Blackjack myths, tested rather than repeated.

Blackjack tables come with advice: blame the player at third base, wait for a hot dealer to cool, assume a ten in the hole, never bust. We put each one through our engine, by exact calculation or by dealing millions of rounds, and kept score.

Most of them fail. The bad player at third base changes nothing for you: over 20 million rounds a table, your result was −0.66% alone and −0.67% beside a player who hits every stiff. Two are only half true: cards do come due, a little, and even money is sure money, at a price.

See the nine myths Deal the third-base test

Nine myths, measured

What the tables say, and what the cards do.

A typical six-deck table throughout: the dealer hits soft 17, blackjack pays 3:2. The simulated figures come from 20 million rounds a table; the rest are exact.

False

The bad player at third base costs the table

Your result: −0.66% alone · −0.67% beside a player who hits every stiff · ±0.05%

The card third base takes is unknown to everyone until it is turned, so taking it is no more likely to rob the dealer of a bust than to save the dealer from one. On average, the dealer draws the same cards either way. Deal the test yourself.

False

After a few losses you are due, and a hot dealer stays hot

Next round won: 43.3% after any round · 43.3% after four losses in a row

Each round is dealt from what is left in the shoe, not from what happened on the last one. Losing runs happen, and they predict nothing. The full count.

Half true

Cards come due

A ten next: 30.8% in a fresh shoe, 31.8% after ten small cards · your edge −0.67% → +0.19%

Results do not carry over, but cards do: every small card that leaves makes the rest of the shoe a little richer in tens. That is what card counting tracks. It moves slowly, and the shuffle resets it.

False

Assume the dealer has a ten in the hole

A ten underneath 30.9% of the time · 33.1% behind a 10 once the dealer has checked for blackjack

More than two hole cards in three are not tens. Basic strategy already prices every card the dealer could have, so a rule of thumb that assumes the worst only throws away the cases it ignores.

False

A dealer showing a 5 or 6 will bust

Busts 42.0% with a 5, 43.9% with a 6 · makes a hand 56.1% of the time with a 6

They are the dealer’s worst cards, which is why you stand on stiff hands against them, but even a 6 makes a hand more often than it busts. Every upcard.

False

Play it safe: never bust, or play like the dealer

House edge: never bust 5.98% · copy the dealer 5.89% · basic strategy 0.67%

The dealer’s edge comes from one rule: you act first, so when you both bust, you lose. Copying the dealer keeps that disadvantage, and neither style doubles or splits, which is where basic strategy wins its money back.

Half true

Even money is a sure thing, so take it

Even money pays $10 on $10 · playing on is worth $10.39 on average

It is a sure $10, but even money is insurance in another form: with a blackjack against an ace, turning it down is worth 39 cents more per $10 on average.

False

Perfect basic strategy beats the house

The house keeps 0.67%, or 0.46% where the dealer stands on soft 17

Basic strategy is the best you can do without tracking cards, and it still loses, slowly. Only a count can push the edge your way, and only briefly.

False

Card counting is illegal

In your head: legal · with a device: a crime in Nevada and many other places

Casinos can refuse to deal to a player they think is counting, or shuffle earlier, but keeping track of the cards in your head breaks no law. Our card counting guide covers what it takes.

The player at third base

Third base takes a card, not your odds.

The complaint goes like this: third base hits a 16 against the dealer’s 6, takes the 10 that would have busted the dealer, and the whole table loses. Sometimes that is exactly what happens. Just as often the card is small, and the 10 left behind is the one that busts the dealer.

That card is unknown to everyone until it is turned, so taking it is no more likely to rob the dealer of a bust than to save the dealer from one. We dealt 20 million rounds at each table below, from the same shuffled shoes, with you playing basic strategy in the first seat. Your result barely moved, and every gap is inside the simulation’s margin of ±0.05%.

What a bad player costs is their own money. Hitting every stiff costs them 6.2% of every bet, and never risking a bust 6.0%. Where you sit does not matter either: at the table where both seats played basic strategy, third base averaged −0.67% and you −0.68%.

Your result per hand, 20 million rounds a table

Alone at the table −0.66%

Third base plays basic strategythird base’s own result −0.67% −0.68%

Third base hits every stiffthird base’s own result −6.19% −0.67%

Third base never risks a bustthird base’s own result −5.97% −0.68%

You play basic strategy in the first seat. Every table is dealt from the same shuffled six-deck shoes, three-quarters of each shoe before the reshuffle. Margin ±0.05%.

