kelly betting in a heat-less casino.

what would be the ideal bet spread for a 5 deck game for high card counts? should i bet more for a +8 count as for a +5, or past +5, should i simply have one maximum bet? am i better off with two hands of 5 units, or one hand of 8 units or so?

i'm trying to take advantage of my casino situation--lousy rules, but no heat. i was wondering how many units i should bet on those rare moments when the count is really high (over 7 or 8). let's assume i'm willing to take a 5% risk of losing my bankroll.

i'm a new counter and i just see myself betting haphazardly high when the count is above 5. i don't want to overbet on the one hand, but i don't want to miss out on such great opportunities on the other hand. in a casino with no heat, what should my bet spread look like. assume my bankroll is $10,000 and the minimum bet is 10, with a maximum bet of $500.

thanks, if anybody wants to really help me out. i'm sure the answer is in one of my BJ books...but i don't think any of them consider completely heat-less conditions at extremely high counts.

the artist formerly known as 'toecheese'

Max bet should be made at +6. And max bet is max bet.

BJRM software will give you the exact answers with the rules you face. It is a good investment if you are serious about the game. Someone could probably run the BJRM for you here, but as your bankroll grows or under various conditions you will probably want to be able to run the numbers yourself.

From Tcheese

----in a casino with no heat, what should my bet spread look like. assume my bankroll is $10,000 and the minimum bet is 10, with a maximum bet of $500.

According to Stanford Wong, your maximum bet should be 1/150th. of your available bankroll, which would make your's $66.66. or $70., on one hand...You said that your highest bet was about 6 times your minimum...This is not enough to eliminate the approx. 0.40% advantage in a 5 deck shoe, and construct a 1% advantage for yourself.

Cheers

thanks for the reply. i have Sage Blackjack right now. i'll run a couple simulations comparing bet spreads. what is BJRM (what does it stand for?) and what are the main advantages of this program (what does it have that my Sage program lacks)?

a max-bet is a max-bet? darn. i was hoping i could get away with bigger bets at +8 TCs. i guess i just increase my risk doing that.

thanks again.
cheez whiz.

i hope my advantage is higher...i wong (when i back-bet) and when i sit at the table i make frequent trips to the toilet on low counts. the casino personnel must think i have some kind of digestive problem!

i wonder how much i save by not playing any -2 TC hands. one night i lost so much money on my minumum bets, i decided to be more liberal in going to the bathroom. now, i credit the toilet for my daily wins. (i haven't lost in about 20 hours of play...my longest streak ever).

i'm new to counting...and there are days where i feel like i'm taking candy from a baby. and other days i feel like david against Goliath, or Rocky against Apollo Creed. but when i win again and again, i feel like Rocky against that big Russian guy in Rocky 4. i mean, it's kind of boring after a while--we all know how the damn movie is going to end, but we sit through it anyway.

when i lose big, i feel like Rocky in Rocky 3, where Mr T kicks the living s--t out of him AND threatens to rape his wife AND gives his trainer a fatal heart attack. it's not easy being Rocky. but Rocky always wins. that's the one thing you can count on. in the end...whenever that 'long run' will be, Rocky will be there, on the steps of that place in Philadelphia, with his hands in the air, pontificating--"Adrian!"

the cheeze

This is a nifty little program that calculates optimum bet spread, risk of ruin, disirablilty index, SCORE, hourly win rates, trip win statistics, and lots more. MOre info can be found at:

http://www.bjrm2000.com/

Also of interest are a couple of new programs from QFIT software, CVCX and CVData, which perform similar functions and more. CVCX also includes a high speed simulator. I have not had a chance to evaluate these programs yet, but since they are from the developer of Casino Verite, IMHO the best practice program on the market, I am rather excited about them. More info at:

http://www.qfit.com/

The modern theory of optimal constrained betting is based upon the premise that you are required to stay within a certain spread for cover reasons. For example, you use a 1-10 spread which means your smallest bet must be at least 10% of your max bet.

