The hands worth arguing about
Every blackjack table has the same arguments, and they are always about the same handful of hands. It turns out that is not a coincidence: the spots people argue over are, almost exactly, the spots where the correct play barely wins.
A hand-versus-upcard spot is on this list for one of two reasons, both measured rather than chosen: the correct play beats the runner-up by less than 2% of your bet, or the correct play actually changes across the 6 rule sets modelled here. The other 319 spots are not close and not contested — Hard 20 vs 8 is decided by 1.64 times your bet, which is not an argument, it is arithmetic.
Ranked by how little the right answer is worth
| Hand | Correct | Margin over runner-up | Your return |
|---|---|---|---|
| Hard 16 v 10rules decide it | Hit | 0.06% | -54.0% |
| Soft 15 v 4rules decide it | Hit | 0.09% | 5.9% |
| Soft 18 v 2rules decide it | Stand | 0.20% | 12.2% |
| Hard 12 v 4 | Stand | 0.24% | -21.1% |
| Pair of 3s v 2rules decide it | Split | 0.26% | -13.8% |
| Soft 17 v 2 | Hit | 0.65% | -0.1% |
| Soft 18 v Ace | Hit | 0.73% | -9.3% |
| Soft 13 v 5 | Hit | 0.74% | 13.3% |
| Pair of 4s v 5rules decide it | Split | 0.95% | 8.0% |
| Pair of 3s v 8 | Hit | 1.25% | -21.7% |
| Hard 9 v 2 | Hit | 1.33% | 7.4% |
| Soft 14 v 5 | Double | 1.41% | 12.6% |
| Pair of 2s v 8 | Hit | 1.48% | -15.9% |
| Hard 13 v 2 | Stand | 1.50% | -29.3% |
| Soft 16 v 3 | Hit | 1.63% | 0.9% |
| Soft 19 v 6rules decide it | Stand | 1.64% | 49.6% |
| Hard 12 v 6 | Stand | 1.68% | -15.4% |
| Pair of 7s v 8 | Hit | 1.71% | -37.2% |
| Soft 13 v 6 | Double | 1.80% | 18.0% |
| Soft 16 v 4 | Double | 1.84% | 5.8% |
| Hard 12 v 3 | Hit | 1.86% | -23.4% |
| Hard 9 v 3 | Double | 1.95% | 12.1% |
| Pair of 2s v 2rules decide it | Split | 2.60% | -8.9% |
| Pair of 4s v 6rules decide it | Split | 3.10% | 14.6% |
| Hard 16 v 9rules decide it | Hit | 3.38% | -50.9% |
| Hard 11 v Acerules decide it | Hit | 3.39% | 14.3% |
| Pair of 6s v 2rules decide it | Split | 3.48% | -21.9% |
| Hard 15 v 10rules decide it | Hit | 3.60% | -50.4% |
| Pair of 3s v 3rules decide it | Split | 4.34% | -6.4% |
| Pair of 2s v 3rules decide it | Split | 5.70% | -2.6% |
| Hard 16 v Acerules decide it | Hit | 14.99% | -51.7% |
Margins are expected return per unit staked, under S17 / DAS / 3:2. A margin of 0.06% means playing the “wrong” way costs you that much of your bet, on average, every time the situation comes up.
Famous hands that are not actually close
These get argued about too, and they should not. Each one has a reputation for being a hard call and none of them is — the gap is wide enough that the decision is settled.
Pair of 8s v 10 Splitting 8s against a ten looks like throwing good money after bad. | Split by 5.03% over hit |
Pair of 10s v 6 Splitting tens against a weak dealer is the classic greedy play. | Stand by 12.84% over split |
Pair of Aces v Ace Splitting aces into an ace feels like asking for trouble. | Split by 12.96% over hit |
Hard 11 v 6 Doubling 11 is the one every table agrees on — and it is agreed on for a reason. | Double by 33.37% over hit |
A close call is a call the shoe can take away
There is a second way to measure how contested a hand is, and it agrees with this one. Work out the Hi-Lo true count at which each of these decisions reverses, and rank the hands by how far the count has to move — that order matches the margin order above at a Spearman correlation of 0.76. The narrower basic strategy’s win, the thinner the shoe has to get before it stops being right at all. Hard 16 against a ten, tightest on this page, reverses 0.44 of a true count from neutral — near enough to even that an ordinary shoe crosses it constantly. The card counting page has the crossover for every hand here.