Understanding heat tables.

[Skull]

Hi Norm!

I am trying to understand some portions of the Heat Table in CVBJ. The reason is that I would like to use the thresholds you defined to assign a "heat index" for every entry in my bet ramps (I know they vary by Casino, but I could use an average).

For instance, you assign 0.5 (page 119 of Modern BJ1) for the action of decreasing the bet after a win and you assign 1 to decreasing a bet to less than a third of the previous one. So, this means you add 1.5 degrees to the heat temperature if I win and then decrease to 1/4 my original bet? Or are these numbers multipliers somehow?

You assign 12 for a bet larger than 90 (I assume it's $90), so in my previous example, if I decreased from $400 to $100, would you add 0.5+1+12 = 13.5 to my heat temperature? This calculation seems a bit out of proportion, so I don't think that's how you do it...

Also, the heat doesn't take into account the TC variation, does it? For instance, decreasing a bet after a win in a rising TC would not be as bad for cover as doing it in a falling TC, would it? At least as far as surveillance is concerned, I guess...

I just want to quantitatively evaluate the bet ramps I create somehow, because it's easy to see that certain "betting paths" generate much more heat than others and I want to minimize the possibility of that occurring by tweaking the tables. So I know that from each cell, you have two entire rows of possible follow-ups. Each of these follow-up cells will have a bet size, which will represent a certain bet variation from the previous bet, which will be done following a given TC variation. Also, each TC variation has a given probability, so if I could weight every action by its probability of occurrence and "heat index", I would be able to see where the "hot spots" of my tables are and then try to get rid of them. Makes sense?

Thanks in advance for any help!

Best!

[Norm]

Every casino is different. Every person is different. Don't spend too much time trying to formulize instinct.

[Skull]

I will spend a fraction of what you spent developing the heat functionality Norm, don't worry.

I find a lot of pleasure thinking about this kind of stuff. I like finding patterns in things. I particularly love receiving a number matrix as input and outputting colored patches of homogeneous classes that you can classify based on an n-dimensional space of attributes. I think it's a thing of beauty when meaningful hidden patterns inherent to the data are uncovered by careful analysis.

It's a hobby for Christ sake!!

A nerd's hobby, but nevertheless, a hobby!

If we try to fit our model to everyone, we will for sure have an over-fitted useless model, but we don't do that, do we? We minimize overall discrepancies, that's all. So, it doesn't matter if "Every casino is different. Every person is different.". We search for something that works "on average", don't we?

Maybe you are tired of thinking about BJ Norm, I understand that. Like you said in a past post, "you have been there and done that".

But I haven't!

So c'mon Norm, stop being so grumpy and please answer my OP .

Is my previous calculation of 0.5+1+12 = 13.5 correct as far as heat is concerned?

Don't tell me the math is right!

If the answer is NO, I will then find a nicer way to ask "Why not?" in my next post, OK?

[Skull]

http://www.blackjacktheforum.com/ent...ing-strategies

Difficulty coefficients?!?!?

What did you say about "formulizing instinct"?

[Norm]

If you bet more than $90, then the heat increases by the value in the Bet>90 row. If you bet less than one-third the previous bet, the heat increases by the value in the < third bet row. You can change the values to whatever you wish.

[Skull]

OK Norm, thanks a lot!