Just out of curiosity...

...Why are bets not simply raised or lowered SLIGHTLY on EVERY hand based on what cards have come out on just the previous hand? As opposed to raising or lowering "units" only when a "count" changes?

My first thought is that no single hand in itself changes the advantage of the deck significantly enough in itself. I know its the cummulative effect that makes counts change and thus you raise or lower units accordingly. But what if you started with a bet that allowed for upward and downward "creeping" on either end, based on what happens on every hand. Say you start with a $12 bet. Then the first hand is played. You and the dealer both push with 20's on two card tens. Now knowing the deck is ever slightly for the worse from this current moment, you go "one below" your starting bet and now bet $11 on the second hand. Let's say that a couple two's and a few 4's come out between you and the dealer. You gather that that was a rather positive hand on your behalf as far as shifting the deck back again. So on the third hand you raise a dollar and now bet $12. On the third hand, you get a five card 20. Through awareness of several small-mid cards having just been removed, you now raise just one and bet $13 on the fourth hand. Etc etc... In effect, you are trying to "tweak" bets in proportion to just one hand rather than in "unit increments" whenever a count gets to a proper level.

Now I'm sure there's probably a million good reasons why this doesn't work, that is right under my nose. My first guess is that positive or negative changes in a deck by one hand alone doesnt translate into nice $1 bet shifts, so your bets would be "mis-aligned" to what the deck really is. Still, what technically would happen if you tailored your bets BY THE HAND, literally, as opposed to changing them only when you hit a count number? Besides probably looking silly and ending up broke... LOL.

Answered Your Own Question

You are correct in your above thoughts. You are describing a very weak form of counting that won't get an advantage.

:joker::whip:

Essentially what you could do is actually count, but not round. So if you are in a shoe, and there are 3 decks left, let's say the running count is +8. So the true count is 2.67. You'd bet exactly $36. Then you get 2 10's. Next hand, the count is +6, with 2.9 decks left, which is a true count of 2.07 (which might be an optimal bet of $31).

So you can be more precise and granular of your actual counts.

Yet several things:

1. The end result won't change that much.
2. Doing those calculations in your head will be very tricky and time consuming, and you'd get much better results playing faster with rounding (you'd get to your N0 quicker, versus trying to lower your N0 by not rounding).
3. You mentioned sort of "eyeballing" the count, but you may not be accurate enough to know how much to raise or lower your bet, which could have negative effects.

I'm sure that's true but I'm still not quite sure why it wouldn't be following what the deck is doing each step of the way (at least semi-accurately). I can see how it would be inprecise to truly COUNTING the cards yet it also seems that in another sense, it would be more "manageable" or fine-tuned to just adjust bets in small proportion to what happens as each isolated hand goes along. Assuming the tiny bet raises or decreases could be at least roughly estimated. Maybe THAT'S the ultimate question - what is an average positive or negative hand worth in bet terms for the NEXT hand?

I'm sure all of that would be too simple if it could be figured, so it must not work. Yet it still seems it would be a much closer/better estimate than simply flat-betting regardless of hand outcomes, or just betting low on one hand and wild on another regardless of what hand just transpired.

I'll give you a counter example:

Let's say the true count is -10. It is very very bad. There are many low cards left.

So you are betting $10.

Then, you see a few low cards come out. You think that the count must be a little bit positive now, so you bet $15, or $11, or whatever is >$10.

Yet you don't know that the low cards came out because the count was so low, and it still might be at -9 or -8. And now you are raising your bet when you definitely shouldn't be (i.e. you're overbetting).

The low cards coming out means the count might have increased a bit, but it doesn't mean it isn't still very negative.

On the contrary, the count may be +10 and you have $100 out. So you get a blackjack and deduce that you should decrease your bet to $90. But the count is still +9, and you are now underbetting!

Ah yes. Thanks! That makes sense now. I figured it must be something right under my nose that I just wasn't getting.

So I'd actually have to keep a "running count" of good hands VERSUS positive hands for this to vaguely work. As one would if they were literally counting the cards specifically, and more accurately.

That's what we call card counting

Instead of remembering the running count, you may infer the previous running/True count from your last bet..

That's an intriguing idea. I never considered that! So in other words, if I was fortunate enough to have witnessned, say, five "positive hands" in a row and I had lightly raised my bet up from the $5 minimum in proportion on each hand, I might now be betting, say, $10-13 on this sixth hand. Now ON this sixth hand, I get a NEGATIVE makeup of cards. But because I know I'm betting at a $10-13 level, I can "infer" that no negative hand should cause me to lower that bet for the seventh hand (because I'm, thus far, still in a fairly positive situation). If I'm understanding this correctly?

Although, I'd still have to keep mental "running counts" to remember when I've seen enough negative hands to where that $10-13 bet is now OVERBETTING into a neutral or, worse, negative situation. If I'm following this correctly?

Kelly Theory

While accounting for variance:

If you have a 1% advantage you bet 1% of bankroll; regardless of table min or max's, one bets what they can.

:joker::whip:

You are right not to assume that every counting methodology has been tried. Unlikely, but you may be the one to find a new and better approach. But the way to go about it would be to run your system through a simulator, or if that will not work, create (or have created) a simulator that will handle your system. Without testing your ideas, it's all conjecture, although I think you got some really good feedback in the comments in this thread.

