Ok, In Wong’s book Pro. BJ , for a2 v 5, look at his basic strategy chart vs his strategy chart (deviations).
On Basic strategy it’s a hit. His index number on the deviation chart is 0 which is telling me to double.
Isn’t a true 0 a basic strategy play?
Absolutely correct - the cards on the table give composition dependent basic strategy - you know the dealer's up card and your two cards.Originally Posted by bigplayer
A2 v 5 double is a very marginal double as are A4 v 4, 3,3 v 2 DAS and 4,4 v 5 DAS. All of these indices are very close to zero.
For A2 v 5 the HL indices are -0.51, 0.03, 0.38. 0.42 and 0.55 for 1, 2, 6, 8 and infinite decks respectively.
Also you can see that for 6 or 8 decks, the index was rounded down to 0.
Last edited by bjanalyst; 10-03-2020 at 06:53 PM.
I am sure that you have heard of risk averse indices. For doubling, for example, right at the index, the expected value of hitting and doubling are the same.
But why risk extra money by doubling at the index when there is no extra return - you have increased risk and no increased expectation. What I do is increase the doubling or splitting expected value indices by one or two true count points - I do this for all doubling and splitting indices which is what is called risk averse indices.
In this case you are doubling with an index that is basically zero. So I would wait to double using an index of 1 or even 2 since at true counts close to the index of zero the expected value of doubling and hitting are the same.
The farther you are away from the index the more is your gain from doubling or splitting and that is where you get most of the gain from the double or split. .
The formula is pa(t) = AACpTCp*(t - Idx) where t = true count (HL true count here), AACpTCp = Average Advantage Change per true count point, pa(t) = player's advantage of true count "t", and Idx = index.
So for A2 v 5 doubling right at the index of zero just increases variance with no extra gain in expected value as doubling and splitting have the same expected value at the index. For A2 v 5 AACpTCp = 1.1%.
So for the four very close basic strategy doubles and splits that I mentioned, namely, A2 v 5, A4 v 4, 3,3 v 2 DAS and 4,4 v 5 DAS, I would wait to double or split until tc(HL) >= 1 where tc(HL) = High Low True count. If you want to be more conservative you can wait to double or split until tc(HL) >= 2.
Last edited by bjanalyst; 10-04-2020 at 08:04 AM.
You cannot calculate indices mathematically to two decimals, and results expressed to more decimals at which they are accurate are wrong. And what penetration are you talking about with such exact numbers?
Last edited by Norm; 10-04-2020 at 08:23 AM.
"I don't think outside the box; I think of what I can do with the box." - Henri Matisse
Or, you could use the correct RA values. It has nothing to do with "wanting to be more conservative". Waiting longer than the RA index will decrease SCORE.
Last edited by Norm; 10-04-2020 at 08:27 AM.
"I don't think outside the box; I think of what I can do with the box." - Henri Matisse
The issue of EV Maximizing versus Risk Averse is an interesting one.
The theoretical lost opportunity of not doubling at index is compensated by RA reduced variance, allowing in theory higher max bets, thus increasing profits. Practically, as Don pointed out one one occasion, the RA user might maintain his max bet using the gain from RA to be further applied against variance.
As Norm has stated, waiting beyond RA index absolutely reduces SCORE.
SCORE, which is such an important concept, is not necessarily the PRACTICAL end all be all solution. Case in point(s) would CVCX sims whereby user at a multiplayer table and experiments with various true counts at when to move to 2 hands. A typical sim at .5 house edge would show maximum SCORE at +1. I never do. I might only get 1 hand and by sim, revert back to 1 hand. I wait until true 2, sometimes 3, which allows for a few rounds of play at 2 hands.
Another SCORE reduction would be 1st bet at new shoe above your min bet. For exampLe, I recently described my initial foray in Covid time at a $5-$200 table - playing $10 off the top, not reducing to $5 until -1 or crappier if winning. That Cost can be handled nicely with a 1-12 spread. I played that table to a 2x90 and 2x150 lower and upper ramp respectively - effect on SCORE being virtually nothing.
Sims and SCORE should be utilized to fashion strategy.
Last edited by Freightman; 10-04-2020 at 09:16 AM. Reason: Adding last point
Posting Permissions
Reply With Quote
Bookmarks