Most conversations about blackjack mistakes are really conversations about last night. Someone stood on 16, the dealer flipped a 6, and the story ends with a shrug. That anecdote is worthless as evidence, and it is roughly how most players decide what their errors cost them.
There is a better unit of measurement, and it is the one every published figure in this field uses: expected cost per dollar wagered, stated as a percentage. It says nothing about tonight and everything about the long run, which is exactly the right way round.
Cost is measured per dollar, not per hand
When a rule or a decision is said to cost 0.14 percent, it means that over a large number of hands, that choice gives up about fourteen cents for every hundred dollars put at risk. The number is an average, and no single hand ever pays out exactly that amount. It is the same accounting used to state the house edge itself, which is why the two can simply be added together.
The percentages look tiny. They are tiny per hand. What makes them worth caring about is repetition: a player at a normal table sees, on commonly published estimates of table pace, somewhere around sixty to a hundred decisions an hour, and a habit repeated across a few hundred hands stops being a rounding error.
Why outcome bias hides the cost
The reason bad habits survive is that blackjack gives loud, immediate, and completely misleading feedback. A correct play that loses feels like a mistake. An incorrect play that wins feels like insight.
Hitting 16 against a ten is correct and will still bust more often than not. Declining insurance is correct and will feel foolish on the hands where the dealer turns over a ten. Neither result says anything about whether the decision was right, because a decision is a bet on a distribution, not a prediction of a card.
The practical consequence is that you cannot audit your own play from memory. Memory keeps the vivid hands, not the representative ones. The only reliable audit is a record of every decision you made and whether it matched correct strategy, which is what a trainer's accuracy history is for.
The expensive ones, with numbers
Taking insurance. Insurance is not part of your hand. It is a separate side bet that the dealer has blackjack, paying 2:1, and as Wizard of Odds calculates its house edge is about 5.92 percent in a six deck game and about 6.01 percent at eight decks. Compare that with roughly half a percent on the main game and the scale of it is clear: this is one of the worst bets on the table dressed up as prudence. Basic strategy declines it every single time, including on your own blackjack. Only a counter holding a true count around plus three has a mathematical reason to take it, which is covered in our piece on insurance.
Sitting at a 6:5 table. This is the biggest one, and it is not a playing mistake at all. As Wizard of Odds calculates, paying 6:5 on a blackjack instead of 3:2 costs the player about 1.39 percentage points. A table paying even money on blackjack is worse still, about 2.27 percentage points. No amount of correct play compensates for that, because the leak is in the pay table rather than in your decisions. Reading the felt before sitting down is the single highest value habit in the game, and it takes two seconds. Our comparison of 3:2 versus 6:5 goes through the arithmetic.
Standing on 16 against a 10 out of fear. The most common playing error there is. Sixteen against a ten is a bad hand under any choice, and standing does not make it less bad, it just moves the losing to a moment you do not have to watch. Hitting loses less. Where late surrender is offered, surrendering is better still. Our full treatment of 16 versus a ten works through why.
Splitting tens. Twenty is the second best hand in blackjack. Breaking it into two hands that each start at ten trades a near certain strong position for two mediocre ones. It looks aggressive against a dealer 5 or 6 and it is simply worse. Counters split tens in a couple of narrow high count spots, which is a deviation with a published index, not a general license.
Not splitting eights. The mirror image. Hard 16 is the worst total in the game, and splitting turns it into two hands starting at eight, each of which can become something. Basic strategy splits eights against every upcard in most rule sets, and players who refuse to split against a nine or a ten are choosing to keep the worst hand available.
Skipping surrender where it is offered. Late surrender is worth about 0.08 percent to the player, as Wizard of Odds calculates, and only if you actually use it. Many players never surrender because giving up half a bet feels like quitting. On a small set of hands, half a bet is genuinely better than what the hand is worth. See our surrender strategy guide for the short list.