I'm sure there are better ways yet to be discovered. KO and TKO have only been out a few years. Spanish 21 was once thought unbeatable. The beat goes on. Keep an open mind, and maybe you'll be the one to find the next breakthrough.

True

Correlation with advatage is light. No counting system can give you 100% betting co-relation with the advantage because of human mind weakness. If you are betting 10-13$, not you may lower it to 9-12$ after 7th round.. This will work better in fewer decks, as you'll be playing with running count.. This can easily beat any single deck game.. Try it in single deck
and then keep increasing/decreasing your bet based on count in the last round. You shall be good..

Hi
Fun at 21,
could not resist to comment on this, in my view you are describing the premise behind progressive betting, which relies on the hand just played, if this is a winner then the next bet is raised, we presume that there are runs of positive and negative hands, and we ride the positive ones, if this hand is lost then we revert back to the minimum bet.
Perhaps not scientific but it works.
Elkobar..

...some of the time.

So does betting fifty times your min bet at random times! You can make a fortune if you guess right several times in a row. And if you parlay the amounts, you will be living on easy street. :grin:

Hi Elkobar,

Thanks for the reply although I was thinking more in terms of "following the count" hand by hand based on if the last hand was beneficial to the deck (low cards/high quantity hands) or non-beneficial (several high cards came out). In such a scenario, one could actually win a given hand but the hand may have been bad (for the current state of the deck). For example, you might win a hand with a five card 21 but it was negative for the state of the deck (because several low cards may have been exhausted in the process). So I wouldn't think this is progression betting in the traditional sense. My original thought/curiosity was just in wondering if a shoe couldn't be betted on by a hand by hand basis which mirrors the true but approximate count at any and all given moments. Although, it was pointed out that such an approach may have limited value since at certain points, a hand could have been negative for the deck yet you still shouldn't be lowering your bet, or vice versa. Basically, the idea is if one can "count cards" through hand QUALITIES without truly counting the cards literally. It's a concept that naturally jumped out to me as a Basic Strategy player just looking to merely even out his inherent disadvantage without expecting any clear-cut edge of course. I then learned that authors like Dubey had long ago studied and explored the same ideas - that just the TYPE of hand played (double downs, splits, pushes, high quantity hands) can tell you
if the deck is now more positive or negative because of it. Of course, it appears there is limited effect to following the "count" in such a general fashion. Still, for a maximumly-playing Basic Strategy player, the concept still seems potentially helpful to me, as opposed to simply not knowing such information from each hand. Rather than ALWAYS flat-betting 100% of the time despite the hand types that may or may not be appearing.

Oops... I just realized in my example above this that I had the good/bad thing mixed up. What I meant to say is that you could WIN a hand with a 21 but the hand may have been "bad" for the deck because you and the dealer exhausted three face cards and an Ace in the process. Obviously a 5 card 21, as in my example above, would be a GOOD sign (typically) for the current deck, not bad. Regardless whether you won or lost that hand in absolute terms. So it's never betting based on a win or loss but betting based on what KIND of hand just happened. At least that was the original idea anyways.

Hi
Fun at 21, yes I understand what you are saying, but the cards do not know what you are doing, and you still can get runs of positive and or negative hands, regardless of what the count is, so in reality we play each hand according to our system and hope that the outcome is in keeping with our strategy.
We place an lot of reliance upon our particular form of mathematical theory, and forget that somtimes the cards do not fall the same way.
Elkobar..

That's true. I didn't mean to imply that ensuing hands would always work out just because I took note of a particular hand type and slightly changed my bet. Surely the cards will do what they do. In that sense, the cards don't know or care what real counters do either. Since one can be in an extremely rare but favorable count and still lose most hands. But like real counters, except less effective, I was just speaking in terms of following the state of the deck hand by hand and betting on the odds when they might be in favor (based on what has been observed). If a deck happens to get in a state where I might currently have, say, a 54% edge, you will still lose roughly half the time in such a situation, yet I would assume betting something slightly more than flat (or minimum) would be preferred on average and in the long run. Of course, I would just be somewhat "guessing" or inferring such an advantage if following something vague like hand types. Likewise, you could still hit a winning streak where you're at a 56% disadvantage on a hand, then win again while it's at a 58% disadvantage, then win again and it's now at a 61% disadvantage. Yet the long run odds would dictate that we'd still bet a minimum at all times if/when we know we dont have a current advantage in the deck makeup. At least that's what I've always been led to believe...

This has happened to many times to remember. Winning streaks with the min bet seems to rule at times.

Yeah, I guess Murphy's Law can be alive and well even in Blackjack...LOL

Or, on the other side, it seems all too common to watch a bunch of low cards or high quantity hands come out, only to lose 20-19 etc to the dealer on all those remaining face cards and aces...LOL

Oh well. As mentioned, I guess it's believing in the LONG run average that hopefully will win/even out in such situations.

Oh yeah, losing with 20 on max bets happens too. It is no doubt you will come out ahead if counting properly without too many mistakes. It helped me tremendously to keep an up to date journal of my playing activity. As I went back and looked at the results over time, it was a confidence builder.