BlackjackInstitute.com

AndyBloch · Joined: 08 Mar 2007 Posts: 4

I guess I should have said that the betting strategy that MJ described was not the MIT team strategy. (I don't know what's in the home study course since I'm not part of BlackjackInstitute.com.) First, we would almost never sit through negative shoes, and if we did, we'd bet as little as we could get away with. When I teach blackjack, I recommend having a minimum when playing shoes of no more than half a unit (preferably 1/4 or less), and leaving at a true of -1. And, we would reach our maximum bet much earlier than a true of 17. If we thought we could get away with a 32-1 spread, we'd have lowered our minimum, not stretching our betting increases out to a phantom maximum at a count that never happens. Also, with other players at the table we'd be playing two hands whenever the true count was above +1.
_________________
Visit my home page at andybloch.com
Play poker with me at FullTiltPoker.com

Marco · Joined: 27 Nov 2006 Posts: 6

Alright at this point what way of betting should be used. It sounds as if the optimal way makes a good point, why then would we want to use the TC-1 if the simulation and simulation expert says otherwise.

AP · Joined: 27 Jul 2005 Posts: 56

Andy, what you described is exactly what 's in the Blackjack Institute's home training course. The training manual does not say that your max bet occurs at TC of 17. It states that you should have a max bet of at least 6 betting units, which given a minimum bet of 1/2 unit is a bet spread of 12/1. MJ arbitrarily picked a max bet of 16 units. A more accurate comparison would be to state a max bet occurs at a true count of +7. In line with what you said Andy, the manual states that your minimum bet should be at MOST 1/2 your betting unit and that your max bet should bet at LEAST 6 betting units, in order to get a bet spread of at least 12/1. Of course if you can get away with a bigger spread, the stronger your game. The manual also states that you should look to drop out or leave at a TC of -1.

MJ · Joined: 03 Jul 2004 Posts: 35

Thanks for the feedback Andy! Smile

For those of you who don't know, Andy managed the other BJ team at MIT....the Amphibians.

As far as I can tell, the comparison was apples to apples. Let me explain why. Whenever we compare bet schedules the ROR, BR, and bet spread should all be the same. It is not possible to have the same minimum and maximum bet, along with the 3 variables I mentioned above to all be equal when comparing 2 different bet schedules. If I increase the minimum bet of the optimal bet schedule from $20 to $50 like you suggest, then that will increase the ROR way above full kelly.

Please note that if I alter the MIT minimum bet from 1/2 unit to 1/4 unit like you suggest, then the ROR is reduced far below full kelly. That is all well and good but the assumption here is that we are comparing the optimal schedule (at full kelly) to the MIT schedule (based on full kelly but higher then 13.5%). Remember, BR/266 = full kelly. Thus, the full kelly unit is $100. I then followed the linear bet ramp that MIT uses.

MJ
_________________
"Thou shalt NOT cheat. Thou shalt, however, take FULL advantage of dealer and casino errors. Thou shalt NOT tip dealers" -JP

Tbonez · Joined: 12 Jul 2006 Posts: 4

I'm not quite sure Semyon Dukach would agree about who managed the amphibians.

Anyway I was wondering after reading the posts from Andy Bloch and AP, is it fair to compare these systems manipulating the MIT style to fit the other one. As they said even though they would use a TC-1 like you simulated, they would not wait for a TC of 17 to put a max bet out. The idea expressed in the HTC is to put out 6 units as a top bet. Usually giving a 12-1 spread with at most .5 unit minimum. Your simulation is not a correct representation of what the MIT guys teach. I don't know the answer to what running a simulation of that would bring but it would be at least close to what is in the manual. Otherwise I think you are comparing apples to oranges. If they don't really play that way why run a simulation to find out what would happen if they did. I'm no expert, but I think the fellas that run this website are and I would tend to listen to there advice. I'm sure there is good reasons for betting the way they do.

MJ · Joined: 03 Jul 2004 Posts: 35

Tbonez and AP,

The HTC does NOT state to put out a max bet of 6 units or that the max bet should be placed at TC = +7. The whole point is to wager a fraction of your BR which is in direct proportion to your advantage. Page 60 states as follows:

"Ideally, your minimum bet should be, at most, 1/2 your betting unit. Shoot for a spread of at least 6/1 between your max bet and your betting unit".

The minimum bet is $50, or 1/2 of $100 which is the betting unit. Finally, rather then spread from 6/1 units I spread from 16/1 units. Nowhere in the manual does it state you cannot exceed a 6 unit spread. For example, at a TC of +9 the MIT Team would bet 8 units or $800.

MJ
_________________
"Thou shalt NOT cheat. Thou shalt, however, take FULL advantage of dealer and casino errors. Thou shalt NOT tip dealers" -JP

AndyBloch · Joined: 08 Mar 2007 Posts: 4

AndyBloch · Joined: 08 Mar 2007 Posts: 4

AP · Joined: 27 Jul 2005 Posts: 56

MJ, have the re-run the sims with the changes Andy mentioned? BTW, with regard to the max bet, what I was referring to was the bet frequency of the max bet. You stated that a max bet of 16 units rarely occurs. Of course this is the case. No mater what betting ramp you use how often is the TC going to reach +17? A fair comparison would be to compare the amounts bet at corresponding trues. For example, under "optimal" betting you essentially max out at a true count of +6 at a bet of 580. This should be compared to a bet of 500 at the same TC of 6 betting TC-1.

MJ · Joined: 03 Jul 2004 Posts: 35

Ok Andy,

I tried to implement your suggestions. The BR size and bet spread are still the same, and we only play hands in which the TC >=0. For all instances when the TC drops below 0, we sit at the table and simply observe.

According to the optimal bet schedule as calculated by the software, the max bet is now placed at a TC of +8 rather then +6 (unlike the other simulation). As per your request, I placed the max bet at the same TC as the optimal bet schedule.

The performance is virtually identical across ALL statistics including CE. Huge difference from the last simulation. But I feel that all we really did is more or less make the MIT bet schedule conform to the optimal bet schedule.

MJ
_________________
"Thou shalt NOT cheat. Thou shalt, however, take FULL advantage of dealer and casino errors. Thou shalt NOT tip dealers" -JP

semyon · Joined: 20 Jul 2004 Posts: 1

>Andy managed the other BJ team at MIT....the Amphibians.

Yeah, it was definitely a team effort, though I ended up doing more managing than most in the first couple of years. Besides Andy, Roz C. did a lot of accounting and Steve H. a lot of kvetching and Dan N. a lot of simulator coding and Peter H. a lot of dealing, among many others. I think there were at least 10 people doing significant management.

Semyon Dukach

zetadog123 · Joined: 11 Feb 2007 Posts: 7

QFIT, read your post bout getting kicked out of 3 casinos on a trip to vegas,then you spent a afternoon at reverse house learning camouflage techs.what is reverse house and please share some of the techs you learned with us .thanks for the help!!!

RJT · Joined: 01 Jan 2006 Posts: 157

Revere's house, was the house of the very famous author and player Laurence Revere. Wrote 'Playing Blackjack As a Buisness'.

RJT.

mdlbj · Joined: 27 Sep 2006 Posts: 143

Dave has the software package come to fruition? I'm looking forward to seeing what it is like.
_________________
Regards,

mdlbj