Updated blackjack and card counting simulator

I updated my blackjack and card counting simulator

https://www.blackjack-hack.com

It has updated graphs and data for outcomes and $s of risk.

I have also added outcome tables for strategy plays. When an index number is assigned to a strategy cell the program produces overall hand win/loss/push and $ results for each strategy play. It also produces detailed outcomes at each count level for each play. These tables are shown after the overall results.


The program comes up with default count values, bet levels, game conditions, strategy, hands to play per visit, and number of visits. All or any of this can be changed as needed by the user.

Maximum hands per run are currently 500,000. Overall results however are totaled and averaged for multiple runs.  The detailed index results are not currently totaled and averaged for multiple runs.

Thanks,

Phil
 

Nice piece of work! Suggestion: You need to be able to reckon the TC by half decks remaining as well as full decks, otherwise, those of us who use a level-2 or higher count, such as RPC, will have to double all indices (losing accuracy), because our indices are created to be used with half-deck divisors.

Don

Thanks very much! 

Thank you for the suggestion as well. I will add an option for the true count to be determined by dividing by the number of half decks remaining. 

Phil

I added a choice for divide count by number of full decks remaining or number of half decks remaining. It is below the begin count at field. For running counts this choice has no effect.

The program calculates the number of full decks or half decks remaining and rounds this calculation to the nearest integer resulting in the number of full or half decks remaining.

The true count is then the running count divided by the number of full or half decks remaining. The true count is then rounded to the nearest integer with 1/2 values rounding up away from zero. (1.5 is rounded to 2, -1.5 is rounded to -2). 

Thanks,

Phil

Not a great idea to round the true count up. Most researchers now prefer flooring. When you round a true count up, you often are then overbetting your advantage and/or using indices before they ought to be used.

Don

Thanks, I will change it to floor. With flooring for strategy variations though at a given index number, there will be fewer plus count variations and more minus count variations as compared to rounding. I assume one can adjust their index numbers with this in mind.

Thanks,

Phil

I changed the true count calculation to use flooring instead or rounding.

The decks remaining calculations are still using rounding.

Thanks,

Phil

Don, correct me if I am wrong. Didn't Revere introduce the half-deck calculation in the early 70s of which I have practiced faithfully?
Thanks
Barry

Yes. Why do you ask?

Don

A couple of other comments. Are there maximums for hands per casino visit or number of sessions? I get some messages in red if I make the numbers too large, apparently.

For your 16 vs. T example, you STAND if the TC>=0; you don't hit. Don't give examples as to what you do BELOW an index. Give every example this way: If the TC>=the index, surrender, insure, split, double, or stand. ALL indices are generated to be interpreted that way. Don't reinvent the wheel.

Don

I just saw your comments, question. Currently there is a maximum of 500,000 total hands per run. If the number of hands per visit multiplied by the number of visits (total hands) is greater than 500,000 the red message is displayed. The message essentially asks to lower the number of hands per visit or number of visits or both. But whatever combination is fine for a run as long as the total hands is 500,000 or less. I need to clarify the message.

Results for multiple runs are totaled, so data for greater than 500,000 hands is available. The number of runs is not limited. The configuration can be named to keep multiple sets of runs. The configuration can also be reset to start over.

The current limit is due to a timeout in the hosting plan I have. At some point I may upgrade the hosting plan which should either remove the timeout or greatly increase it. I need to look into that.

I will look into my example text, thanks. Who invented the wheel 

Thanks,
Phil

Resetting or naming the results is on the right below the first table of results, above shuffle settings.

Phil

Simulator question:

When recording hand results after strategy decisions - if there are multiple decisions during a hand should I attribute the result of the hand to each decision or just the last decision (or in between decisions)?

Thanks,
Phil

I increased the maximum number of hands per run from 500,000 to 1,000,000.

Phil

A high winning percentage sounds really good. How much would be high, and how does one do that? 

I have updated my blackjack and card counting simulator.

 

Results are more clearly displayed at the top of the page. Results for each run are shown along with accumulated results for multiple runs.

 

Results can be named. Or a default name will be generated based on the count. The results can be reset to start over or can pick a new name (reset is under the first long results table).

 

Tables with outcomes for strategy variations are generated. Outcomes are also shown for each strategy variation at each count. These tables are below the result tables and graphs.

 

The simulator comes up with a 6 deck game, zen count, and basic strategy. All inputs can be changed as you'd like.

 

https://www.blackjack-hack.com

 

Phil

Great work!  I tried your simulator a lot of times last year but couldn't get a consistent result.  I will try again, but I don't use Zen. Can you make HiLo as the default count, also make the illustrious-18 as the default index deviation? 

Congratulations, Phil. This is a lovely piece of work, and it keeps getting better. I've never seen the results of the deviations broken down the way you've done it. Really great feature.

Further comments/suggestions:

 

1. You need to add surrender as a rules choice. That's rather important.

2. In the top middle, if I have started my lowest bet, below, in the ramp, at $50, why would I see a $25 bet here? That's confusing.

3. Since we base so many of our studies and stats based on 100 hands per hour, can you not make the breakdown default to 100 hands per hour and then add a line for number of hours per visit? That way, the user gets to define a visit as a single hour, four hours, a weekend, a week, or whatever.

4. More importantly than anything, you need to find a way to speed up everything so that the "Run Simulation" button gives us hundreds of millions, if not billions, of hands. As lovely as all of this is, the stats simply aren't terribly reliable and won't converge until a run consists of a great deal more than 400,000 hands.

5. While the count tags are simple enough to change, yes, the default should be Hi-Lo, which is used by more players than all the other counts combined.

6. You need to allow for the possibility of a reverse index, such as when splitting 8,8 vs. T.

 

Keep up the good work!

 

Don

Hi Don,

 

Thank you very much for the compliments! I appreciate it very much! And thank you for trying it out and offering all the great suggestions! 

 

I need to work on adding surrender.

 

The $25 bet is showing up because the base bet at the top is $25. The base bet can be changed. The base bet is what is bet at counts below where the betting ramp starts. There are options (above the betting table) for quitting the shoe at or below a specified count. 

 

I will have to think about the 100 hands per hour and number of hours vs the hands per visit and number of visits. I like the concept of playing x number of hands each time I visit the casino and I am going to visit the casino y number of times. I am wanting to show as one of the main results how much money do you need to bring to the casino. A visit however could possibly be defined as a number of hours and how many hands per hour would give me my hands per visit. Need to think about what makes sense here.

 

I usually run a million hands per run by changing the inputs but it is slow. I need to work on the performance of the whole thing.

 

88 v. T and 88 v. A are actually reverse index options in the program. It stands on 88 at the index or above and splits below the index.

 

Thanks,

 

Phil Satz

I set the default count to Hi-Lo. Thanks!