Less precise estimates win more money (long)

Alex

Member
This is the reason why: I already know for long time that there is practically no difference if you deliberately use the wrong index, off by one point or so. In live games I use half a deck estimates in 2D, not for any greater accuracy, but it is usually easy because of the few cards in the discard tray compared to the shoe game. I usually divide the RC by 2 until a clear half deck is in the discard tray, then by 1.5, and after that point I just use the RC when there is about a full deck dealt. Usually the penetration is not good enough to place excessive worry on getting more accurate than that. If you do have good penetration then it may be worth going to quarter decks into second half while playing a 2D game.

Using less precise deck estimation and being conservative on the numbers of decks dealt is like using RA (risk averse) index without having to learn them. All of the hit/stand, double down, and split indexes automatically will be pushed up a little. We did a study and come up with the conclusion that less precise estimates win more $$.

Another point that I would like to make is that you guys have got to pay attention to this concept: If you cannot get enough "edge" over the house to make a decent income then don't worry and go crazy with bigger spreads to accomplish that. Just use the POWER of your bankroll to do just that. Play with a smaller edge but with bigger denomination chips. Mathematically it has been proven that this professional technique is working great and is harder to be detected too. For example: If you need to make $100 per hour or $100,000 per year as a full time income then you have to play with a 1.0% edge and have an average bet of $100 per hand. Now, to do that, you figure in your sims that the game in question has got to be attacked with a 1:6 spread in order to generate that a 1.0% edge. Wow, a 1:6 spread is a huge spread, and if you have a problem applying that spread in real games then you start playing and thinking like a pro does. Reduce the spread until the sims tells you that you have 0.50% edge and then bet into that game with $200 average bets. If, for the 0.50% your sims tells you that you still have to spread to 1:3 and if you find that to be a problem then you got to run the sims again and find out what spread gives you 0.25% edge over the house. Let's say that 1:2 gives you a 0.25% edge. So, How much should your average be? --- $400 per hand! Play with that average and you will make the same amount of money in the long run like the player that is struggling with 1:6 spread while betting $100 average. You see, the pro way is to use the power of your bank unit to make the same income with a much smaller edge and obviously with a much smaller spread too. This gives you longevity in business, while your income is the same. Now, if you don't have the bank then you have to be happy with less income per hour applying the same principle. Don't get upset by the edge. Adjust the average bet to make up for!

Hope that my point about using the deck estimation for TC conversion and the concept of chip denomination related to the edge will open up some new ways for you guys to attack this game.

AlexD30

PS: Now, our great host "The Mayor" is a PhD in mathematics, he is a accomplished computer programmer and most important for all of us is that is a pro BJ player. He can give some light into this issue if he so desire. We all will benefit from it. I guess.
 

The Mayor

Well-Known Member
Great post -- thanks for your nice comments about me, but they won't lure me into contributing much to your excellent post. I just want to point out an essay I wrote a while back, which has a similar theme -- that we worry too much about precision:

(Dead link: http://www.cardcounter.com/Essays/Mistakes.htm)

I am not sure how you arrive at the conclusion that less accurate play results in a higher win rate EV). Fewer errors? But, I know many pros that do what you are advocating, e.g. on double deck they either use the original numbers or double those numbers at the 1-deck point. Similarly on single deck at the 1/2 deck point. Although the win rate is not higher, your point that the by using this method, the indices are pushed up towards their RA values, does reduce variance, which increases the DI of the game. That is, though EV is reduced a bit by using this scheme, SD is reduced even more, making the ratio EV/SD higher.

Now, on your thesis that we should raise our minimum bet to accomplish our income goals, with a fixed bankroll, your comments completely avoid the issue of increased variance and ROR, and are EV related only. With sufficient bankroll, such issues are not relevant, but the 100k bankroll you gave is insufficient for a $400 min bet on a weak game with a weak spread.

There is a program called BJRM (Blackjack Risk Manager) that very precisely tells you bet size for income goals, given specific game advantages and risk tolerance. Great little program.

--Mayor
 

Alex

Member
I agree with you totally about the bankroll. The bank has got to have the right number of average bets to get you even less then 1% ROR. Over 1,000 average bets will do a great job. For a bank of 100K I would say that a steady income of $60/hour is easy achievable even in a 0.60% edge if your average bet is $100. Now, if you play a 1:2.5 spread that probably will give you a 0.60% edge and make $60/hour as a full time player but that will be depending how many hours you can play per year. I suspect you can earn a little less then 100K/year as income. Now, of course, you can spread more and get a bigger edge but how about if you get barred from playing and run out of places? I think you are better off to build your bank and play with less spread and smaller edge but with bigger chips and make same money in the long run.

AlexD30

PS: Now, Mayor, How I post under the name of AlexD30? I created a profile and now if I use the AlexD30 it says that the password is not correct. What is my password? I don't know what password I have. Can you give me a password?
 
Top