Deal it yourself

A million rounds at each table, dealt in your browser from fresh shuffles. Run it more than once: the two figures trade places, and neither stays ahead.

Third base

You, alone at the table–

You, beside a third base who hits every stiff–

Press the button to deal. Each figure is your average result per hand.

Streaks and due cards

Luck doesn’t carry over. Cards do, a little.

From the table where you played alone: every round after a losing run, and what happened next. A losing run changes nothing, whether you call it being due or the dealer running hot; the small differences are the noise of fewer rounds.

Your next round after a losing run: how often you win, push and lose, and how many rounds the figure rests on
Before the round	You win	Push	You lose	Rounds
Any round	43.3%	8.7%	48.0%	20,000,000
After two losses in a row	43.3%	8.7%	48.0%	4,612,035
After four losses in a row	43.3%	8.7%	48.0%	1,062,107
After six losses in a row	43.2%	8.7%	48.1%	244,411

A win or a loss is the round’s net result, doubles and splits included. The rarer the run, the fewer rounds behind its row, so the last row can wobble by a few tenths.

What does carry over: the cards

A shoe remembers what has been dealt. Each small card that leaves makes the rest a little richer in tens, which helps the player; each ten or ace that leaves does the opposite. The effect is real, it is small, and it vanishes at the shuffle. Tracking it is card counting, not a hunch.

Next card a ten, fresh six-deck shoe: 30.8%
After ten small cards have gone: 31.8%
After twenty small cards have gone: 32.9%
Your edge, ten small cards gone: +0.19%
Your edge, eight tens and two aces gone: −1.62%

The dealer’s upcard

A weak card, not a certain bust.

Two of the most repeated table rules are about the dealer: a 5 or 6 means a bust, and there is always a ten underneath. Here is how often each is true, for every upcard.

For each dealer upcard: how often the dealer busts, how often the dealer makes a hand, and how often a ten is underneath once the dealer has checked for blackjack
Upcard	Dealer busts	Makes a hand	A ten underneath
2	35.7%	64.3%	30.9%
3	37.7%	62.3%	30.9%
4	39.9%	60.1%	30.9%
5	42.0%	58.0%	30.9%
6	43.9%	56.1%	30.9%
7	26.2%	73.8%	30.9%
8	24.4%	75.6%	30.9%
9	22.9%	77.1%	30.9%
10	23.0%	77.0%	33.1%
Ace	20.1%	79.9%	None

Six decks, the dealer hits soft 17. With a 10 or an ace showing the dealer checks for blackjack first, so the figures are for the hands that play on: behind a 10 the hole card cannot be an ace, behind an ace it cannot be a ten.

Even a 6 makes a hand more often than it busts: 56.1% against 43.9%. You stand on 12 to 16 against it because standing loses less than hitting would, not because the dealer is sure to break.

Questions about blackjack myths.

Does a bad player at the table hurt my odds?

No. Over 20 million rounds a table, your result playing basic strategy was −0.66% alone, −0.67% beside a third base who hits every stiff and −0.68% beside one who never risks a bust, all inside the margin of ±0.05%. Their choices change which cards you see, not the odds of them.

Does it matter where I sit?

No. At the table where both seats played basic strategy, the player at third base averaged −0.67% and the player in the first seat −0.68%, the same within the simulation’s margin.

Is the dealer more likely to bust with a 6 showing?

Yes, it is the dealer’s worst card: a bust 43.9% of the time, against 35.7% behind a 2 and 23.0% behind a 10. But the dealer still makes a hand 56.1% of the time with a 6.

Should I assume the dealer has a ten underneath?

No. A ten is underneath 30.9% of the time behind a 2 to 9, and 33.1% behind a 10 once the dealer has checked for blackjack. Play the chart, which allows for every card.

Are blackjack streaks real?

Runs of wins and losses happen all the time; they just predict nothing. After four losses in a row you win the next round 43.3% of the time, the same 43.3% as after any round.

Is it better never to bust?

No. Standing on every hard 12 or more and never doubling or splitting costs 5.98% of every bet, against 0.67% for basic strategy. Busting sometimes is the price of making hands that beat the dealer.

Before you take table advice

Test it, don’t repeat it.

Three things the tests say.

Other players don’t cost you

Over 20 million rounds a table, your result stayed between −0.66% and −0.68%, whoever sat at third base and however they played.

Streaks don’t predict

After four losses in a row you win the next round 43.3% of the time, the same as after any round. Only the cards carry over, and only a little.

Safe play is expensive

Never busting costs 5.98% of every bet and copying the dealer 5.89%, against 0.67% for basic strategy.