The Hands Nobody Wants, and the Hands You Get Constantly
Key Takeaway: Stiff hands are losing hands no matter what you do. Your job is not to win them but to lose the least. Against a dealer 7 through Ace, standing on 12–16 means you win only when the dealer busts, which happens roughly 17–26% of the time. Hitting is worse in the short run because you bust in front of everyone, and better in the long run because it is the only play that gives your hand a real chance to win. When Late Surrender is available, hard 16 vs 9/10/A and hard 15 vs 10 should be surrendered, which caps the loss at exactly −0.500.
Every player who has sat at a blackjack table knows the feeling. You get a ten and a six, the dealer turns up a ten, and the next decision feels like choosing between two ways to lose. That feeling is accurate. Hard 16 against a 10 is a losing hand under every possible strategy. What most players never learn is that the two ways to lose cost different amounts, and over a playing lifetime that difference adds up to real money.
This guide covers the full stiff-hand zone, hard 12 through hard 16. These are the totals where one more card might bust you, but where standing leaves you below the dealer's minimum finishing total of 17. I will cover how often these hands occur, why the dealer's upcard decides almost everything, why hard 12 is the exception to the rule, and why hitting 16 against a 10 is correct even though it feels wrong. If you have not memorized the full chart yet, start with our learn basic strategy guide and come back. What follows assumes you know the chart exists and want to understand why it says what it says.
How Often You're Stuck: The 38% Problem
Players tend to remember stiff hands as unusually frequent because they are so unpleasant. Here the memory is correct. Using infinite-deck probabilities, where each rank 2 through 9 and the Ace appears with probability 1/13 and ten-valued cards appear with probability 4/13, here are the two-card hard totals from 12 through 16:
P(pair) = p(a)²
Hard 12: T-2 (8) + 9-3 (2) + 8-4 (2) + 7-5 (2) + 6-6 (1) = 15/169
Hard 13: T-3 (8) + 9-4 (2) + 8-5 (2) + 7-6 (2) = 14/169
Hard 14: T-4 (8) + 9-5 (2) + 8-6 (2) + 7-7 (1) = 13/169
Hard 15: T-5 (8) + 9-6 (2) + 8-7 (2) = 12/169
Hard 16: T-6 (8) + 9-7 (2) + 8-8 (1) = 11/169
--------------------------------------------------------------
Total stiff starts = 65/169 = 0.3846 → 38.46% of all opening hands
More than one opening hand in three is a stiff. The count above includes 6-6, 7-7 and 8-8, which you will often split instead of playing as totals. Even with those removed, you still face a two-card stiff on about 36% of hands, and that is before counting the stiffs you draw into from hands like 2-3-10 or 4-2-7.
Each individual stiff-hand decision is often close, but the hands are so common that small errors repeated hundreds of times per session become a large part of the house edge. A player who follows a "never bust" approach and stands on every stiff faces a house edge somewhere around 4%. A basic strategy player in a decent six-deck game faces about 0.5%. Most of that gap comes from the stiff-hand zone.
The Decision Line: Bust Cards vs. Power Cards
When you hold a stiff and stand, your outcome depends entirely on one event: whether the dealer busts. You cannot push, because the dealer never finishes on 12 through 16. You win only if the dealer goes over 21. So your standing equity is set by the dealer's bust probability, and that probability depends heavily on the upcard.
| Dealer Upcard | Dealer Bust % (S17, after peek) | Your EV Standing on Any Stiff | Classification |
|---|---|---|---|
| 2 | ≈35.4% | −0.293 | Weak bust card |
| 3 | ≈37.4% | −0.252 | Weak bust card |
| 4 | ≈39.5% | −0.211 | Bust card |
| 5 | ≈41.7% | −0.167 | Strong bust card |
| 6 | ≈42.3% | −0.154 | Strong bust card |
| 7 | ≈26.2% | −0.475 | Power card |
| 8 | ≈24.5% | −0.510 | Power card |
| 9 | ≈22.8% | −0.543 | Power card |
| 10 | ≈23.0% | −0.540 | Power card |
| A | ≈16.7% | −0.667 | Strongest power card |
Look at the drop between 6 and 7. The bust rate falls from about 42% to about 26%. The reason is that a dealer showing 7 has a 4/13 chance that the hole card is a ten, which makes a pat 17 immediately. A dealer showing 6 who catches that same ten has 16 and must draw again. Small upcards force the dealer into stiff totals of their own. Big upcards usually make a pat hand.
With a stiff, standing EV equals exactly (2 × P(dealer bust)) − 1. You win the bust cases and lose every other case. No push is possible. Against a 10, that works out to 2(0.230) − 1 = −0.540. Standing on 16 against a 10 means you have already accepted losing about 77% of the time. Players think of standing as avoiding risk. In fact it locks in a worse outcome and simply hands the bust to the dealer, where it happens quietly.
The Hard 12 Anomaly
The basic strategy rule for stiffs is short: stand against 2–6, hit against 7–A. Hard 12 is the one exception. You hit 12 against a 2 and a 3, and stand against 4, 5 and 6.
