How much EV do you lose for 9 vs 2?

[seriousplayer]

The illustrious 18 said to double 9 vs 2 at TC >= 1 as EV max and risk averse indices for Hi-lo. Instead of doubling on 9 vs 2 at TC >= 1 what if I double on that play at TC >= 3 will there be significant lost in EV? I am taking about the game 6 deck S17, DAS, LS 75% penetration. It seem like the deviation to double 9 vs 2 at TC >= 1 is only for Hi-lo because in Hi-opt II it said to double at TC >= 3. Not sure how doubling on 9 vs 2 at TC >= 3 will be applicable to a level 1 count in general. I am not using Hi-lo I just want to know how doubling at TC >= 3 and not TC >= 1 for 9 vs 2 affects the level one count in general. Would it result in lose?

[Three]

It seem like the deviation to double 9 vs 2 at TC >= 1 is only for Hi-lo because in Hi-opt II it said to double at TC >= 3. Not sure how that will be applicable to a level 1 count.

My index for HIOPT II is +2 but this is a play were risk aversion ids a consideration. The ace adjustment is -1 per deficit ace.

The only help I have for your question directly can be found here:

http://www.card-counting.com/cvcxonlineviewer3.htm

Choose the S17 game and the 9v2 matchup. Looks like a 0.017 gain in advantage per true count as you progress above the EV maximizing index. It is up to person setting the RA indices to decide how much extra EV makes double the risk worthwhile. .02 or .03 are common thresholds. Risk aversion is tied to the counts correlation to the EOR's for the matchup. The 7 and 8 are the most important low cards and are not counted among the lows using HILO but the counted cards are well behaved with none counted in the wrong direction and only the 3, 4 and A being relatively weak counted cards. The correlation isn't too bad but as you can see from the aces weak status the play is strongly correlated to dealer bust probabilities. The 7 and 8 help the dealer (making his 2 card total 9 or 10) and hurt you (giving you a stiff and 17) making them key cards, but the 9 helps both of you (giving you 18) but the dealer more (giving a 2 card total of 11) making it a somewhat minor low card. HILO is pretty well correlated but missing the 2 most important key low cards makes risk aversion a good idea. The T's are the only key high cards and are counted as high. Risk aversion to either +2 or +3 is a good idea if you have a need for RA indices.

Sorry. Still no direct answer but I hope that helps.

[seriousplayer]

My index for HIOPT II is +2 but this is a play were risk aversion ids a consideration. The ace adjustment is -1 per deficit ace.

The only help I have for your question directly can be found here:

http://www.card-counting.com/cvcxonlineviewer3.htm

Choose the S17 game and the 9v2 matchup. Looks like a 0.017 gain in advantage per true count as you progress above the EV maximizing index. It is up to person setting the RA indices to decide how much extra EV makes double the risk worthwhile. .02 or .03 are common thresholds. Risk aversion is tied to the counts correlation to the EOR's for the matchup. The 7 and 8 are the most important low cards and are not counted among the lows using HILO but the counted cards are well behaved with none counted in the wrong direction and only the 3, 4 and A being relatively weak counted cards. The correlation isn't too bad but as you can see from the aces weak status the play is strongly correlated to dealer bust probabilities. The 7 and 8 help the dealer (making his 2 card total 9 or 10) and hurt you (giving you a stiff and 17) making them key cards, but the 9 helps both of you (giving you 18) but the dealer more (giving a 2 card total of 11) making it a somewhat minor low card. HILO is pretty well correlated but missing the 2 most important key low cards makes risk aversion a good idea. The T's are the only key high cards and are counted as high. Risk aversion to either +2 or +3 is a good idea if you have a need for RA indices.

Sorry. Still no direct answer but I hope that helps.

I did see the index chart viewer. Those indices only apply to Hi-lo count and are not generic. Meaning that it might not apply to other systems.

[Three]

I did see the index chart viewer. Those indices only apply to Hi-lo count and are not generic. Meaning that it might not apply to other systems.

I thought you were using HILO. My bad.

[seriousplayer]

I thought you were using HILO. My bad.

I should've made this clear I will reedit the post.

[DM21]

In BJA3 the RA and Max EV index for 9 vs 2 are both +1 Hi-Lo.

9 vs 2 in a 6D game only has a frequency rate of .0027 or .27% of the time.

The true count of 1 and 2 have a combined frequency rate of 14.4% in a 6D game with 75% pen.

This decision should come up on average of .03888% of the time or about once every 2700 hands or so.