Playing a chart built for the wrong rules. Silent and common. Basic strategy is computed against a specific rule set, and a chart memorized for one game misfires at another. The rules that move it are known quantities: as Wizard of Odds calculates against a baseline of eight decks, dealer stands on soft 17 and double after split allowed, a dealer who hits soft 17 costs the player about 0.22 percent, losing double after split costs about 0.14 percent, and moving down to a single deck gains about 0.48 percent. Memorize one chart, sit at a different game, and you are giving away part of the reason you memorized it.
21 TrainerFree rule-aware charts plus a built in house edge calculator: see what each rule and each error is worth.
Get the app
How imperfect play adds up
It is tempting to want one number for "how much the average player loses to mistakes." That number does not exist in any trustworthy form, because it depends entirely on which errors a given person makes and how often.
There is a useful anchor, though. As Wizard of Odds calculates, the Wizard's deliberately simplified strategy, a shortened chart designed to be easy to memorize, gives up only about 0.14 percent relative to full basic strategy under liberal Vegas Strip rules. That is what a well designed set of approximations costs. Inexpert play, which mixes hunches with half remembered rules and the occasional insurance bet, gives up substantially more than that, though how much more varies far too widely to put a figure on.
The comparison is still instructive. Getting most of the chart right is worth a great deal, and the remaining gap between good and perfect is small. The expensive things are the categorical errors: the 6:5 table, the insurance habit, the refusal to split eights.
The arithmetic that actually matters
Every one of these costs scales with hands played. A quarter of a percentage point is nothing across ten hands and quite real across ten thousand. That cuts both ways, and it is the encouraging half of the story: fixing a habit pays every time the hand comes up again, forever, with no further effort.
Which is why the efficient move is not to re-learn the whole chart. It is to find the eight or ten hands you personally get wrong and drill those. A trainer that records decisions can show you accuracy by hand type and dealer upcard as a heat map, and a Weakness Drill can deal from your worst spots instead of dealing uniformly. That turns a vague sense of "I should study" into a short, finite list.
And when the list is done, be clear about what you have bought. Correct play under good rules shrinks the house edge to a small number, roughly half a percent. It does not create an edge, it does not stop losing sessions, and no habit fixed here changes the fact that variance dominates any single night. What it buys is that you are no longer paying extra for the privilege.
Frequently asked questions
What is the most expensive mistake in blackjack?
Sitting at a table that pays 6:5 on blackjack instead of 3:2. As Wizard of Odds calculates, a 6:5 payout costs the player about 1.39 percentage points, far more than any single playing error you are likely to repeat. It is a table selection decision rather than a decision made with cards in front of you, which is what makes it so easy to overlook.
How much does taking insurance cost you?
Insurance is a separate side bet on whether the dealer has blackjack, and as Wizard of Odds calculates it carries a house edge of about 5.92 percent in a six deck game and about 6.01 percent at eight decks. That is roughly ten times the edge on the main game. Basic strategy declines it every time. Only a counter with a true count around plus three has a reason to take it.
Is standing on 16 against a 10 a mistake?
Yes, under plain basic strategy in a multi deck game, with surrender preferred where it is offered. Both hitting and standing lose money on 16 against a ten. Hitting simply loses less, because standing hands the dealer an uncontested chance to make a total you cannot beat. The exception belongs to counters, whose index plays turn the hand into a stand at a high enough count.
Why do people say a correct play that loses is still correct?
Because a decision is judged by its expected value over many repetitions, not by one card. Hitting 16 against a ten is right and will still bust most of the time. Judging plays by the last outcome is outcome bias, and it is how good players talk themselves into expensive habits. The scoreboard that matters is your decision accuracy, not last night's result.
How do you find out which blackjack hands you get wrong?
Track them rather than guess. A trainer that records every decision can show accuracy by hand type and dealer upcard as a heat map, and drill the hands you are worst at instead of dealing uniformly. Most players find their real error list is short, eight or ten hands, which makes it a very cheap thing to fix once you know what is on it.