The reason comes down to two numbers. Hitting hard 12 busts you only on a ten-valued card, which is 4/13 = 30.8%. Every other card (A through 9) gives you 13 through 21 with no chance of busting. Hard 12 is the least risky stiff to hit. Meanwhile, the dealer showing a 2 busts only about 35% of the time, and with a 3 about 37%. Those are the two weakest bust cards on the board.
So against a 2 or 3, standing gives up your hand's improvement potential in exchange for a dealer bust that is not much more likely than your own bust when hitting. The EVs bear this out:
- 12 vs 2: Stand −0.293, Hit −0.253. Hitting saves about 4.0% of your bet.
- 12 vs 3: Stand −0.252, Hit −0.234. Hitting saves about 1.9%.
- 12 vs 4: Stand ≈ −0.211, Hit ≈ −0.211. The difference is a fraction of a hundredth of a unit. Standing is marginally correct in a neutral shoe, and this is one of the most count-sensitive plays in the game.
- 12 vs 5 and 6: Standing is clearly correct, because the dealer's bust rate of over 41% beats anything hitting can achieve.
Now compare 13 vs 2. Your bust risk when hitting rises to 5/13 = 38.5%, because a 9 now busts you as well. That one additional bust card is enough to make standing correct (Stand −0.293, Hit ≈ −0.309). The boundary between hitting and standing falls right between 12 and 13 against a 2, and it is decided by a single rank.
The Hard 13–16 Protocol
From 13 upward the rule is simple:
- Against 2 through 6: stand. The dealer will bust 35–42% of the time, and every extra point in your hand increases your bust risk by 1/13 (13 busts 38.5%, 14 busts 46.2%, 15 busts 53.8%, 16 busts 61.5%). Let the dealer take the risk.
- Against 7 through Ace: hit (or surrender where indicated below). The dealer will make a pat hand roughly 74–83% of the time. Standing loses all of those. You need to draw toward a hand that can actually win.
Many players accept this against a 10 but hesitate against a 7, which is a mistake. Hard 16 vs 7 is where hitting adds the most value of any stiff decision.
STAND: EV = 2(0.262) − 1 = −0.475
HIT: Draw A→17, 2→18, 3→19, 4→20, 5→21 (each p = 1/13)
Draw 6,7,8,9,T → bust (p = 8/13)
EV(17 vs 7) ≈ −0.107 EV(18 vs 7) ≈ +0.400
EV(19 vs 7) ≈ +0.616 EV(20 vs 7) ≈ +0.774
EV(21 vs 7) ≈ +0.927
EV(hit) = (1/13)(−0.107 + 0.400 + 0.616 + 0.774 + 0.927) − 8/13
= (2.610 / 13) − 0.6154
= 0.2008 − 0.6154 = −0.415
Hitting gains ≈ 0.060 units per hand (6.0% of the wager)
You bust 61.5% of the time when you hit, which feels terrible. But the 38.5% of the time you make a hand, you are often the favorite, especially with 18 through 21 against a dealer whose most likely result is 17. Standing gives you just a 26% chance to win and nothing else.
The Notorious Hard 16 vs 10
This is the most debated decision in blackjack, and it deserves the attention because it is genuinely close. In a multi-deck game, with a two-card 10-6 against a dealer 10 (after the dealer checks for blackjack), the numbers are:
STAND: EV = 2 × P(bust | no BJ) − 1 = 2(0.230) − 1 ≈ −0.540
HIT: EV ≈ −0.535
Edge from hitting: 0.005 units (0.5% of wager)
Frequency check (T-6 or 9-7 vs 10, infinite approximation):
P(non-pair two-card 16) = 10/169 = 0.0592
P(dealer upcard 10) = 4/13 = 0.3077
P(spot) ≈ 0.0182 → ~1 hand in 55
At $100 flat, 100 hands/hr: 1.82 × $100 × 0.005 ≈ $0.91/hr saved by hitting
Over 2,000 hours: ≈ $1,820 kept in your pocket
Half a percent sounds negligible, and on any single hand it is. You will often hit, bust, and watch the dealer flip a 6 and bust. That does not change the math. Over millions of shoes, standing on this hand gives up money with no compensating benefit.
Two refinements matter here. First, in an infinite deck the gap shrinks to roughly 0.2%. The multi-deck 10-6 gap is wider because the two cards you hold are both cards that would bust you if you hit, and they are no longer in the shoe. Second, if your 16 is made of three or more small cards (for example 4-4-3-5 or 2-4-10), you have removed low cards that help a hitter, and standing becomes correct in multi-deck games. Our composition-dependent hands guide covers these exceptions in full. The underlying derivations are laid out cleanly in the stiff hands loss minimization data on BlackjackMath.org, which is worth working through line by line.