I don't think there will be a big difference over a lifetime if you double at +1 or +3.


JMHO,

DM



****** Edit *********

Are you asking about hi-lo or hi-opt II indices?

[Freightman]

The illustrious 18 said to double 9 vs 2 at TC >= 1 as EV max and risk averse indices for Hi-lo. Instead of doubling on 9 vs 2 at TC >= 1 what if I double on that play at TC >= 3 will there be significant lost in EV? I am taking about the game 6 deck S17, DAS, LS 75% penetration. It seem like the deviation to double 9 vs 2 at TC >= 1 is only for Hi-lo because in Hi-opt II it said to double at TC >= 3. Not sure how doubling on 9 vs 2 at TC >= 3 will be applicable to a level 1 count in general. I am not using Hi-lo I just want to know how doubling at TC >= 3 and not TC >= 1 for 9 vs 2 affects the level one count in general. Would it result in lose?

Essentially saying the same thing as everyone else, except in a different way. Each doubling decision has a point which makes it plus EV over the course of a lifetime. 9v2 is profitable at +1, more profitable at +2, still more profitable at +3 and so forth. The same is true any doubling decision. So (don't have exact numbers in front of me) +1 double captures 55% of EV, +2 captures 60% of EV and so forth. The long term loss at not doubling at +1 is marginal, and not doubling can easily be construed as risk averse. The higher the TC, the more expensive that non double becomes in lost EV.

[Three]

You are right about BJA 3 having no RA index. It doesn't surprise me too much as the counted cards for HILO correlate very well to the play. I would prefer a handle on the density of 6, 7 and 8 for the play or risk aversion. I use +2 for HIOPT II because I use the block side count and adjust by +2 for deficit cards in that group. That is my risk aversion, a very accurate decision for the deck composition. If I didn't have the side count or lost it I would use some risk aversion by using +3 or +4 as my HIOPT II index. That might not be optimal but I like RA plays when key cards are uncounted.

Nice post DM.

[Three]

Now that I look through my notes, HIOPT II EV maximizing index for 9v2 is +2 so +3 is already RA to some degree.

in Hi-opt II it said to double at TC >= 3.

[seriousplayer]

In BJA3 the RA and Max EV index for 9 vs 2 are both +1 Hi-Lo.

9 vs 2 in a 6D game only has a frequency rate of .0027 or .27% of the time.

The true count of 1 and 2 have a combined frequency rate of 14.4% in a 6D game with 75% pen.

This decision should come up on average of .03888% of the time or about once every 2700 hands or so.

I don't think there will be a big difference over a lifetime if you double at +1 or +3.


JMHO,

DM



****** Edit *********

Are you asking about hi-lo or hi-opt II indices?

I am asking about doubling on TC >= 3 for 9 vs 2 use on a level one count like Hi-lo,KO, Hi-opt I, Silver Fox, etc.

[Thirdbaseman]

[QUOTE=Tthree;134786]
The only help I have for your question directly can be found here:
http://www.card-counting.com/cvcxonlineviewer3.htm

This is a great chart.
But is there a glitch re Surrender 8,8 v 10. Should be +1, not +4?
and Surrender 8,8 v 9. Should be +7, not +16?

[Three]

I am asking about doubling on TC >= 3 for 9 vs 2 use on a level one count like Hi-lo,KO, Hi-opt I, Silver Fox, etc.

He gave you the answer to that. Frequency of hand times frequency of the true counts of +1 and +2 for HILO. The graph on the CVCX online viewer gives the cost. Once every 2700 hands you play you will lose the bet size at the TC of +1 or +2 times the advantage differential between hitting and doubling.

So if I have this figured right if you bet $25 at TC +1 and $50 at TC +2 your average will be slightly less than the average due to falling TC frequency as you get further from TC 0. Let's just use $35 to keep the numbers round and an advantage gain of .017. You lose about 60 cents in EV every 2700 hands you play. So if you put in a 1000 hours a year if we use the optimistic average we will get 100 hands/hour it costs $22 per year. EV per hour depends on a lot of things but we will say the hourly EV is $35. That puts the cost in EV at 0.63% of your EV.

I am not certain I got that right. I have one of those nagging feelings I missed something. That nagging feeling is usually right.

[Three]

http://www.card-counting.com/cvcxonlineviewer3.htm

This is a great chart.
But is there a glitch re Surrender 8,8 v 10. Should be +1, not +4?
and Surrender 8,8 v 9. Should be +7, not +16?

Norm, we have a problem here?