The One That Matters: 6:5
A natural blackjack traditionally pays three to two: bet $10, get $15. The 6:5 variation pays $12. Nothing else about the game changes, which is exactly why it spreads — the table looks identical and the felt says "Blackjack pays 6 to 5" in the same typeface as everything else.
The cost is straightforward to compute. In an infinite shoe, a two-card blackjack arrives with probability 2 x (1/13) x (4/13) = 4.73% of hands. A small share of those push against a dealer blackjack and are unaffected by the payout, leaving 4.51% that get paid. The payout difference is 1.5 − 1.2 = 0.3 units.
Published figures for a six-deck shoe put this at about 1.39%; the small gap is the infinite-deck assumption, which slightly overstates how often a blackjack appears. Either number leads to the same conclusion. Perfect basic strategy is worth roughly half a point against the house. This rule is worth nearly three times that, against you, and it arrives before you have made a single decision.
There is no strategy adjustment that recovers it. You cannot play your way out of a payout change, because it applies to hands you have already won. The only correct response is to play somewhere else.
The Full Ranking
Effects on the house edge, from most expensive to least. Negative numbers are worse for you.
| Rule | Effect on edge | Source |
|---|---|---|
| Blackjack pays 6:5 | −1.35% | computed here |
| Dealer hits soft 17 (H17) | −0.22% | published |
| No double after split | −0.14% | published |
| Double on 10 and 11 only | −0.18% | published |
| No resplitting aces | −0.08% | published |
| Eight decks instead of six | −0.02% | published |
| Late surrender offered | +0.09% | computed here |
| Dealer peeks for blackjack | +0.11% | published |
Read the order rather than the individual values. Everything below the first row is a rounding error next to the first row. A table that hits soft 17, forbids double after split and uses eight decks is worse than the ideal by about 0.38% — noticeable over a long session, and still less than a third of what a single 6:5 sign costs you.
This is the same shape as the RTP band problem on slots: the visible game is identical, the price is set by a configuration you were not shown, and the difference dwarfs anything your play can control.
Surrender: Four Hands, and That Is All
Late surrender lets you forfeit half your bet and end the hand. It sounds like an escape hatch and it is offered at every hand, which makes it feel more useful than it is.
Surrender returns exactly −0.5000. So it is correct only where every other option is worse than −0.5000. Across all 340 hand-versus-upcard situations, that happens four times:
| Hand | Best play otherwise | Its value | Gained by surrendering |
|---|---|---|---|
| Hard 16 vs 10 | Hit | −0.5398 | +0.0398 |
| Hard 16 vs Ace | Hit | −0.5171 | +0.0171 |
| Hard 16 vs 9 | Hit | −0.5093 | +0.0093 |
| Hard 15 vs 10 | Hit | −0.5044 | +0.0044 |
Weighted by how often those hands actually occur, the whole rule is worth +0.09% — a real gain, and the smallest number on this page. Published figures put it nearer +0.07%; again the infinite shoe accounts for the difference.
A note on 8,8 against a ten. It is not on the list, because the correct play is to split, not to surrender or hit. Splitting returns −0.4895 against −0.5398 for hitting. It still loses. Losing less is the entire game.
Reading a Table in Ten Seconds
Which of These We Computed
Two rows of the table above were derived here, from the same expected-value engine that produces the per-hand data behind all 340 hand pages: the 6:5 payout and the value of late surrender. Both calculations are shown in full in this article so they can be checked.
The rest are published values for multi-deck games, reproduced rather than recomputed. Doing those properly means modelling each rule variant end to end, which is a larger piece of work than this article; when it is done, the code will be published alongside it as the rest has been. Where a figure is someone else's, this site says so.
Reality check: no combination of rules on this page makes blackjack a positive-expectation game. The best commonly available table still holds an edge of about half a percent against perfect play. The point of reading the felt is to pay the smallest available price, not to stop paying.
Play with the math on your side
SlotDrop's free tools show you the real house edge, EV and variance behind every bet — before you place it.
Open the Black Box →Reality check: no strategy on this page turns a negative-expectation game into a positive one. Strategy reduces how much the house takes and how fast your bankroll disappears — it does not guarantee profit. Only ever bet money you can afford to lose entirely.