The cells everyone gets wrong

A handful of cells account for most mistakes at real tables, because in each of them the correct play feels wrong. The Checkpoint tests these ten and nothing else.

Hit 16 v 10. Standing wins only if the dealer busts, about 23% from a ten. Hitting busts you more often than not and is still less bad, by less than a cent per dollar. It is the closest call on the chart, which is why counters flip it as soon as the count rises. At a neutral table, hit. Stand 12 v 4. Only four ranks bust you, and your 12 beats nothing the dealer makes. Your only win is their 40% bust, and hitting risks that win. Hit 12 v 2 and 12 v 3. The dealer's bust rate drops just enough to flip it. The 12s are the closest calls in the stiff region. Memorize the border: hit against 2–3, stand against 4–6.

Hit soft 18 v 9. You hold 18, and a dealer showing 9 finishes 19 or better almost half the time, so your 18 loses slowly and the draw that cannot bust is free equity. Hit 9 v 2, do not double. The 2 looks weak and is not weak enough. It is the least-busting weak card, and the double loses money. Double A-7 v 2 at our default H17. At S17 tables it is a stand, and the rule-variations Lesson is about knowing which rules you are under. Hit A-6 v 2, never stand. Soft 17 stands nowhere, against anything. Split 8-8 v 10. One 16 against a ten loses about 54 cents per dollar. Two hands from 8 lose less combined, even with double the money out. It feels like putting a second bet on a losing hand. It is paying to get out of the game's worst total. Stand 9-9 v 7. Your 18 beats the 17 a 7 usually makes. Splitting breaks a winning hand to fund two coin flips.

Pass this at 8 of 10 and you know every cell that trips people up. What is left is speed.