When to Split 8s and Aces

Of the ten pair hands in basic strategy, eight have answers that change depending on what the dealer shows. Two do not. A pair of 8s and a pair of aces are split against every dealer upcard, from a 2 through an ace, without exception.

That is unusual enough to be worth understanding rather than memorising, because the case that feels most wrong, splitting 8s against a dealer 10, is the one players get wrong most often.

Aces: the obvious one

A pair of aces played as one hand is a soft 12. That is close to the worst starting total in the game: you cannot bust, but you also cannot get to a competitive number without drawing several times.

Split them and each ace becomes the start of a hand where any ten-value card makes 21. Roughly three cards in thirteen do that. Two live hands with that upside beat one hand stuck at 12 against any upcard, which is why the answer never changes.

8s: the one that feels wrong

A pair of 8s is a hard 16, statistically the worst hand in blackjack. You bust more often than not if you hit, and standing on 16 against a strong upcard loses most of the time.

Splitting does not turn 16 into a good hand. It turns one terrible hand into two mediocre ones, each starting at 8, which is a legitimate base. Against a dealer 10 you will still lose money on average. You will lose less than you would playing 16 as a single hand, and that is the whole argument.

Losing less is the correct goal

This is the part that trips people up. Basic strategy does not always hand you a winning play. On some hands every option loses money and the correct move is simply the one that loses least.

Splitting 8s against a 9, 10 or ace is exactly that situation. Players who refuse to split there because "you are throwing good money after bad" are choosing a worse expected result to avoid the feeling of doubling their exposure on a bad hand.

What about surrender

If your table offers late surrender, 16 against a 9, 10 or ace is a surrender hand, and a pair of 8s is a 16. Where surrender is available, surrendering the 8s against those three upcards is slightly better than splitting.

Most tables do not offer it, and the calculator on this site assumes it is unavailable. Without surrender, split.

Doubling after the split

Both of these plays get better when the house allows doubling after a split, which most do. Split 8s and catch a 3 on one of them for an 11, and you want to double that hand.

The strategy this site uses assumes doubling after split is permitted. If your table forbids it, splitting 8s against a 10 gets marginally worse, though it remains the correct play.

The one rule that never changes

Never split 10s, and never split 5s. A pair of 10s is a 20, which wins on its own more often than two hands starting at 10 will. A pair of 5s is a hard 10, which you should double, not split.

Those two, along with always splitting 8s and aces, are the four pair decisions worth knowing cold. The rest you can look up.

Enter the hand into the calculator to see the recommendation for any pair against any upcard, or browse the full pair splitting chart.

Why Two Hands of Eight Beat One Hand of Sixteen

It helps to look at what each of the two hands actually is, rather than at the total.

A hard 16 has no good continuation. Stand and you win only when the dealer busts. Hit and you go over on any card from 6 upward, which is more than half the deck. There is no card you can draw that makes 16 into a strong hand, only cards that make it into a slightly less bad one.

A hand starting at 8 is entirely different. Every ten-value card takes you to 18, and roughly 30 percent of the deck is ten-value. A 3 gives you 11, which you can double. A 9 gives you 17. There is a wide band of cards that leave you somewhere useful, and almost nothing that ruins you outright, because you cannot bust an 8 with one card.

So the split is not turning a bad hand into a good one. It is trading a hand with no future for two hands that each have one. You are paying a second bet for that, which is why it only makes sense when the improvement is as large as it is here.

The Number Behind "Losing Less"

Against a dealer 9, 10, or Ace, splitting 8s still loses money. It is worth being specific about that rather than pretending otherwise, because the honesty is what makes the play stick.

Playing the 16 against a dealer 10 loses roughly 54 cents on the dollar. Splitting into two hands of 8 against the same upcard loses less than that per unit wagered, which is why the chart says split. You are not converting a loss into a win. You are reducing the size of an expected loss.

This is where a lot of players get off the bus, and the objection is always some version of throwing good money after bad. The response is that the second bet is not being thrown after the first. It is being placed on a new hand with better prospects than the one you would otherwise be playing. The two bets are independent from the moment you split.

If your table offers late surrender, that changes the answer against 9, 10, and Ace: surrendering caps your loss at exactly 50 cents on the dollar, which beats splitting. Most tables do not offer it.

The Aces Case Has One Wrinkle Worth Knowing

Split Aces receive exactly one card each, and then the hands are finished. That restriction is nearly universal, and it is severe: catch a 2 on a split Ace and you are stuck with a soft 13 that you cannot improve.

Splitting Aces is still correct against every single upcard, which tells you how bad the alternative is. A pair of Aces played as one hand is a soft 12. It is close to the least valuable holding in blackjack: too low to stand on, and every hit leaves you somewhere unremarkable.

The second wrinkle is the payout. An Ace and a ten-value card on a split hand is not a blackjack. It counts as an ordinary 21 and pays even money rather than 3 to 2. That feels like a penalty, and it is, and the math behind the split already accounts for it. Split anyway.

Some casinos allow resplitting Aces if you draw a third one, and a few allow drawing more than one card to split Aces. Both rules are player-friendly and both are uncommon. Where they exist they make the play slightly better; where they do not, it remains correct.

The Two Pairs That Look Similar and Are Not

A pair of 4s and a pair of 9s both invite the same reasoning as 8s, and both go the other way, which makes them useful for testing whether you understand the principle rather than the rule.

A pair of 4s is an 8. Splitting gives you two hands starting at 4, which is worse than one hand at 8, since an 8 can be improved by almost anything while a 4 needs two good cards. So 4s split only against a dealer 5 and 6, where the dealer's weakness is extreme enough to justify it, and only where doubling after a split is permitted. Otherwise you hit.

A pair of 9s is an 18. That is already a decent hand, so splitting has to buy something real. Against a dealer 7 you stand, because a 7 most often finishes on 17 and your 18 already beats it. Against 2 through 6 and against 8 and 9 you split, because 18 is not good enough there and two hands starting at 9 have genuine upside. Against a 10 or an Ace you stand, because the dealer is too strong for two mediocre hands to be worth the extra bet.

Both of those show the same underlying test the 8s pass so clearly: split when the hand you are giving up has no future and the hands you are getting do.

Practising the Decision Away From the Table

These four pair rules, always split Aces and 8s, never split 10s or 5s, are the only ones worth knowing cold, and there is a reason to drill them rather than look them up.

Pair decisions arrive fast, they involve putting more money out, and they are the decisions casual players are most likely to talk you out of at a live table. Somebody will tell you not to split 8s against a 10. If you are looking the answer up in that moment, you are more likely to be persuaded than if you decided in advance.

Everything else on the pair chart can be read from a printed card, which most casinos permit at the table. Use one. It is more reliable than half-remembering a cell while three other people wait.

A note on risk. Basic strategy reduces the house edge to roughly half a percent. It does not remove it. Nothing here is gambling advice or a prediction of results. If gambling has stopped being entertainment, the National Problem Gambling Helpline is 1-800-522-4700, or ncpgambling.org.