How to work with Spanish decks in CVData?

[duanedibley]

Hi,

I have been playing around a bit with the CVData demo, and I think I am finally getting close to making the purchase, but first I have some concerns about using it for Spanish 21.

In various places it seems that "1 deck" is still being defined as 52 cards, despite having a Spanish 21 strategy attached. For index generation I would like the True Count to be calculated by dividing by 48-card decks. This way my playing indices will allow me to calculate the TC by looking at either the discard tray or the shoe.

Is there a way to do this?

And for clarification: I should buy CVData if I want to generate indices, and CVCX if I want to run a simulation with those indices to see my win-rate?

Thanks.

 

[assume_R]

CVData version 5 checks your "modified deck" and knows how many cards are in each deck. CVData version 4 didn't.

cvdata will give you the win rate and simulation result. cvcx allows you to play around with post-sim things such as bankroll requirements, changing penetration on the fly, etc. but doesn't support Sp21 last I checked. CVData v5 includes some cvcx results which supports Sp21.

 

[duanedibley]

CVData version 5 checks your "modified deck" and knows how many cards are in each deck. CVData version 4 didn't.

cvdata will give you the win rate and simulation result. cvcx allows you to play around with post-sim things such as bankroll requirements, changing penetration on the fly, etc. but doesn't support Sp21 last I checked. CVData v5 includes some cvcx results which supports Sp21.

Awesome. Thanks for the reply.

Can I get by with just purchasing CVData for generating Risk-Averse Spanish 21 indices and estimating my win-rate? I am still a little confused about this since there seems to be a lot of interdependence between the two packages.

 

[assume_R]

Awesome. Thanks for the reply.

Can I get by with just purchasing CVData for generating Risk-Averse Spanish 21 indices and estimating my win-rate? I am still a little confused about this since there seems to be a lot of interdependence between the two packages.

Yes you can.

Remember - the indices you'll get from cvdata must be generated using the "Composition Dependent" option, and are 2-card indices. The relative indices (how much to add to each index when a possible bonus is presented, or when you have multi-card hands) is given in Kat's book. Therefore, your actual win rate will be slightly higher than what the sims give you, since will be presumably using relative indices in actual play. However, in my tests I've found that relative indices don't add too much.

 

[duanedibley]

OK I now have the full version of CVData! However, I am getting some strange results generating indices for Spanish 21.

I am generating Risk Averse indices with multiple passes, and so far everything is coming out to about half of what it should be based on other people's results. I have +4 where I was expecting a +9, a few +3's where I was expecting +7's, -1's where I was expecting -2 and -3. etc.

On the first pass the sim reported a score of -15, which I assume was for the basic strategy I attached, then on the second pass it reported a score of -120! I am guessing this is from using the bad indices.

My decks seem to have the right composition, and my IRC's are set to what I wanted, so I am lost as to where to start looking for the problem.

Any advice??

 

[duanedibley]

And what on earth is up with the included "Spanish 21 HiLo for index generation" strategy. It doesn't look anything like basic strategy. Is this intended?

 

[Automatic Monkey]

OK I now have the full version of CVData! However, I am getting some strange results generating indices for Spanish 21.

I am generating Risk Averse indices with multiple passes, and so far everything is coming out to about half of what it should be based on other people's results. I have +4 where I was expecting a +9, a few +3's where I was expecting +7's, -1's where I was expecting -2 and -3. etc.

On the first pass the sim reported a score of -15, which I assume was for the basic strategy I attached, then on the second pass it reported a score of -120! I am guessing this is from using the bad indices.

My decks seem to have the right composition, and my IRC's are set to what I wanted, so I am lost as to where to start looking for the problem.

Any advice??

My advice is to set up a separate configuration for index generation that just uses one bonus- the auto-win on 21 rule. Generate all your indices using the comp dependent feature (10,3 for 13, etc.) and transfer your results into a from-scratch playing strategy that you actually run in another configuration that has all the bonuses active.

Make sure you have the DDR function set if you are generating doubling indices because the software recognizes that, but it doesn't handle multi-card double so you're going to have to get creative if you want accurate doubling indices for anything lower than hard 10.

 

[duanedibley]

My advice is to set up a separate configuration for index generation that just uses one bonus- the auto-win on 21 rule. Generate all your indices using the comp dependent feature (10,3 for 13, etc.) and transfer your results into a from-scratch playing strategy that you actually run in another configuration that has all the bonuses active.

Make sure you have the DDR function set if you are generating doubling indices because the software recognizes that, but it doesn't handle multi-card double so you're going to have to get creative if you want accurate doubling indices for anything lower than hard 10.

Thanks, Monkey, I will try this approach.

 

[duanedibley]

OK, I had "Custom True Count Calculation" checked and for some reason the table was filled out to divide by twice the remaining number of decks in all cases - so that explains pretty well why all my indices appeared to be half of what I was expecting.