Card Counting Strategy Question

Depending on the answers to my following questions, I think there is a way to use card counting much more efficiently than current strategies.

Q1) In any useful card counting method, isn't there a specific true count value where your odds are exactly 50%? (i.e. A true count higher than this value gives the player an edge)

Q2) Assuming this is true, doesn't it make more sense to put down the maximum bet as long as the count is above that number?

The caveat is that you need a substantially larger bankroll to accomodate losing streaks and it might be easier to identify you as a counter. Then again, if no one uses this strategy then it might be more difficult to identify you. Theoretically, I don't see a problem with the concept although I'm surprised I haven't heard about anyone using this technique considering it could offer much higher returns.

Any thoughts?

Sure, but it also makes sense to put down the minimum bet when the count is below that number. Or not to bet at all. If you spread $5-$1000 at a blackjack table the house will immediately ID you as a counter, and then make you flat-bet. Then your advantage goes away.

If you're after total financial gain, the goal isn't necessarily efficiency, it's long-term advantage. A +1% EV for a year is better than a +10% EV for 15 minutes.

Each true count value decreases the house edge by about 0.5%. The break-even point is therefore around +1.

The question is should you always bet the max if the count is +2. If your bankroll is large enough, I suppose, that makes sense. Kelly criterion says that the optimal bid amount is equal to the advantage you have. So, if the max bet is, say $1000, and true count is 2 (for 0.5% advantage), then betting max makes sense as long as your bankroll is 200K or more, otherwise, you are better off with a more conservative bet of 0.5% of your total bankroll.

deleted

Quote: Ibeatyouraces

Q1) ...as the count rises, the player actually LOSES more hands than normal and the frequency of ties increases... Actually its what I believe a "monetary advantage".

Alright so is there a specific value when the "monetary advantage" shifts to the player's favor? If so, doesn't it still make sense to bet the maximum at the table, assuming (of course) that you have a huge bankroll and won't get booted for counting? It still sounds like those are the only two drawbacks to this strategy, probably making it useless but still interesting.

deleted

yes, it is +2

What gives the player his advantage in high counts is that doubles and splits have a much higher winning percentage, due to the fact that the dealer will bust his stiffs more often, and that the most favorable doubles and splits (hard 11/10, split A's, 9's, etc.) will make good hands that much more often. The player won't see much, if any, of an increase in win FREQUENCY, though. He will just be winning his big hands (where he increases his bet AFTER the cards are dealt) more often.

Quote: mkl654321

What gives the player his advantage in high counts is that doubles and splits have a much higher winning percentage, due to the fact that the dealer will bust his stiffs more often, and that the most favorable doubles and splits (hard 11/10, split A's, 9's, etc.) will make good hands that much more often. The player won't see much, if any, of an increase in win FREQUENCY, though. He will just be winning his big hands (where he increases his bet AFTER the cards are dealt) more often.

Actually, AFAIK, the biggest gain in advantage is from increased FREQUENCY of blackjacks.
But I don't understand why is everybody fixated on the frequency of the wins. As long as bankroll is sufficiently large to support the bet size, it doesn't matter how often you win, it only matters how much.

Doubles and splits are great but it is also crucial that a BJ is more probable where you get paid 3:2 and the dealer gets paid 1:1.

weaselman beat me to it

Thanks for these great responses! Two more questions:

Using the Ace-Five counting method and betting strategy proposed by the wizard it sounds like he's talking about the actual count (as opposed to the true count) when he says to double on a count of 2. Is that true? Can anyone suggest a better betting strategy?

Quote: clarkacal

Doubles and splits are great but it is also crucial that a BJ is more probable where you get paid 3:2 and the dealer gets paid 1:1.

Yes. That, too. Though the increase in blackjacks isn't that large, quantitatively, compared to the increase in profitable doubling/splitting opportunities. Plus, you don't have a higher winning percentage in high counts for dealt blackjacks (i.e., they win about 96% of the time either way), but you do have a higher winning percentage for doubles and splits in high count situations.

Quote: mmichelman

Thanks for these great responses! Two more questions:

Using the Ace-Five counting method and betting strategy proposed by the wizard (https://wizardofodds.com/blackjack/wizardcount.html), it sounds like he's talking about the actual count (as opposed to the true count) when he says to double on a count of 2. Is that true? Can anyone suggest a better betting strategy?

I'm sure he's referring to the true count, not the running count (what you refer to as the "actual" count). For example, a running count of +2 would be much more significant if three out of four decks had been dealt than if say, only five or six cards had been dealt. For example, using a four deck shoe, if you had a running count of +5, you would not double if only one deck had been dealt and the threshold number for doubling was a true count of +2, but you WOULD double if two+ decks had been dealt.

There are "better" strategies, in that they are more precise, but the Ace/Five count was always meant to be a first step. It is an improvement over Basic Strategy, but it is nowhere near as effective as even a simple Hi-Lo count. It's more useful as a tool to get you used to the idea of counting and using strategy alterations than as a winning playing system.

Quote: mkl654321

Though the increase in blackjacks isn't that large, quantitatively, compared to the increase in profitable doubling/splitting opportunities.

If the true count is, say, +4, you'll get a bj about once in every 18 hands, compared to 1 in 21 with 0 count. That's a 16.7% increase in frequency.

You'll get a 10 when you double or split (which is usually good, but not always - e.g. soft 15 vs. 4 or split 2s against a 3) about 1/3 of the time compared to 4/13, which is 8.3% increase.

You'll get a hand totalling ten or 11 less often than you would when the count is neutral (I don't see an easy way to estimate the actual frequency right now, but it's obvious it'll be lower because there are more 10 value cards to bump up your total).
The soft hands will be a little bit more likely (2*4/48*32/48=1/9=0.11 vs. 2*9/169=0.107, that's a 2.8% increase) (this includes split aces as well).