Benchmark Table: The Stiff-Hand Decisions That Matter Most
| Hand vs Upcard | Stand EV | Hit EV | Surrender EV | Best Play (LS available) | Cost of Standing |
|---|---|---|---|---|---|
| 12 vs 2 | −0.293 | −0.253 | −0.500 | Hit | 0.040 |
| 12 vs 3 | −0.252 | −0.234 | −0.500 | Hit | 0.019 |
| 12 vs 4 | −0.211 | ≈−0.211 | −0.500 | Stand (razor-thin) | ≈0.000 |
| 13 vs 2 | −0.293 | ≈−0.309 | −0.500 | Stand | — (hit costs 0.016) |
| 16 vs 7 | −0.475 | −0.415 | −0.500 | Hit | 0.060 |
| 16 vs 8 | −0.510 | ≈−0.458 | −0.500 | Hit | 0.052 |
| 16 vs 9 | −0.543 | ≈−0.509 | −0.500 | Surrender | 0.043 |
| 16 vs 10 | −0.540 | −0.535 | −0.500 | Surrender | 0.040 |
| 16 vs A | −0.667 | ≈−0.517 | −0.500 | Surrender | 0.167 |
| 15 vs 10 | −0.540 | ≈−0.502 | −0.500 | Surrender (marginal) | 0.040 |
The "Cost of Standing" column shows how much a frightened player gives up compared with the best available play. Standing on 16 against an Ace is the worst of these errors, at nearly 17% of the wager. Plenty of players freeze in exactly that spot because the Ace looks intimidating, and that hesitation is the costly part.
Surrender: Paying Half to Leave the Table
Late Surrender lets you give up half your bet after the dealer has checked for blackjack. The EV is exactly −0.500, no more and no less. The rule for when to use it follows directly: surrender whenever your best playing option is worse than −0.500.
- Hard 16 vs 9: Hitting is about −0.509. Surrender saves roughly 0.9% compared with hitting and 4.3% compared with standing.
- Hard 16 vs 10: Hitting is −0.535. Surrender saves 3.5% compared with hitting. This is the most valuable routine surrender because the spot is so common.
- Hard 16 vs A: Hitting is about −0.517. Surrender saves about 1.7%. In H17 games the case for surrendering is stronger still, and 15 vs A and 17 vs A also become surrenders.
- Hard 15 vs 10: Hitting is about −0.502. Surrender gains only a fraction of a percent, but the gain is real, and over thousands of occurrences it adds up.
If you have 8-8 against a 10 or Ace, split it in most S17 games. The pair rules take priority over the hard-16 rules. For the complete rule-set breakdown, including Early Surrender and H17 adjustments, see our late surrender guide.
Sharp Tip (Marcus Vance): Before you sit down, ask the dealer "Surrender allowed?" and read the felt for H17 or S17. Those two rules change your stiff-hand plays more than any other conditions. When you surrender, say the word clearly and draw a horizontal line behind your bet with your finger. Dealers sometimes miss quiet requests, and a misread can turn a −0.500 surrender into a −0.540 forced stand. If you are using a Hi-Lo count, memorize four indices that cover most of the value in this zone: stand 16 vs 10 at true count 0 or higher, stand 15 vs 10 at +4, stand 12 vs 3 at +2, and stand 12 vs 2 at +3. Build a printable version for your exact game in the Strategy Chart Builder.
Count and Composition: Where the Lines Move
Stiff decisions are among the most count-sensitive plays in blackjack because they sit so close to the breakeven point. A high true count means more tens remain in the shoe. That raises the dealer's bust rate and also raises your own bust risk when hitting, and both effects favor standing. A negative count does the opposite. At −1 you hit 13 vs 2. Below 0 you hit 12 vs 4. Hard 16 vs 9 becomes a stand at about +5. If you are not counting, ignore these and play the neutral chart. If you are counting, the 16 vs 10 index alone is worth more than many of the obscure plays players spend time memorizing.
Composition matters even without a count. In single-deck and double-deck games, the exact cards in your hand change the probabilities of what remains. A 10-2 against a 4 in single deck is a hit, because the 10 you hold is a bust card that is no longer available to the dealer. Treat these as second-tier refinements. Learn the neutral totals perfectly first.
Conquering the Table: Peer Pressure and Busted 16s
Now for the hardest part: hitting 16, drawing a king, and having the player at third base groan because "that was the dealer's bust card."
That claim is mathematically false. The next card in the shoe is unknown to everyone, and its probability distribution is identical whether you hit or stand. Your hit is just as likely to remove a dealer-saving 5 as a dealer-busting 10. Across all possible card orders, other players' decisions have zero effect on your expectation or theirs. Anchorman complaints are memory bias. People remember the time you "took the bust card" and forget the times you "took the dealer's five."
Here is how to handle it in practice:
- Decide before the cards come out. Know your play for every stiff against every upcard so the decision is automatic and nobody's opinion enters into it.
- Judge decisions, not outcomes. A correct hit that busts is still correct. A wrong stand that wins was still a mistake that happened to pay.
- Don't argue at the table. Shrug, say "that's the book," and play on. You are not going to teach anyone combinatorics in the middle of a shoe.
- Remember the scale. Stiffs make up 38% of your hands. Giving in to social pressure on each of them would cost you more than almost any rule variation the casino could impose.
Stiff hands will never feel good, and they aren't supposed to. You are dealt a losing position more than a third of the time, and the skill is in losing exactly as little as the math allows. Stand against the bust cards, hit against the power cards, treat hard 12 as its own case against 2 and 3, and surrender where the rules allow and the numbers favor it.
