. Your original bet wins or loses on its own merits regardless of the insurance outcome.
How much can I insure?
Up to half your original bet. Insuring the full half is the standard offer; smaller amounts are usually allowed but do not change the underlying maths.
Is insurance worth it if I have a lot of money on the hand?
No. The expected value is negative per dollar staked, so a larger hand means a larger expected loss on insurance, not a stronger reason to take it.
Do the odds change with more decks?
Slightly. More decks dilute the effect of the cards you can see, pushing the probability closer to 30.8% and making insurance marginally worse than in a single deck.
Why do dealers sometimes seem to encourage insurance?
Dealers are required to offer it when an Ace shows; it is a procedural step, not advice. A good dealer will not tell you how to play, and the house benefits when you accept.