So, the increase in blackjacks has by far the biggest impact (also once you've got a bj, you have almost certainly won, or, in the worst case pushed, while all the other cases are not so certain at all, so the real contribution of increase in bjs relatively to all these other things is even greater). I am guessing that next biggest contribution is the increased frequency of dealer busts, but I am too lazy and distracted to actually do that math now - I am sure I have screwed up enough times already in the above estimates :)

Quote: weaselman

If the true count is, say, +4, you'll get a bj about once in every 18 hands, compared to 1 in 21 with 0 count. That's a 16.7% increase in frequency.

You'll get a 10 when you double or split (which is usually good, but not always - e.g. soft 15 vs. 4 or split 2s against a 3) about 1/3 of the time compared to 4/13, which is 8.3% increase.

You'll get a hand totalling ten or 11 less often than you would when the count is neutral (I don't see an easy way to estimate the actual frequency right now, but it's obvious it'll be lower because there are more 10 value cards to bump up your total).
The soft hands will be a little bit more likely (2*4/48*32/48=1/9=0.11 vs. 2*9/169=0.107, that's a 2.8% increase) (this includes split aces as well).

So, the increase in blackjacks has by far the biggest impact (also once you've got a bj, you have almost certainly won, or, in the worst case pushed, while all the other cases are not so certain at all, so the real contribution of increase in bjs relatively to all these other things is even greater). I am guessing that next biggest contribution is the increased frequency of dealer busts, but I am too lazy and distracted to actually do that math now - I am sure I have screwed up enough times already in the above estimates :)

The increase in blackjacks is real, but its effect isn't also magnified by an increase in win frequency for that hand--it wins as often as it always does. Doubles and splits, however, win more often in high counts. In fact, it is more helpful to the player that the DEALER is more likely to draw a ten than that the PLAYER is. The real benefit for double downs and splits is their increased win frequency, which comes both from a increased likelihood of making a good hand, and the increased likelihood of dealer busts. The strength of the latter factor can be seen in the strategy changes for soft doubling; in high counts, you are LESS likely to make a hand when soft doubling A4, A6, etc., but the greatly increased chance of the dealer busting (when showing a low card) offsets that.

Another reason that doubles and splits have such strong effects in positive counts is that you are getting more money on the table in an advantageous situation. You can't do that with a blackjack (unless you're drunk).

Quote: mkl654321

I'm sure he's referring to the true count, not the running count

I thought so too ... until I read the page, and it is written pretty clearly and unambigously referring to the running count, without even a brief mention of the true count anywhere. So, that got me thinking ...

If you assume that tens appear about four times more often than aces, and low cards (2,3,4,6) appear four times as often as five, then the *running* ace-five count is roughly equivalent to the true count about halfway of an eight-deck shoe. So, it seems like a fairly reasonable thing to do to use the running value of the ace-five count as an approximation of the Hi-Low TC in a 6-8 deck game.

Quote: weaselman

If you assume that tens appear about four times more often than aces, and low cards (2,3,4,6) appear four times as often as five, then the *running* ace-five count is roughly equivalent to the true count about halfway of an eight-deck shoe. So, it seems like a fairly reasonable thing to do to use the running value of the ace-five count as an approximation of the Hi-Low TC in a 6-8 deck game.

Very insightful, good catch. I linked the website because I figured people would assume he meant the true count even though I read it like you did, but now at least it makes sense :)

Quote: weaselman

I thought so too ... until I read the page, and it is written pretty clearly and unambigously referring to the running count, without even a brief mention of the true count anywhere. So, that got me thinking ...

If you assume that tens appear about four times more often than aces, and low cards (2,3,4,6) appear four times as often as five, then the *running* ace-five count is roughly equivalent to the true count about halfway of an eight-deck shoe. So, it seems like a fairly reasonable thing to do to use the running value of the ace-five count as an approximation of the Hi-Low TC in a 6-8 deck game.

If the Wiz was referring to running count, it sounds like he was simplifying things for the sake of beginning counters (who would be the only ones to use an Ace-Five anyway). Using your criterion, the difference is actually between tracking forty cards (A,K,Q,J,10,6,5,4,3,2) versus tracking eight cards (A,5). So the appearance of an Ace or 5 in the A/5 count is five times as significant as the appearance of a high or low card in one-level Hi-Lo counts. Therefore, the A/5 running count will diverge in accuracy as the number of decks remaining increases, and become more accurate as the number of decks remaining decreases. The Wiz probably realized that the A/5 running count overemphasized the count early in the shoe, and gradually became more accurate as the shoe was dealt out, but the running count would be right SOME of the time; this was probably better than also introducing the concept of the true count at the same time and having his readers' heads explode.

Correct, my ace-five count is all based on the running count. There is no true count conversion. It is meant to be easy to use, at the expense of making less money compared to more robust count counting strategies. I know from many questions I have received over the years is that true count conversions throw most beginners. In my opinion, it is tough going from basic strategy to a true count strategy like the hi-low. My ace-five is meant to be a bridge between basic strategy and a serious card counting strategy.

About the reason card counting works, please read my introducton' rel='nofollow' target='_blank'>https://wizardofodds.com/blackjack/count/index.html]introducton to card counting. That breaks down the reason it works. More blackjacks, for example, accounts for only 7% of it.

Regarding just betting the maximum when the count is +2 or more, and the minimum otherwise, yes, that would maximize money won per hand. It would not maximize advantage on a percentage basis. It would also be at a huge expense in volatility and raise huge red flags.

Quote: Wizard

About the reason card counting works, please read my .

When you say 40% of the benefit is from standing, does that mean that it's actually attributed to the dealer busting after you stand?