Doubling A,8

[matt21]

I have a question and would value anyone's feedback.

I have the following Hi/Lo-based index levels for A,8:

Double against 6 @ 1+
Double against 5 @ 1+
Double against 4 @ 3+

How do I evaluate the increase in EV from doubling (as opposed to standing) AND the change in ROR of doubling versus standing. Is there some established approach for assessing or 'grading' the decision to double from a EV/ROR standpoint i.e. to enable you to weigh the increase in EV against the increase in ROR?

Many thanks in advance for any comments and good luck to you all
(off to bed i go!)

 

[Automatic Monkey]

I have a question and would value anyone's feedback.

I have the following Hi/Lo-based index levels for A,8:

Double against 6 @ 1+
Double against 5 @ 1+
Double against 4 @ 3+

How do I evaluate the increase in EV from doubling (as opposed to standing) AND the change in ROR of doubling versus standing. Is there some established approach for assessing or 'grading' the decision to double from a EV/ROR standpoint i.e. to enable you to weigh the increase in EV against the increase in ROR?

Many thanks in advance for any comments and good luck to you all
(off to bed i go!)

Yes what you are looking for are risk averse indices. CVData will calculate them automatically. If you are using some other method of generating indices, you want the index to be at the point where the added advantage from the double/split is equal to the overall advantage at the count.

 

[moo321]

As far as I can remember, risk averse indices weren't worth a whole lot compared to regular indices. 10 v. 10 was the only really important one.

 

[kewljason]

I have a question and would value anyone's feedback.

I have the following Hi/Lo-based index levels for A,8:

Double against 6 @ 1+
Double against 5 @ 1+
Double against 4 @ 3+

How do I evaluate the increase in EV from doubling (as opposed to standing) AND the change in ROR of doubling versus standing. Is there some established approach for assessing or 'grading' the decision to double from a EV/ROR standpoint i.e. to enable you to weigh the increase in EV against the increase in ROR?

Many thanks in advance for any comments and good luck to you all
(off to bed i go!)

I'm wondering if you have considered throwing "cover" as well as EV/ROR into the equation of when to make these plays.

I frequently take the train to/from Atlantic City, which depending on the time of day can be filled with casino employees. Their endless chatter is usually mundane, however occassionally I pick up useful information from them. (BTW I don't consider this eavesdropping if I can hear them several rows away while trying to relax or nap)

Anyway one night a couple suits and dealers where discussing counters. One suit said he spots counters more by their plays than their spread. He said lots of people jump their bets all around but only counters split tens, double on soft 20 or soft 19. I refuse to give up these plays altogether, but for this reason, I've always been a bit more conservative making these plays usually doubling at a slightly higher count than the indices call for. In the case of splitting tens, I also end my session at the end of that shoe.

He also said how a person plays soft 18 vs a 9, 10 is a immediate tell as to whether they know basic strategy. He claimed people who are complete ploppies are very satisfied with an 18, (even though 18 is a loser more often than a winner). just a thought for what it's worth.

 

[creeping panther]

Kewl

I'm wondering if you have considered throwing "cover" as well as EV/ROR into the equation of when to make these plays.

I frequently take the train to/from Atlantic City, which depending on the time of day can be filled with casino employees. Their endless chatter is usually mundane, however occassionally I pick up useful information from them. (BTW I don't consider this eavesdropping if I can hear them several rows away while trying to relax or nap)

Anyway one night a couple suits and dealers where discussing counters. One suit said he spots counters more by their plays than their spread. He said lots of people jump their bets all around but only counters split tens, double on soft 20 or soft 19. I refuse to give up these plays altogether, but for this reason, I've always been a bit more conservative making these plays usually doubling at a slightly higher count than the indices call for. In the case of splitting tens, I also end my session at the end of that shoe.

He also said how a person plays soft 18 vs a 9, 10 is a immediate tell as to whether they know basic strategy. He claimed people who are complete ploppies are very satisfied with an 18, (even though 18 is a loser more often than a winner). just a thought for what it's worth.

