BackTesting Blackjack Counting Stratergy

Hi All,
I have been a silent follower of this site and WizardOfOdds for a long time. Over the last few months I think I have mastered the Basic Strategy. Now I want to move over to some very basic Card Counting Strategies like Ace/Five or Simple Count with a Bet Spread of say 1-4 or maybe come up with a very simple strategy of my own. My Idea is not to generate consistent +ve EV, but to enjoy maximum time in Casino, and any $$$ i make in the process is just a bonus. It would be great if my card Counting skills help me pay for my trip to Vegas/Reno or gain some comps in the casino.

One Idea of a Strategy is not maintaining any running count, but to bet the next hand based on the count of the current hand. Was looking at options of executing the same and running a few simulations to see how it performs.

My Question to all of you is "How do you backtest your Card Counting strategies". Does CVData give me the option of creating my own card counting strategies and run simulations against it? I tried writing a program to back-test, but have kind of given up as there were too many bugs in it. Would not like to spend more time on that if there is any tools/product available for me to test my ideas. My day job is Software Development, so I can write code, if needed.

Quote: LearningToCount

<snip>My Question to all of you is "How do you backtest your Card Counting strategies". Does CVData give me the option of creating my own card counting strategies and run simulations against it?<snip>

LearningToCount,

Yes, CVData is designed to test card-counting strategies, both canned and user-created.

Dog Hand

For someone like you that isn't even "overly" worried about generating EV, I wouldn't recommend CVData since it might overwhelm you. Don't get me wrong, if you want to get in to counting to actually generate and make money, then you should get it. Otherwise it's just overkill. Simply get 6 decks of cards together, shuffle them, and deal 3 hands to yourself playing and counting. Do this over and over until you know the last card 10/10 times in a row. I talk about "how to train" a little bit in my 3 articles on A-Z Counting Cards in Blackjack in the articles section of this very site. There's also free online resources for training (counting systems tho, not A-5).

I also talk about how you should not learn the A-5 count, at least it's not my recommendation. There's a lot of reasons. The system, at best, is a good one for earning comps and trying to break even... but that's at larger bet spreads/etc. Collin from Blackjack Apprent. just put out a video about it as well:



I would definitely urge you to review the 3 articles I've written. EVEN IF you don't want to be a full time pro, learning the "wrong" way is never a good idea. Even with the A-5 count you still need a bankroll. You still need to understand Kelly Betting and Risk of Ruin, etc. The system itself isn't going to save you from ANY of that, so why not learn to add and subtract 1's instead of counting Aces and 5's? Then you'll actually generate a small amount of EV over the long haul and you can have a lesser close to break even spread for generating comps all the same.

https://wizardofvegas.com/articles/A-to-Z-Counting-Cards-in-Blackjack/
https://wizardofvegas.com/articles/A-to-Z-Counting-Cards-In-Blackjack-2/
https://wizardofvegas.com/articles/A-to-Z-Counting-Cards-in-Blackjack-3/

Quote: LearningToCount

Hi All,
I have been a silent follower of this site and WizardOfOdds for a long time. Over the last few months I think I have mastered the Basic Strategy. Now I want to move over to some very basic Card Counting Strategies like Ace/Five or Simple Count with a Bet Spread of say 1-4 or maybe come up with a very simple strategy of my own. My Idea is not to generate consistent +ve EV, but to enjoy maximum time in Casino, and any $$$ i make in the process is just a bonus. It would be great if my card Counting skills help me pay for my trip to Vegas/Reno or gain some comps in the casino.

One Idea of a Strategy is not maintaining any running count, but to bet the next hand based on the count of the current hand. Was looking at options of executing the same and running a few simulations to see how it performs.

My Question to all of you is "How do you backtest your Card Counting strategies". Does CVData give me the option of creating my own card counting strategies and run simulations against it? I tried writing a program to back-test, but have kind of given up as there were too many bugs in it. Would not like to spend more time on that if there is any tools/product available for me to test my ideas. My day job is Software Development, so I can write code, if needed.

As far as easy to learn counts go, Olaf Vancura and Ken Fuchs KO count are about the easiest you can learn. I used Arnold Snyder's Red Seven for a time too and because of the way they are designed you don't have to keep a true count, which for me made counting many, many times easier. Wishing you the best with it and would also advise you take whatever advice Romes gives ya. He's really experienced and knowledgeable on the topic!

Hi Romes,

Thanks for your advice. I will surely read all the 3 articles. I carefully follow the videos/blogs from Michael Shackleford and Phil Collins. I did see the video from Collins on Aces/Five Count just after posting in this forum, and realized that the risk of ruin is too high from my comfort. I guess i will practice the KO Count then, I somehow really get confused adding and subtracting numbers, will practice it however.

On running simulations, I got my Program running with Basic Strategy and should be able to customize with any card counting technique i need. I am getting a house edge or around 0.40% with the following Rules
8 Deck
Penetration 75%
Dealer Stands on all 17s.
DAS Allowed
Surrender Allowed
No Resplit Aces
Split Upto 2 Hands (Would like to Improvise on this).
Late Surrender.
Blackjack Pays 3:2.

This is close to the Theoretical number of 0.42% from Wizard of Odds Website. I am willing to share the code if anyone is interested.

Finally Once i am ready, is it a good idea to practice my strategy on a Table where the BlackJack pays 6:5. I feel that pit boss would not be too bothered about heat on these tables compared to 3:2. Also the minimum at these tables are $15 compared to $50 at the 3:2 tables.

Quote: LearningToCount

Hi Romes,

Thanks for your advice. I will surely read all the 3 articles. I carefully follow the videos/blogs from Michael Shackleford and Phil Collins. I did see the video from Collins on Aces/Five Count just after posting in this forum, and realized that the risk of ruin is too high from my comfort. I guess i will practice the KO Count then, I somehow really get confused adding and subtracting numbers, will practice it however.

On running simulations, I got my Program running with Basic Strategy and should be able to customize with any card counting technique i need. I am getting a house edge or around 0.40% with the following Rules
8 Deck
Penetration 75%
Dealer Stands on all 17s.
DAS Allowed
Surrender Allowed
No Resplit Aces
Split Upto 2 Hands (Would like to Improvise on this).
Late Surrender.
Blackjack Pays 3:2.

This is close to the Theoretical number of 0.42% from Wizard of Odds Website. I am willing to share the code if anyone is interested.

Finally Once i am ready, is it a good idea to practice my strategy on a Table where the BlackJack pays 6:5. I feel that pit boss would not be too bothered about heat on these tables compared to 3:2. Also the minimum at these tables are $15 compared to $50 at the 3:2 tables.

Don't play 6:5 ..EVER
If you want to avoid heat while you are practicing, keep a low spread.. or practice at home. Practicing at a 6:5 is very expensive.
With min bet and basic strategy, no counting, you will lose more on the 6:5 playing 15$ than 3:2 playing 50$

PS: If these are your only 2 options... Sorry I mean if this is your only option (6:5 is something you should always pretend doesn't exist) I can't recommend starting there. 50$ min is a bit steep for a beginner and unnecessary risk if your goal is to make money. That is unless you are already playing at that level and enjoying the comps. Don't care about ROR and just trying to improve your odds a bit, than that would be gambling while counting the cards to improve your odds a bit..