As I understood this question, there is no such restriction for play at this casino. In which case, Max Bet is nearly meaningless. For example, at Hi Lo TCs of

+9 you would bet over 4% of KEB
+6 you would only bet 2.4% of KEB

This is for playing 1 spot at a Midwest shoe game: S17,DAS, 75% dealt.

I computed these numbers with BJRM.

yes, you understood my question right. i jack up my bets when the count get's above +6. more than my normal "max bet" at around +5. psychologically, i can handle it when i lose my max bet. but psychologically, i feel the butterflies when i lay out twice my max-bet at high counts. of course, i've been winning doing that (perhaps over-betting though) and it's really paid off so far. i want the mathematically justified answer for how much i should bet--that would put me more at ease when such infrequent bets do arise.

Mathprof...what's KEB? kelly betting or something? i'm a new counter and i couldn't find KEB in the site glossary.

if you have the time...could you make a guess of what my bet spread should look like: 5 decks, 75% penetration, no DAS, s17, no resplitting Aces.

assume i have a bankroll of $10,000 and i'm willing to take a moderate amount of risk (i can afford to lose the bankroll, but i'd obviously rather not!). minimum bet at the casino is $10, maximum bet...sometimes $250, sometimes $500, sometimes $1000 (i back-count and wong the higher limit tables when possible). also, i can spread to two hands sometimes.

for example...last time i played, the cut card was the first card dealt to box 1--the count was +8. i threw out double my maximum bet on two hands. got a 20, and then two aces against a dealer 6. i split the aces (god, that made me nervous since i only get one card for each ace). i got a 21 and a 17. dealer busted and i breathed a sigh of relief. but then...the question always hits me...should i have bet more?

i would really appreciate knowing what i should do for counts up to +10. the rules i play with suck, but there's no heat so i want to take advantage of the situation. thanks prof! (PS i'm in grad school...an aspiring prof...but not in math).

By KEB, I meant Kelly equivalent bank. Suppose that you have a 20,000 bank but play half-Kelly. Then your KEB is 10,000.

You didn't specify if you bet Full Kelly, and I would dsicourage you if you did.

If you want to use a fixed level of betting with Risk of Ruin R, and your bank size is B, then your KEB would be

- 2 B / Ln (R)

For example, if you had a 10,000 bank and wanted a 5% RoR, you KEB would be - 2*10K/lln(.05) = $6,670.

How often does a +9 count occur? By saying you are going to bet more than a 'max bet' at +9, how much more money will you have after 2000 hours of play?

How often does a +9 count occur? By saying you are going to bet more than a 'max bet' at +9, how much more money will you have after 2000 hours of play?

More than if you did not. It is never to your financial advantage to deviate from Kelly. Why are you disputing this?

Here is an example: In a game I play frequently, at a +4 count my KEB bet is 4.97 units. Since the casino doesn't make .97 unit chips, I bet 5 units. Mathematically, I am overbetting. If the casino did make .01 unit chips, would I make a 4.97 unit bet instead of 5 units? NO. Because then I would have to factor in the time lost with the dealer making change and paying off odd amount bets. That loss in hands per hour outweighs the .03 unit overbet.

Another example: Kelly betting is based on betting a percentage of your total bankroll. With each win or loss of a bet, the optimal bet changes. Should I whip out my calculator and figure out my next optimal bet? No, even with no heat, the time spent calculating and placing optimal bets, would cost me hands per hour.

The pratical considerations in the above two examples that prevent me from making ideal mathematically correct optimal bets occur far more frequently than a +9 count. Why do we accept the above limitations, but draw the line at a max bet at +6 concept?

Again I ask, how often does a +9 (or +8) count occur? If it probably won't happen in 2000 hours of playing time, have I lost anything in that year's worth of playing time by planning on not increasing my bet on something that didn't occur?

