How much is the Illustrious 18 really worth?

[hopson77]

As I've said before, I'm a relative noob to AP, but I'm getting there. I've got BS down, and I'm using simple Hi-Lo. I know BS leaves roughly a 0.5% advantage for the house.

The 3 big plays in the I18 are roughly 60% of the advantage you get from using I18, and knowing when to take insurance is half of that 60%. It makes sense to me that making the right I18 plays in higher counts is more advantageous than in lower counts because there is more money on the table, but at the same time you will run into lower good TC's than higher TC's.

All of this having been said, what's the percentage gain from using the I18?

Many thanks!

 

[Sonny]

The exact value of the Ill18 will depend on the game and your bet spread. Using indices will increase your EV while reducing your SD, which is really the goal of the game. Here is a link to a recent discussion:

http://www.blackjackinfo.com/bb/showthread.php?t=3087

-Sonny-

 

[hopson77]

Thanks, Sonny. I appreciate the link.

I understand the concept of standard deviation, but it's been a loonnngg time since I've had to calculate it. EV = something like $ put in play times your advantage (or disadvantage), correct? How might I calculate standard deviation for how I play based on all the variables that get thrown in?

I really do appreciate all the help, guys. Hopefully soon I'll be able to be a true AP.

 

[Sonny]

EV = something like $ put in play times your advantage (or disadvantage), correct?

Right on. EV = Advantage * Money

How might I calculate standard deviation for how I play based on all the variables that get thrown in?

SD is a bit trickier. As a general guidelilne you can multiply the square root of the number of hands by 1.33. For example:

SD for 1 hour = 1.33 * sqrt(100 hands per hour) = 13.3 units per hour
SD for 5 hours = 1.33 * sqrt(500) = 29.7 units

The SD will depend greatly on your playing style so the formula above is just a general estimation. A more accurate answer would require a computer simulation. For most games your hourly SD will be around 26 units.

-Sonny-

 

[sagefr0g]

The SD will depend greatly on your playing style so the formula above is just a general estimation.
-Sonny-

by depending greatly on your playing style do you mean action such as backcounting, wonging out, bet ramp? any other factors?

best regards,
mr fr0g

 

[hopson77]

And if I understand SD correctly (which is a long shot, it's been ages..), something like 65% of the time I will win/lose 13 some units in an hour of play. 95% of the time, I will win/lose 26 units in an hour, and 99% of the time, the most I could be expected to win/lose in an hour of play would be 39 units.

Is that correct? And do your SD figures account for bet spreading? I would assume so? Thanks so much.

I almost feel like I'm joining a secret society or something.

 

[Sonny]

And if I understand SD correctly (which is a long shot, it's been ages..), something like 65% of the time I will win/lose 13 some units in an hour of play. 95% of the time, I will win/lose 26 units in an hour, and 99% of the time, the most I could be expected to win/lose in an hour of play would be 39 units.

That's right. The exact confidence intervals are:

1 SD = 68.3%
2 SD = 95.4%
3 SD = 99.7%
4 SD = 99.994%
5 SD = 99.99994%
6 SD = 99.9999998%
7 SD = 99.9999999997%
8 SD = Why bother?!

As you can see, 3 SDs is usually a close enough estimate for most players. Anything beyond that would be fairly unusual.

-Sonny-

 

[Sonny]

by depending greatly on your playing style do you mean action such as backcounting, wonging out, bet ramp? any other factors?

Exactly. The variance will depend on things like how big your bets get, how often you make big bets, the variance at each TC you make your bets, etc. Think of it this way: You bet $1 for 10,000 hands then suddenly bet $10,000 on the next hand. Your variance is going to be greatly affected by that large bet and will gravitate towards the higher bet even though you only did it one time. Such skewing is common with a large bet spread.

The number 1.33 is a general estimation of a card counter's variance. For flat betting you could use 1.1.

-Sonny-

 

[ScottH]

That's right. The exact confidence intervals are:

1 SD = 68.3%
2 SD = 95.4%
3 SD = 99.7%
4 SD = 99.994%
5 SD = 99.99994%
6 SD = 99.9999998%
7 SD = 99.9999999997%
8 SD = Why bother?!

As you can see, 3 SDs is usually a close enough estimate for most players. Anything beyond that would be fairly unusual.

-Sonny-

These intervals are assuming perfectly symmetrical data. Does blackjack have symmetrical results? If not, then those numbers are not nearly that close to 100%.

 

[hopson77]

What is the correlation between bet spread and the number by which you mulitiply the square root of hands played per hour? I would be comfortable with a spread of 5:1, possibly 8:1. I would likely be doing spreads of the following:

TC 0-1 = 1 unit
TC 2-3 = 2 units
TC 4-5 = 3 units
TC 5-6 = 4 units
TC 7-8 = 5 units
TC 9+ = 6+ units

I imagine TCs of 9+ are pretty rare in 4+ deck games, hence why I figure my spread will be 5:1. I know enough to stay away from the 8 deckers, and will try to stick to 6D or DD with decent rules.

 

[kender]

Statistic books

With all these questions and talk about SD and how to figure it, I was wondering if there were any books that would help with learning all this BJ math. Blackjack specific ones would definetly be awsome but a standard stats textbook that is really good for learning the stats most commonly used by BJ players would be good as well.

Also, I don't remember who said it, but I totally understand the "secret society" feeling :laugh:

 

[avs21]

With all these questions and talk about SD and how to figure it, I was wondering if there were any books that would help with learning all this BJ math. Blackjack specific ones would definetly be awsome but a standard stats textbook that is really good for learning the stats most commonly used by BJ players would be good as well.

Also, I don't remember who said it, but I totally understand the "secret society" feeling :laugh:

Blackjack Attack by Don Schlesinger

 

[sagefr0g]

With all these questions and talk about SD and how to figure it, I was wondering if there were any books that would help with learning all this BJ math. Blackjack specific ones would definetly be awsome but a standard stats textbook that is really good for learning the stats most commonly used by BJ players would be good as well.

Also, I don't remember who said it, but I totally understand the "secret society" feeling :laugh:

i don't know of any books that have the pertainent math placed in an appendix so that one could easily refer to it. in most books you find the math dispersed amongst the various chapter within various explainations . it would be an excellent thing for some future author to undertake.
the bjmath site used to be pretty good for a mathematics primer but they may have sold their site..... not sure.

best regards,
mr fr0g

 

[Sonny]

With all these questions and talk about SD and how to figure it, I was wondering if there were any books that would help with learning all this BJ math.

Blackjack Attack is definitely a great resource. Also, pretty much all you need to know is on this free website:

http://www.bjmath.com/bjmath/toc.htm (Archive copy)

I spent about two months reading through all of those articles when I first started learning the math of BJ. The first month I was reading them, then it took me another month to actually understand some of them! :laugh:

-Sonny-