Resplitting aces is valuable, but not valuable enough to rescue an otherwise poor blackjack game. Under common multi-deck assumptions, allowing RSA—resplitting aces—is often worth roughly 0.03 percentage points of house edge. The exact figure changes with deck count, dealer rules, split limits and what happens after the split.
That makes RSA a tie-breaker between similar tables, not the first rule to inspect. A 3:2 blackjack payout, dealer standing on soft 17, surrender and double after split usually have a larger effect.
The rare sequence that activates RSA
The rule matters only after a specific chain of events:
- You receive two aces.
- You split them and place a second equal wager.
- One of the new hands receives another ace.
- The table either allows that new pair to be split again or forces it to remain as a restricted split-ace hand.
Without RSA, the new ace often leaves a hand such as A-A that receives only one additional card and cannot be played normally. With RSA, the pair can become two more starting hands, each built around an ace.
The rule does not change the probability of receiving the original pair. It improves the value of the unusual occasions when another ace arrives after the first split.
Why an ace is worth separating
Two aces played together begin as soft 12. That is an awkward total: hitting can produce a useful hand, but the two aces are more valuable when used as separate starting cards. Each split ace has many ways to become a strong total with one additional card.
Most tables restrict split aces more heavily than other pairs. Common restrictions include:
- one card only to each split ace;
- no hitting after that card;
- no doubling on a split-ace hand;
- ace plus a ten-value card paying as ordinary 21 rather than blackjack;
- no resplitting, or a cap of three or four total hands.
RSA removes only the resplitting restriction. It does not automatically remove the others. A table may allow a second split while still dealing only one card to each resulting ace.
For the full procedure rather than its mathematical price, see resplitting aces and blackjack splitting rules.
Turning the rule into dollars
Suppose two otherwise identical games have these estimated house edges under correct basic strategy:
| Rule set | Estimated house edge |
|---|---|
| No resplitting aces | 0.50% |
| Resplitting aces allowed | 0.47% |
The illustration assumes RSA improves the game by 0.03 percentage points. Expected loss is:
Expected loss = total action × house edge
For $10,000 of total wagering:
- without RSA:
$10,000 × 0.0050 = $50; - with RSA:
$10,000 × 0.0047 = $47.
The expected difference is $3 per $10,000 wagered. That does not mean RSA pays you $3, nor that one session will finish $3 better. It means the long-run average cost is slightly lower if every other assumption stays the same.
The example also shows why percentage points must not be confused with percentages. Moving from 0.50% to 0.47% is a change of 0.03 percentage points, not three percentage points.
RSA is not the same as hitting split aces
Three rules are often blended together even though they have different values:
| Rule | What it permits |
|---|---|
| Resplit non-ace pairs | Split a new pair formed after an earlier split |
| Resplit aces | Split A-A again after the original ace split |
| Hit split aces | Draw additional cards after the first card on a split ace |
Hitting split aces is generally more powerful than merely resplitting them because it converts the restricted hand into a playable hand. It is also much less common. A rules sign saying “resplit to four hands” is incomplete unless you know whether aces are included.
The Massachusetts approved blackjack rules illustrate this discretion: a licensee may permit repeated splitting while separately prohibiting aces from being split more than once, provided notice is given. The same rules retain the one-card restriction on split aces. See the Massachusetts blackjack rules for the jurisdiction-specific wording.
Published blackjack rule comparisons commonly place RSA near a three-hundredths-of-a-percentage-point improvement under standard assumptions, while emphasizing that the value is ruleset-dependent. The blackjack rule-variation calculations provide one established reference point.
What to compare before choosing the table
Read the rules in this order:
- Blackjack payout. A 6:5 payout is usually far more damaging than RSA is helpful.
- Dealer soft 17. Standing on soft 17 is better for the player than hitting it.
- Doubling rights. Check which totals can be doubled and whether double after split is allowed.
- Surrender. Late surrender can matter more than RSA.
- Deck count and dealing procedure. These affect both edge and strategy.
- Split restrictions. Only after the larger items should RSA decide between close alternatives.
A table with RSA, 6:5 blackjack and dealer hitting soft 17 is not a good game because one minor favorable rule appears on the placard. Compare the full package with house edge by blackjack rules or the blackjack rule variation calculator.
The decision RSA changes
There is no new strategic debate when the opportunity appears. If basic strategy calls for splitting the original aces and the rules permit another ace pair to be split, the additional split is normally the value-producing action. The practical problem is not memorizing a complicated play; it is recognizing whether the table allows it and budgeting for the extra equal wager.
A $50 original hand can become $100 after the first split and $150 after one resplit. The mathematical improvement does not remove the short-term exposure. RSA is better for expected value, but it can create more simultaneous money at risk when the rare sequence occurs.
The useful description is therefore precise: RSA slightly reduces the house edge and occasionally increases the number of active hands. It is a favorable rule, not a guarantee, a betting system or a substitute for checking the rules that affect every round.