6:5 or 3:2? The split-minimum table

One configuration, run through our bet-spread engine. The plus-count ramp is held fixed. The only thing that changes between the options below is which bet you make in the low counts, and each option plays exactly one hand, so they all deal cards at the same rate and the comparison is direct.

As a rule, 6:5 is worth avoiding. A 3:2 blackjack pays one and a half times your bet; a 6:5 blackjack pays one and a fifth. A natural comes about once every twenty-one hands, and over many hands that payout gap costs a flat 1.4% of every dollar wagered on the 6:5 line. That is the standard 6:5 penalty.

The trade on this table is different. In a minus count you are betting into a disadvantage, so you want the smallest bet available. Here that is $5, and it costs 6:5 to place. The question is whether putting out half the money is worth the worse payout. Halving a $10 bet to $5 only saves money when the disadvantage is steeper than the 1.4% penalty you take to do it. That sets the break-even: move to the $5 / 6:5 line once your 3:2 edge is worse than about −1.4%.

Here is the same $5 / 6:5 and $10 / 3:2 bet placed on identical shoes at each low true count, with the better of the two marked. “Edge/$” is your per-dollar edge on the 3:2 line; “Δ” is what one round earns (loses less) by choosing the $5 line instead of the $10 line.

Turn that into a rule, “move to the $5 line at or below true count X,” and run each rule end to end against the same flat-$10 game. “Δ vs flat $10” is each rule’s gain over always betting the $10 minimum, measured on the same shoes so the small numbers are reliable.

The end-to-end rules match the count-by-count table. Moving to the $5 line at true count −1 and below is the strongest option, +$0.01 a round over flat-betting the $10 minimum; waiting until true count −2 is nearly identical. Moving too early, at true count 0 while you are barely behind, loses ground (−$0.01 a round), because you pay the 6:5 penalty to shrink a bet that was not costing you much. Moving too late, only in the deep counts, leaves a little on the table.

The reader had a second idea: in the low counts, spread to several hands of the cheap $5 / 6:5 line, up to a full table of seven, to use up the bad cards faster and reach the shuffle sooner. We ran the same engine with two through seven hands in the negative counts, all else equal.

Measured per round, it looks harmless. The win rate per round barely moves from one hand to seven, all of it near $0.23 a round. But the per-round comparison is the wrong one. A seven-hand round wagers seven times as much and deals about seven times as many cards, so it is not the same unit of play. A dealer pitches cards at a roughly fixed rate, so the real limit on a table is cards per hour, not rounds per hour. The honest question is how much you earn per card dealt.

Per card dealt, every extra hand costs you, and steadily. Going from one hand to seven drops the win rate from about $22.78 an hour to $17.06 at the same deal speed, roughly a quarter of the total. The reason is simple: the shuffle arrives at the same point in the shoe however you split the cards among hands, so playing more of them does not bring the good cards any sooner. It only puts more money on the negative-count hands you were trying to get past. Spreading only in the deepest counts costs less, but only because it rarely happens.

Multi-hand does not even buy a calmer ride for the lower return. The hourly swing is about the same either way, so the only effect is that the edge takes longer to overcome it: N₀ rises from about 520 hours on one hand to 920 on seven. There is no true count at which spreading to extra $5 / 6:5 hands helps. If you want fewer negative-count hands, the way to get them is to sit out, not to play more of them.

Everything above assumes you will play every hand, sitting through the minus counts so that the only question is which chip to put out. That assumption is the costly part. The most profitable option in a negative count is not the $5 line or the $10 line. It is no bet at all.

Counters call this Wonging out: backing off, sitting out, or leaving the table when the count turns negative, and betting only when the shoe is rich. In the count-by-count table, every row at true count 0 and below is a loss. The $5-vs-$10 decision only picks the smaller of two losses. Sitting those rounds out with no money on the felt turns each of them into $0, which is better than either bet. You keep the full plus-count ramp, where the profit comes from, and stop paying to play the dead counts.

The scale is the point. Playing the low-count bet perfectly instead of flat-betting the $10 line is worth about a cent a round. Not betting the negative counts at all removes the loss from those rounds entirely, and they fall across roughly a quarter of all hands. That is a far larger amount than the bet-sizing choice. If you can back-count a shoe and enter only when it turns positive, or get up and leave, that is the better play.

You cannot always Wong out. You may need cover, the table may be the only one open, a back-counted shoe may never turn positive, and entering and leaving repeatedly draws its own attention. When you do have to play the minus counts, the $5-vs-$10 question is worth getting right, and that is what the rest of this post covers.

Bet the $10 / 3:2 minimum at true count 0 and above, and move to the $5 / 6:5 line at true count −1 and below. That is the best you can do with the low counts on this table.

Keep the size of this in perspective. Playing the low counts perfectly instead of flat-betting the $10 minimum is worth about $0.01 a round, on the order of $1.13 an hour. It is worth doing, but it is not where the table is won or lost. Two things matter more:

• Do not take 6:5 on a bet you want to have out there. Moving to the $5 line helps only because it shrinks a bet you would rather not place. The meaningful money goes on the $10 line at 3:2 in the plus counts. Keep it there.
• The ramp does the work. Nearly all of the profit on this table comes from the plus-count bets. The 6:5-vs-3:2 choice in the low counts is small by comparison: worth getting right, not worth much deliberation.