Insurance Bets in Blackjack: When to Avoid Them
The dealer slides an Ace out. Chips stop. The table goes quiet. “Insurance?” the dealer asks. Regulars shake their heads. A new player leans forward. It sounds safe. It sounds smart. It is not, most of the time.
Insurance is a side bet. It feels like a shield. It is not about your hand. It is a guess on the dealer’s hidden card. The math looks kind at a glance. It is not. This guide shows why, when it can flip, and how to make the call in ten seconds.
Ten-second answer: If you do not count cards, skip insurance every time. Even when you have a blackjack and the dealer offers “even money,” say no.
First, what insurance really is
Insurance is a separate bet. You can bet up to half of your main bet. You only get this choice when the dealer shows an Ace. If the hidden card is a 10, J, Q, or K, the dealer has blackjack. If that happens, the insurance bet pays 2 to 1. If not, you lose the insurance bet. Your main hand then plays as normal. See the trick? Insurance is not tied to your 16, your 20, or your blackjack. It is only tied to the dealer’s hole card.
“Even money” is the same thing but dressed up. You have a natural blackjack. The dealer shows an Ace. The dealer offers you “even money.” If you say yes, you get paid 1x your bet now. But the math is the same as taking insurance on your blackjack. You only win that extra bet if the hidden card is a 10-value. The label changes. The math does not.
So the right way to think is simple. Ask: “What is the chance the hole card is a 10-value right now?” If that chance is less than one in three, the bet is bad. The payoff is 2:1. The break-even point is one win in three tries (33.33%).
Math break: why the house wants you to say “Yes”
In a fresh shoe, the chance the hole card is a 10-value is about 30% to 31%. That is lower than the 33.33% you need to break even on a 2:1 bet. This gap is the edge for the casino. Here is a clear walk-through of the mathematics of blackjack insurance if you want to see the full detail.
The exact chance moves a bit with the number of decks. With more decks, the chance goes down a hair, and the bet gets a bit worse. With fewer decks, it gets a bit better but is still bad for you on average. Pros still skip it unless the deck is packed with 10s. That is where counting comes in. More on that soon.
Insurance EV by deck count (fresh shoe)
How to read this: EV (expected value) is the long-run average win or loss per unit bet. A negative EV means you lose on average.
| 1 |
16/51 ≈ 31.37% |
3p − 1 ≈ −5.88% |
≈ 5.88% |
+3 |
| 2 |
32/103 ≈ 31.07% |
≈ −6.80% |
≈ 6.80% |
+3 |
| 4 |
64/207 ≈ 30.92% |
≈ −7.25% |
≈ 7.25% |
+3 |
| 6 |
96/311 ≈ 30.87% |
≈ −7.40% |
≈ 7.40% |
+3 |
| 8 |
128/415 ≈ 30.84% |
≈ −7.47% |
≈ 7.47% |
+3 |
Note how the bet gets a touch worse as the deck count goes up. That is why you should not take insurance by default in big shoes. Pros only take it when the deck has many 10s left. That is a card counting call, not a hunch.
If the idea of EV is new to you, this short expected value primer is a simple way to see how the math plays out over time.
Why “protection” feels good but is not
The word “insurance” is clever. It makes you think you protect your 20 or your blackjack. But your hand does not matter here. The side bet only cares about the dealer’s hole card. You might win the bet and still lose your main hand later, or the other way round. It is two separate games at once.
Our brains hate swings. We want to lock wins. We want to avoid pain. That is why “even money” sounds sweet. But short runs can trick us. Two or three wins in a row can feel like proof. It is not. The long run tells the truth. Avoid the gambler’s fallacy explained by thinking in odds, not in streaks.
When the math can flip: a pro-only corner
There is one clear case when insurance can be good. When you count cards, and the true count is high enough, there are more 10s left in the shoe than normal. The bet can then have a positive EV. The basic rule most pros use is simple.
Take insurance only if the Hi–Lo true count is +3 or more. This is the common “index” for this bet. You can see a clean walk-through of the insurance index for Hi–Lo card counting here. If you are not counting, there is no signal. Skip it.
Also note this key point. The size of your main bet does not change the EV of insurance. A big main bet can tempt you to “protect it.” Do not. The side bet has its own math. Make each bet on its own merit.
- Not counting cards? Always say no to insurance and to even money.
- Counting with a solid true count? If TC ≥ +3, you may take insurance. If not, decline.
- See many small cards on the felt? That hint is not enough by itself. Trust the count, not the vibe.
