bankroll using kelly criterion

Hi,

I'm working with a very limited bankroll here, because I don't make much money and all that, so I have a question about said bankroll. Due to being too prone to losing it all in a play-all situation, my best bet is wonging in and out. I have a decent way to get away with this, in that I visit the casino with a friend who can afford to play all. I would just sit and shoot the bull with my friend while he's playing and whenever I nudge him or whatever on a good positive count, he would add some of my money to his bet. By doing this, my money is only in play when I have a positive EV.
So if I want to bet $5 (no one will notice a jump in his bet of only $5) every time that the count reaches a hi-lo true count of +2, how much does my bankroll need to be?

My thinking is that at +2 TC, the player advantage is 1%. Is this accurate? I lent my books to a friend, so now I don't have them here to refer back to. If I'm right, and the Kelly Criterion says bet an amount of your bankroll equal to your advantage, then I would want that $5 to be 1% of my bankroll, right? So I think I would need $500 to play in this way. Also, since I'm not risking anything in negative counts, etc, what should I expect my win rate to be?

Thanks for help in advance. Correct me on anything I'm wrong about.

There is a lot of info on the board about optimal betting and risk of ruin, and it can get complicated. If you don't have a math or statistics background, think about picking up Sanford Wong's "Professional Blackjack," which has an excellent introductory chapter on this subject. I know just enough to realize what a good teacher he is, and I would recommend it.

I predict you are going to hear a lot about variance, which can get you even when the deck is in your favor. Lets see....with an advantage of 1%, minus a half percent house advantage, figure on an EV of 25 cents for every hand you play, times say 25 hands an hour. Plus a little more at higher counts. The problem is the range of possible values in real life. It is amazing how often the house wins even when they aren't "supposed to." ;)

I've been more or less in the same position as you, and I've given the subject some thought; I have one suggestion that might be of interest. It sounds like you guys are planning on a "spread" of 1-2; in other words, the minimum bet will be one chip and the maximum will be two. If so, then YOU might make money, but between the two of you, there will be a net loss. In other words, your kind friend is going to lose more than you make. You need a bigger spread to beat most shoe games.

You might want to think about switching seats. So you can sit there and bet minimums, and when the count gets good, you call in an airstrike from your rich friend, who will come in at 5 or 10 times your minimum bet and parlay up from there. Between the two of you, you will need about three grand to play this game. The advantage of this approach is the power of a pooled bankroll, and also neither one of you has to worry as much about covering bet variation or playing strategy deviation. The disadvantage is that it wouldn't be hard for casino personnel to figure out what you're doing, so you have to be cool and not get greedy.

Now, if your friend doesn't MIND if he loses, then you are off to the races. I have a friend like that...what a guy! I can usually get away with jumping in on one or two good shoes without getting nasty looks, although I don't know that I would feel comfortable grinding away for several hours using this technique. Its just too obvious. Look around the casino. You will see people standing behind the table, cheering their buddies on. Make sure you act like them.

Either way, dive in and have fun. I started with not much more than you, and I made a little money. I got lucky. Luck counts. Just realize that there is no guarantee of winning if you don't have a sufficient bankroll, and don't be running for the ATM if you get down.

Assuming a 6D, S17, DAS, NSR game with 4.5/6 pen, if you bet $5 at all TCs of +2 or greater and have a $500 BR, then your ROR is 8.8%, which isn't too bad. The bad news is that your EV is $1.25 per 100 hands (played or observed). If you are not able to get your friend to switch tables on negative counts, then you will probably only get about 60 hands per hour at a shoe game, so your hourly EV is 75 cents. At least you are playing with an advantage. Your likely result (68% probability) over those 60 hands is anywhere from a win of $14.55 to a loss of $13.05. Of course, you could get into 2 SD territory (-$26.85 to +$28.35). In short, you are unlikely to win or lose much money.

td's suggested approach is much more profitable.

My buddy is being rated, and comped, etc, doesn't care if he wins or loses, the regular gambler. I just happened to ask if I could add on to his bet every once in a while and he said sure, I don't care. Now my plan here is starting to work itself out... I have a bunch of friends, actually, who like to gamble together. At times they'll all be sitting at the same table, betting their green chips while I'm in over my head betting $10 a hand. So if I can get one guy to do it, why not 5? I give each of them $100 in red chips of my own, and sit there BS'ing with them. Then when I like the count, I'll say "Hey I feel lucky, everybody bet $5 for me!" and voila -- I have $20 - $35 on the line in every +2 count.

I know it might sound far fetched, but with the people I hang out with, I think it is doable. The question is... is it worth it?

How are you planning on keeping track of and divvying up your results? I could see this approach leading to possible confusion about what's who's.

In this situation your best option is to just use your bigger betting drunken(?) friends as cover and spread as big as you can, bankroll permitting. I used a similar ploy frequently last summer in NE casinos and received absolutely no scrutiny while spreading 1-10 with very aggressive table hopping.

Additionally comps & mail offers were suprisingly generous considering my level of play, it seems as though the rating was based more on my big bets, perhaps this was partially due to being amidst my ploppy friends and their larger wagers.

ANS

> My thinking is that at +2 TC, the player advantage is 1%. Is this accurate?

A rule of thumb is that each unit increase in TC corresponds to a 0.5% increase in advantage, but remember that you must overcome the house edge (typically -0.5%). So for a typical game, at +2 TC your advantage will be about 0.5%. Depending on the rules and the system employed this could vary maybe 1.0% either way.

> If I'm right, and the Kelly Criterion says bet an amount of your bankroll equal to your advantage, then I would want that $5 to be 1% of my bankroll, right?

This is true for a game with even payoffs. With uneven payoffs you need to multiply by the variance (which for blackjack is about 1.25) to arrive at the required bankroll.

It is very difficult to make money counting cards with only $500. Would you be able to convince one of your richer, less-skilled friends to stake you for a share of the profits?

Can you get a grand up? If so you can achieve a Win/100 of >$10 with a RoR in the neighborhood of full Kelly in a 4.5/6 game, backcounting with a spread something like this:

TC___Bet
+2 = $5
+3 = $10
+4 = $10
+5 = $15
+6 = $15
+7>= $20

According to my calculations and limited experiences it is *possible* to achieve 100 hands per hour while backcounting in good conditions, though it by no means is an easy feat. I'm sure some may disagree.

ANS

Thanks a ton for your suggestions. I think I can get the grand together, and I'll play this strategy. I don't know how likely it is I'll get 100 hands per hour, but just trying this will be a lot of fun for me. Thanks again, I'll let you know my results after the first time I go. Might be a little, since I've got to put the $1000 together (I didn't even have the $500 set aside yet, actually.)