Variance, EV, Risk on Multiple Hands

C

calleo

Member
I know this has been discussed here:
http://www.blackjackinfo.com/bb/showthread.php?t=3069

But want to confirm some things.

My game is S17, RS4, DAS, ES.

Let's say my 1 unit bet is already double the table minimum. Should I be betting 2 hands @ 1/2 unit each? Will this keep my EV the same and reduce my risk? variance?

Also to confirm when we are talking about risk are we talking about Risk of Ruin? Or variance? Or both?

As I ramp up my betting am I better off opening more boxes instead of increasing the bets on one (or two) box(es)?

Also if I'm playing 1-10 betting spread, and playing multiple hands once I reach max bet, does that mean if I'm betting multiple hands my max bet should be 10 divided by the number of hands on each hand? Or is it 10 per hand?
 
A

assume_R

Well-Known Member
calleo said:
Let's say my 1 unit bet is already double the table minimum. Should I be betting 2 hands @ 1/2 unit each? Will this keep my EV the same and reduce my risk? variance?
Click to expand...
Correct.

calleo said:
Also to confirm when we are talking about risk are we talking about Risk of Ruin? Or variance? Or both?
Click to expand...
The equation for Risk of Ruin (RoR) is a function of variance, amongst other variables (such as bankroll size and your win rate). That means that given all else is equal, if you increase your variance, you increase your RoR. See Blackjack Attack for some of the RoR equations.

calleo said:
As I ramp up my betting am I better off opening more boxes instead of increasing the bets on one (or two) box(es)?
Click to expand...
Given the same wager, it will decrease your variance. But it might bring heat, which could affect your longevity and hence the # of hands you play in your lifetime at a given place. Also it might eat up cards. This is a topic still under research and debate! In my understanding, though, in shoe games the card-eating effect isn't as prominent as in pitch games.

calleo said:
Also if I'm playing 1-10 betting spread, and playing multiple hands once I reach max bet, does that mean if I'm betting multiple hands my max bet should be 10 divided by the number of hands on each hand? Or is it 10 per hand?
Click to expand...
That depends on how you define spread. Correct me if I'm wrong, qfit, but I believe in cvdata and cvcx spread is defined as your total $$$ on the table at the highest count divided by the total $$$ on the table at the lowest count (which would be a 20-1 spread if you spread from 1x1 to 2x10). This could just be semantics, and it depends on your simulation results, whether to be 2x5 or 2x10 in a given situation.
 
A

assume_R

Well-Known Member
Just as a followup, I know this isn't the theory or math section, but the simplest RoR equation (without time constraints, bankroll goals, etc. etc.) is:

RoR = (1 - WinPerHour / StandardDeviationPerHour) / (1 + WinPerHour / StandardDeviationPerHour) ^ (Bankroll / StandardDeviationPerHour)

With everything in terms of "units".

Some books or programs use the exponential function (e^...) to simplify the above formula, but that's it in its truest form.

And remember, like I said, you can add time constraints, bankroll goals, etc. which will inevitably make the above equation more complicated.

Going from 1x10 units to 2x5 units will keep WinPerHour the same but reduce StandardDeviationPerHour and hence reduce RoR. But you might play less rounds per hour, might have to shuffle more per hour (using more cards per round) and the high counts might not occur as frequently (since you could have used all the 10's in your 2 hands causing the count to drop!). The only way to see the effects are through full simulations of 2 hands.
 
C

calleo

Member
I should of mentioned it's a 6-deck shoe game.

And there is no heat at the place because I know many of the dealers and pit people. Also I bring a lot of ploppys and dropped a lot of cash before my counting days.

I can get shuffles in the middle of the shoe, raise the table limit, and even get a private table (heads up against the dealer) and can open as many boxes as I want without increasing the minimum bet.

Counters dream? Maybe, if only they would let me take more than one card on split aces/resplit aces.
 
moo321

moo321

Well-Known Member
calleo said:
I should of mentioned it's a 6-deck shoe game.

And there is no heat at the place because I know many of the dealers and pit people. Also I bring a lot of ploppys and dropped a lot of cash before my counting days.

I can get shuffles in the middle of the shoe, raise the table limit, and even get a private table (heads up against the dealer) and can open as many boxes as I want without increasing the minimum bet.

Counters dream? Maybe, if only they would let me take more than one card on split aces/resplit aces.
Click to expand...
That's good. I would make sure you continue to be welcome at this place, though. Get a good spread down (1-20, say) but don't push the limits too hard and get backed off.
 
psyduck

psyduck

Well-Known Member
Based on my own simulation, at high TCs, playing 2 hands of $10 reduces variance compared to playing one hand of $20. However, the two hand play also has lower winrate and lower SCORE than one hand with the same wager.
 
C

calleo

Member
I found the following passage in "Million Dollar Blackjack":

"There is one key provisio, however. If, by playing multiple hands, you reduce the number of rounds you would have been delt in a given positive deck or shoe, you haven't gained a thing -- in fact, you may have decreased expected value. For example, assume you're playing head-on at a single deck game with three-fourths of a deck left and a highly positive count. If you play one hand of $200, the dealer may deal you four additional rounds. If the count stays positive, you have bet four hands of $200, a total of $800 in action, before the shuffle. But if you spread two hands of $115 and the dealer gives you only two more rounds, your total action has been reduced to $460. You then would lose expected value.

On the other hand, assume you're playing at a table with three other players and the dealer deals three rounds, whether you play an occasional multiple hand or not. "Spreading" to two hands in positive situations will increase your total action and, thus, your expected value.

Multiple hands can also be used to good advantage toward the end of a shoe when the player knows the cut-card will come out on the next round. Not only is more money placed into action with no sacrifice in the number of rounds played (we will get only one more round at any rate), but the dealer is forced to deal further into the shoe, increasing the player's advantage. One time, in a highly positive four deck game at the Fremont, with a half-deck remaining in the shoe, I spread to seven hands of $500. The cut-card was the second card dealt. By the time the round was completed, there were only eight or nine unused cards in the shoe."
 
C

calleo

Member
Estimate penetration?

calleo said:
Multiple hands can also be used to good advantage toward the end of a shoe when the player knows the cut-card will come out on the next round. Not only is more money placed into action with no sacrifice in the number of rounds played (we will get only one more round at any rate), but the dealer is forced to deal further into the shoe, increasing the player's advantage. One time, in a highly positive four deck game at the Fremont, with a half-deck remaining in the shoe, I spread to seven hands of $500. The cut-card was the second card dealt. By the time the round was completed, there were only eight or nine unused cards in the shoe.
Click to expand...
This sounds cool. How can I do this? The shoe at my casino has a small lip where the cards are dealt from, not sure how many cards it covers, but has the effect of covering the cut card when it is about to be dealt.

I would estimate the average penetration to be about 4.5 decks.

What's the easiest way to more accurately gauge the penetration and *know* when the cut card is about to come out so I can spread to all the boxes?
 
You must log in or register to reply here.
Top