“Even money” on your blackjack: the quiet trap
Let’s do the math fast. You have a blackjack. The dealer shows an Ace. You can take even money (get 1x now), or you can wait.
If you decline even money, your outcomes are two. If the hole card is not a 10-value, you get paid 3:2, which is +1.5 units. If the hole card is a 10-value, it is a push (0 units). Let p be the chance the hole card is a 10-value. In a fresh 6-deck shoe, p ≈ 0.3087. Your EV if you decline is (1 − p) × 1.5 ≈ 1.5 − 0.3087 × 1.5 ≈ 1.5 − 0.4630 ≈ 1.037 units. That beats the flat +1 unit from even money. So the best play for non-counters is to say no.
If you do count and your TC says insurance is good, then taking insurance (or “even money” on your blackjack) can be correct. But the rule is the same as above: follow TC ≥ +3, not your fear.
Bankroll and swings: how insurance can sting twice
Insurance adds another bet to your hand. That means bigger swings. Your stack will bounce more, and you will tip into risk-of-ruin sooner if you add bad-EV bets on top. A high-volatility side bet with a negative edge is a double hit. You get more pain with no gain.
If you like to think in steady growth terms, read up on the Kelly criterion overview. Kelly says you should only risk money when the edge is in your favor. With no edge, the best bet size is zero. That is a neat way to remember the insurance rule too.
Why casinos pitch it, and what table rules matter
Dealers offer insurance fast and with a smile. It is a profit center. Many players take it. Some think it is a smart, safe move. It is not safe in the math sense. It only feels that way in the moment. Casinos also know that people hate seeing a blackjack push. “Even money” sells well because of that feeling.
Rules matter for your base game too. A table that pays 6:5 on blackjack is bad value. Skip it. 3:2 is the standard you want. Single-deck games can be better but watch for harsh rules. Shoe “penetration” (how deep the dealer deals before shuffling) can help counters but does not fix bad pay tables. For background on how rules shaped the game, see the UNLV Center for Gaming Research on blackjack rules.
Picking where to play (so you can focus on good choices)
Good play starts before you sit down. You want clear rules, fair payouts, and no hard sell on weak side bets. You also want clear bonus terms if you play online. If you need wagering requirements explained in plain words, that guide helps you see what you must bet through before you can cash out. When you know the house rules and the bonus rules, it is easier to stick to the right blackjack choices, like skipping insurance.
Quick FAQ
Is insurance ever a good bet?
Yes, but only if you count cards and your true count is high. With the Hi–Lo system, most pros use a TC of +3 as the point to take it. Below that, skip it.
Should I take even money on my blackjack?
No, not if you are not counting. It pays less in the long run than waiting for the normal 3:2 payout. “Even money” is just insurance in disguise. You can read a clear proof in this breakdown of why even money is just insurance in disguise.
Does the size of my main bet change the insurance math?
No. The insurance bet stands on its own. A big main bet does not turn a bad side bet into a good one.
Is insurance better in single-deck games?
It is a little less bad there, but still negative on average in a fresh deck. You still need a strong true count to flip it.
Do basic strategy charts ever say “take insurance”?
No. Basic strategy is for non-counters. It never calls for insurance.
Does insurance change online?
No. The math is the same. Live or RNG, the odds do not care about the format.
A short note on terms
- EV (expected value): the average result you would get if you made the same bet many times.
- True count (TC): the running count adjusted for decks left. It tells you how rich the shoe is in high cards.
- 3:2 vs 6:5: blackjack payout rates. 3:2 is better for you.
Common mistakes to avoid
- Taking insurance right after a big win “to protect it.” The bet still has the same EV.
- Taking even money “to lock profit.” It cuts your long-run win on blackjacks.
- Thinking your hand total changes the insurance bet. It does not.
- Letting a short streak guide your next choice. Trust the odds.
Sources and further reading
For the nuts and bolts on insurance math and break-even points, Michael Shackleford’s analysis is a gold standard. For counting context, Blackjack Apprenticeship gives clear, field-tested rules. For EV basics, Khan Academy is a friendly start. For risk and bet sizing, Investopedia’s Kelly primer is a good plain guide. For rule history and research, UNLV’s archive is deep.
Help is there if gambling stops being fun. For tools and support, see the National Council on Problem Gambling.
TL;DR
Do not take insurance unless you are counting and your true count is +3 or more. Say no to even money when you have a blackjack if you are not counting. Keep your edge by keeping it simple.
|