How about this one? How are we determining our advantage? Based on a count? The count is just an approximation of our true advantage. It is not perfectly linear. There is sometimes a greater increase in advantage in a +3 to +4 change in count than +2 to +3. I would assume these same 'bumps' in advantage exist in +7,+8,+9,+10. Will knowing the increased rounded amounts I should be optimally spreading at these rare high counts win me enough to buy me an extra 6 pack at the end of the decade?

I don't know the answer to that question. I could be wrong. Maybe there is a sharp spike in advantage at, say +8, where you should really be increasing your bet. Of course, then we would have to get into the fact that I am estimating remaining decks by 1/2 or 1/4 deck and not the actual number of cards played. Maybe my +8 count isn't really a +8 but a +7.995 LOL.

Thank you BJRM for allowing me to sim away, and figure out how to best smooth the real world limitations, and let me take my non-optimal bankroll, and place close but not optimal bets, using my rough non-optimal system, in a game filled with non-optimal distractions and non-optimal variations.

I don't understand what you're staying there. Your max bet is the biggest bet you make. You can't bet bigger than your max.

adaptation that we use for cover purposes. If you define a constrained spread, such as 1:10, etc, then it can be defined. But if you want to bet bigger than 10 units then it is better to just figure what the optimal spread is for your real max. If you have no need for cover then you don't need to define spread constraints, such as max bet. You just bet what ever is warranted given your advantage. It makes no sense to figure an optimal spread for 1:8 and then pick some random high count to push out a bigger bet.

So, if I calculate my advantage and determine a 10 unit bet is optimal for my bankroll at +6, what do I lose by betting 10 instead of 11 units at +7?

One game I am looking at it would be (.68% freq of +7) X (.35% increase in advantage) X (1 unit not bet) = $0.0595 per 100 hands at a $25 unit.

Geez, at 40 hours a week, 52 weeks a year, that is $100 gain not realized. How many six packs will $100 buy these days?

Thank you guys, for forcing me to do the math. A nickle per hundred hands is cool. I knew an old guy who made a nickle commission on every 100 lbs. sack of potatoes he sold, he lived on a nice large ranch and spent his retirement days fishing.

my original question was hoping i'd get more bang for my buck by betting much more at high counts than what seems to be prescribed in my BJ books (since most BJ authors wisely assume you need to watch out for heat). i was also hoping there'd be a significant jump or spike in TCs above, say, +6. if the money gained by betting progressively (ad infinitum, without regard for any pre-set bet spread) is negligible in the 'long run' (whatever that is), then that's obviously important to know so that i don't whip out my calculator at the end of the shoe to figure out how much to bet on the last dealt hand.

on the other hand, it seems like so much of the BJ literature is devoted to crunching out numbers for the most subtle advantages. i feel like those advantages would add up at the end of the year. and even if they didn't, i'd feel better knowing that those six-packs of beer were in my fridge, and not that of some casino exec.

but thanks to all for answering my question at both the theoretical and practical level. and besides...once i become a high roller (and when will that be?), i'm sure betting big at +6 or +9 will be enough for my yearly supply of french wines and cheetoes.

the cheeze

because you have made a decision that this is what you think will fly at that casino. It has nothing to do with being mathematically optimal for your bankroll. Once you make the determination what type of spread you think you can get away with, the spread can be optimized for these constraints.

But I've done it too, based on the situation in the casino. I may have an optimal 1:10 spread figured with a max @ +5. I get a much higher count at a time when no floor people are around and the dealer doesn't seem to be calling "checks play". So I might bump it up a little bigger. By the same token, I figure the spread based on play all but I will escape several of the worst counts from table departure and restroom breaks. You should probably have your optimal ramp based on what you will be doing the vast majority of the time that you are at the table. Don't think of it as losing anything by not making a bigger bet than your max at very high counts. Think of it as a little bonus on those rare opportunities when you think it is safe to push it a little more.