Optimal profit VS Minimal Risk

jee_pack · May 28, 2007

I just made a bunch of calculations using the kelly formula to see what would be the optimal bet if I spreaded to 2 or 3 hands at each possible TC (in my spread of (1-8). And I know that by spreading to more hands using these bets, I'm improving the EV but the Risk stays the same. Then I know I could also just divide the one hand bet by 3, and put that bet on 3 hands, and the EV will stay the same but the Risk will go down. So if I bet anywhere between this bet (one hand bet divided by 3 on 3 seperate hands) and the optimal bet (calculated for 3 hands). It means it theory that I am both improving the EV and reducing the risk. The thing is I'm clueless at where I should start to value the EV more than the Risk and where should I do the oposite. I was wondering if you guys have any input on this subject. These are the bets I calculated (with a 25$ unit bet):

TC = 2: 25$ / 20$(40$)-NA / 15$(45$)-NA
TC = 3: 50$ / 35$(70$)-25$ / 30$(90$)-15$
TC = 4: 75$ / 55$(110$)-40$ / 45$(135$)-25$
TC = 5: 100$ / 75$(150$)-50$ / 55$(165$)-35$
TC = 6: 125$ / 90$(180$)-65$ / 70$(210$)-40$
TC = 7: 150$ / 110$(220$)-75$ / 85$(255$)-50$
TC = 8: 175$ / 125$(250$)-90$ / 110$(330$)-60$
TC = 9: 200$ / 145$(290$)-100$ / 115$(345$)-65$
TC = 10: 225$ / 165$(330$)-115$ / 130$(390$)-75$
TC = 11: 250$ / 180$(360$)-125$ / 145$(435$)-85$

If you look at line 1, it means that at TC +2, the one hand bet is 25$, 20$ for 2 hands (40$ total on table), the NA means its not possible to divide the 25$ in 2 because of the table minimum (15$),then the 15$ is for the bet with 3 hands(45$ table total), and so on.....

So if you look at TC 5, for 3 hands, the 55$ is the OPTIMAL bet on each hand, to improve EV with same RISK. And the 35$ is the bet to place on each hand to MINIMIZE the RISK while not changing the EV. How should I bet knowing that with 3 hands and a TC of 5, if I bet anywhere between 35$ and 55$, the EV can only be better than playing a 100$ bet on 1 hand, and also the RISK is lower. Of course if I bet 35$, the EV is not improved, and if I bet 55$ the risk is not lowered. But I could improve both by betting, 40$, 45$ or 50$...

That's too much possibilities for me, so I just wanted to know what you more experience counters think of this dilema... Any comments would help, thanks in advance

Sonny · May 29, 2007

The first thing you need to do is decide on an acceptable risk of ruin. Are you comfortable with a 10% change of losing your entire bankroll? How about 5%? 3%? Once you decide how much risk you feel comfortable with you can size your bets to fall within that level. That is the first step.

Another thing to think about is heat. If you start spreading to three hands every time the count is positive you might start to get some attention sooner than you’d like. You aren’t increasing your EV if you can’t play for as long.

Also, be careful not to eat up too many cards during those positive counts. If you spread to 2 or 3 hands you might only get 2 rounds of play, but playing one big hand might get 4-5 more rounds. In that case you would be better off playing one big hand.

-Sonny-

jee_pack · May 30, 2007

thanks a lot sunny!

One thing though, comparing the EV of playing one hand and more rounds in good counts, to playing 3 hands with more money but less rounds in good counts. I'm lost here, I mean, both of these have 1 big PRO and 1 big CON, so how the heck do we find out which method is the best one.

Lets say the count goes to +4, and you win and even get a blackjack, that probably means the count went down since a nice amount of tens and aces came out. So the count drops back to +2 after the round. Well if you play 3 hands, that's 2 more hands at TC +4 than if you would have played only one hand! See what I mean....

