RA index in CVData

[Phoebe]

"Important" risk-averse index plays are a function of three things: 1) how frequently the play arises; 2) the percentage of your bank that is wagered on the hand, and 3) the volatility of the play (i.e., how quickly the advantage changes as the TC rises). Since these plays fail the very first frequency tests, they will never be "important."
Don

Concretely, how to calculate RA index in CVData ?
Where and how to enter point 2 ?

[Norm]

CVData runs a pre-sim which creates data that is used to calculate the percentage of your Kelly bank that should be bet at each TC as well as the volatility. This info is used during the index generation along with the EVs for each TC for each decision to determine the RA index. I'm afraid the process is rather complex.

[Phoebe]

Suppose I play 6-deck regular REKO-F, with 1.83% RoR (Half-Kelly)
If I well understand, I simply enter Max Bet @4 and Ratio=2.0 !?

[seriousplayer]

Talking about REKO-F there is some imperfections regarding some of the playing indices I want to get answer from Norm. I don't think it is appropriate to start a new conversation in this thread but I will start a separate threads on this topic.

[Phoebe]

Norm answered a few months ago.
REKO-F in "Modern Blackjack" are NOT all with RA. Wrong copy/paste.
A least Soft DD in one sheet.

https://www.blackjacktheforum.com/sh...l=1#post182873

[Norm]

Suppose I play 6-deck regular REKO-F, with 1.83% RoR (Half-Kelly)
If I well understand, I simply enter Max Bet @4 and Ratio=2.0 !?

Max bet can be off a bit without a problem. +4 is fine. Leave ratio at 1.0. I need to find a way to explain this better in the documentation. Basically, you increase the number if your bankroll is so large that you are betting below the risk that you are aiming for because you are playing a game that doesn't allow you to bet as much as your bankroll and desired risk allow. Increasing the ratio decreases the risk-aversion.

[Phoebe]

Clear but, yes, documentation should explain better
Ballon said "2.0 would indicate betting at half-kelly"

[seriousplayer]

Leave ratio at 1.0. I need to find a way to explain this better in the documentation.

It seem like the risk aversion ration cannot go below 1.0.

[Norm]

A ratio of 1.0 is RA. An index of infinity is EV-Maximizing. Ratios between them would be varying compromises between them. A ratio below 1.0 wouldn't be EV-maximizing or RA.

[Phoebe]

When, in an other thread Don said:

So my answer is: Any extra gain in e.v. attained by reaching the index has to be tempered by the extra risk. For A,6 vs. 2, you would probably never double. For A,7 vs. 2, there are different r-a indices depending on the size of the wager relative to the bankroll. The larger the bet, the higher the index.

I guess "ratio" is the way to calculate these different indices !?

[Norm]

If you have a very large bankroll, you could tune the RA accordingly by increasing the ratio. However, there are very few indices where this will make a difference -- unless your bankroll is huge, in which case there is no reason to use RA indices.

[BJFan]

Norm,

Phillipe B. posted this somewhere:

CVData without RA Indexes
Soft Double Down - A7 vs. 2: DD >= 1
Soft Double Down - A6 vs. 2: DD >= 2

With RA Indexes (Max bet @6)
Soft Double Down - A7 vs. 2: DD >= 1
Soft Double Down - A6 vs. 2: DD >= 9

With RA Indexes (Max bet @3)
Soft Double Down - A7 vs. 2: No Double Down
Soft Double Down - A6 vs. 2: No Double Down

Why "No double Down" when Max bet @3? There must be an index as there's an index for non RA and for Max bet @6.

[Norm]

I'd have to see the exact sim.