2 hands instead of one...

Many thanks if anyone will answer this ... I know this person reads these pages, I will direct him here, but this may be good for all to see.

=====INSERT LETTER==========
I am reading in all my books that when you decide to
split to 2 hands, you are supposed to put 70% of what
you would normally bet on each of the two hands.
However, I dont understand the logic of doing this.
Why not just bet what you would normally bet but play
two hands instead of one?

for example: the book says that for a $100 bet on 1
circle- if you decided to split to two hands, you
would bet $70 on each circle.

What is the problem of betting $100 on each circle
instead of $70 on each circle? I dont understand the
logic of this... Why do books always tell me when
splitting to two hands to bet 70% of what i would
normally bet on each of the two hands, instead of
betting 100% of what i would normally bet on two
hands?

Is there any harm of betting 100% of the normal bet on
each hand when splitting instead of 70%? What are the
effects of doing each scheme? Variation differences?

The kelly crit suggests that you bet proportional to your br to get the best trade-off for risk and return.

1%ev you bet 1% of br.

However, in blackjack, because sometimes the best strategy you have to split and shit, there is kind of an extra variance. After you made the correct bet, you might or might not split, double and do some other special moves.

An extreme case example is when you get 100% edge and bet 100% but damn.....you get 8,8 vs 4. You got no cash to split.

So.....after factoring in this stupid thing, ideally, normally, you bet 70% of the correct kelly.

Now 2 hands......why 70% of the 70% in each again? This is another extra variance thing, because now you have 2 hands which might split ot not but might not be at the same time.

You calculate how much they will vary n-sync, lose or win together, which is called the covariance.

Because of the extra variance, you have to bet less on each hand to get the correct kelly.

Just kidding. smile.

Also just kidding. :-)

First, I'm assuming the $100 you might have bet on one hand is "optimal" for you, in some sense. In other words, you've figured out your bankroll, and your tolerance for risk, and calculated your bet ramp based on this. Or maybe you're just following advice you read in one of the better books on the subject. Regardless, you have some reason for believing $100 is the right amount.

Now let's think about the two hand problem carefully. Suppose you knew for sure (with metaphysical certitude) that both hands would win or lose together. You can't know that, of course, but just suppose, if hand A wins you KNOW hand B will win, and if A loses, you KNOW B loses, and for pushes - likewise. Now what sense would it make to put $100 on both spots? None. It would be no different than putting $200 on one spot, and you already knew you weren't going to do that. In fact, the only logical thing to do, in this case, is to place $50 on each spot, or else go back to plan A and stick to one hand.

Now let's turn it around. Suppose there was NO relationship whatsoever between the two hands. Suppose they were at different tables with the same TC. Now you have practically the same situation as two consecutive hands at one table where the TC stays the same. Unless you're strapped for cash, there's no earthly reason, in this scenarios, to bet less than $100 on both hands.

The situation in real life is somewhere in between. You don't know that your two hands will win or lose in perfect sync, yet there is some correlation between them. For instance, if the dealer gets blackjack, both hands are very probably dead meat. OTOH, if the dealer has a 5 up and busts, it's PARTY TIME!

Mathematicians have this all decked out and buttoned up. They say two hands on the same round have a certain "covariance." You can read about this in Griffin's Theory of Blackjack, if you like. Or you can be like 85% of the rest of us who just follow instructions and keep our mouths shut! (Not good to move your lips while counting anyhow...)

L:U:C:K
ETF

which leads to an additional question. Do you know if covariance has an affect when playing at multi-action BJ table where you play 2 or 3 hands at the same time. In other words, is it more advantageous to play at a multi-action table than at a regular BJ table due to covariance? or maybe the question is wouldn't it always be better to play 2 hands rather than 1? (I have been playing at a multi-action table in AC with the usual 8 decks, and varying the number of hands played & my bet amount based on hi-low count & wonging out at -2). Bonne Chance.

"which leads to an additional question. Do you know if covariance has an affect when playing at multi-action BJ table where you play 2 or 3 hands at the same time. In other words, is it more advantageous to play at a multi-action table than at a regular BJ table due to covariance?"

See BJA, p. 43-44.

Don