And again, if you EV goes up by placing 3 hands on the table, maybe you play less rounds, but that's not for sure as when a count is at lets say +2, especially far into the shoe, you can't know if the count will still be +2 after the round...

This is confusing me a lot, would you mind explaining a bit more? thanks

mdlbj · May 30, 2007

What sonny is saying is mostly correct. When the count is there spread to more hands in play. As far as risk of rouin, you will base this on your the kelly you are playing. If your bankroll is smaller, your kelly is closer to 1 which means spreading to more than 1 hand will increase your ROR.

In principle, he is dead on.

You will mostly be playing two hands while at the table. One will only be playing one hand if they are running solo with a small bankroll. When playing two hands or more will dismiss heat to a point.

zengrifter · May 30, 2007

mdlbj said:

No, Jee has figured out how to bet more in total across 2hands for the same RoR.

However, what everyone is overlooking is the potential excess hand depletion - there is an optimum level risk/EV for 1-2-3 hand high-count bets that is detirmined by taking into account how many others are playing at the table. zg

jee_pack · May 30, 2007

zengrifter said:

Not anymore! I'm currently completing the spreadsheat for this "optimum level risk/EV for 1-2-3 hand high-count bets that is detirmined by taking into account how many others are playing at the table".

I need to be sure 100% percent of the proportion of cards in a dealer hand compared to the # of cards in a player hand (average). Untill I know that # for sure, I can't make the calculations exact.

Sonny · May 31, 2007

jee_pack said:

I believe it should be about 2.7 cards for both the dealer and the player. That's what Schlesinger used in his analysis in Blackjack Attack.

-Sonny-

Brock Windsor · May 31, 2007

Sonny said:

I also add 0.1 to the number of hands at the table to account for splits. i.e. if there are 7 players at a table you will average 2.7 cards per hand and 8.1 hands per table.

zengrifter · May 31, 2007

jee_pack said:

OPTIMUM BET MAX HANDS -

0-1 other spot = 1 hand
2-3 other spot = 2 hand
4... other spot = 3 hand

jee_pack · May 31, 2007

I'm not sure that's exacly what my results point to (I don't have the spreadsheat here as I only have excel at work), but I can attest that is mostly looks the same. I think though that with 1 other player, the 2 hand spread is already optimal. I'll know friday. But the cool thing about the spread sheat is the portion that Calculates MINIMAL RISK with SAME EV. Cuse the minimum bet you can place for each TC is in direct relation to how much hands are being played. Every single result you get by not taking into account the depletion rate has to be changed to represent the real minimal bet; which is not the case for optimizing, the spreadsheat in this case will only use depletion to see if the EV is increased or decreased, and by what percentage, but it will not modify the optimal bet.

jee_pack · May 31, 2007

Brock Windsor said:

So what is the good number? is it 2.7 for the dealer and 2.8 for each player?

Sunny: the 2.7 vs 2.7 from blackjack attact, does that take into account the splits? I have another thread about this... Someone in the topic I started specificly to find this # (Subject: "For those of you with math skills"), said that it looks like the dealer hands actually contain more cards in average, for many reasons other than splits, like us only drawing 1 card on a double. Or us standing on 12 through 16s... Stuff like that... This is why the subject was: For those of you with math skills... Everyone here has math skills! But so far, no one has said the same #, the first one in the other post is 2.7 for player vs 3 for dealer, sunny's is 2.7 for 2.7, and brock windsor's is 2.7 for the dealer vs 2.8 for the player....

Anyway I apreciate all the inputs on this by everyone, I'm sure there is no where else on the web where I have better odds of getting this help. Jee

Optimal profit VS Minimal Risk

jee_pack

Well-Known Member

Sonny

Well-Known Member

jee_pack

Well-Known Member

mdlbj

Well-Known Member

zengrifter

Banned

jee_pack

Well-Known Member

Sonny

Well-Known Member

Brock Windsor

Well-Known Member

zengrifter

Banned

jee_pack

Well-Known Member

jee_pack

Well-Known Member