Good info there...keep listening

CP

 

[Automatic Monkey]

As far as I can remember, risk averse indices weren't worth a whole lot compared to regular indices. 10 v. 10 was the only really important one.

Right, but they're freebies. If you're going to learn an index you might as well use the risk-averse one. Using the non-risk averse one is exactly the same thing as overbetting.

 

[rukus]

Right, but they're freebies. If you're going to learn an index you might as well use the risk-averse one. Using the non-risk averse one is exactly the same thing as overbetting.

not only are they freebies, but they do add more than just a slight % advantage. they might not add up to much in terms of % advantage gained, BUT when coupled with kelly betting, it becomes more powerful... becomes a multiplier of EV in terms of $s..

Edit for example: say RA adds .1% to your advantage. if you bet by some fraction of kelly, ie proportional betting, you will also bet an additional .1% of your BR. hence your $EV is (1.001*1.001*usual bet) as opposed to just (1.001*usual bet).

 

[nottooshabby]

kewljason,

Good to see you back posting, and thanks for the info

 

[kewljason]

kewljason,

Good to see you back posting, and thanks for the info

Thank shabby. I read the board everyday. Just don't post as much. And I'm just staying out of any personal attack threads. Its just not my purpose here.

 

[jack.jackson]

not only are they freebies, but they do add more than just a slight % advantage. they might not add up to much in terms of % advantage gained, BUT when coupled with kelly betting, it becomes more powerful... becomes a multiplier of EV in terms of $s..

Edit for example: say RA adds .1% to your advantage. if you bet by some fraction of kelly, ie proportional betting, you will also bet an additional .1% of your BR. hence your $EV is (1.001*1.001*usual bet) as opposed to just (1.001*usual bet).

So, is this whats known as Risk-averse betting?

"IN practice, I like to parlay my bets, on wins only. I know this lowers your SCORE and Win-rate, but it also lowers your RoR. If I was short-stacked I would probably use this method.

 

[matt21]

Yes what you are looking for are risk averse indices. CVData will calculate them automatically. If you are using some other method of generating indices, you want the index to be at the point where the added advantage from the double/split is equal to the overall advantage at the count.

Thanks for this response - i will look into this in more detail.

And also for the comments regarding the need to take 'heat' into the equation, and the possibility that casino staff pin-point counters by some of the BS index plays.

Very appreciated everyone!

 

[zengrifter]

IN practice, I like to parlay my bets, on wins only. I know this lowers your SCORE and Win-rate, but it also lowers your RoR. If I was short-stacked I would probably use this method.

Huh? ... I think it INCREASES ROR. Can we gat a fact checker here on Aisle 19? zg

 

[Automatic Monkey]

Huh? ... I think it INCREASES ROR. Can we gat a fact checker here on Aisle 19? zg

It lowers your RoR if you do it on wins, relative to doing it on losses or randomly. By definition, you are increasing your risk only after you've enjoyed good variance.

There's a spread I use for SD games that works a lot like this, where I play a positive progression in positive counts and a negative one in negative counts. It's the equivalent of a pretty damned big spread, and although the RoR is slightly higher than if I just laid the spread openly, I wouldn't be able to get away with that spread in SD.

 

[jack.jackson]

Huh? ... I think it INCREASES ROR. Can we gat a fact checker here on Aisle 19? zg

Huh? ... I think it INCREASES ROR. Can we gat a fact checker here on Aisle 19? zg

Here we go. 6D, 75% , 1-15 spread

This is the Betting Strategy I used. Never more than Double your bet, on a win, and NO increases on lose.

Example: If you bet 5 units(lastbet) and TC jumps +5. Bet 10units on Win and stay 5units if lose. (conservative)

Standard proportion
+2/20
+3/50
+4/100
+5>150

Standard betting(below)
10K
RoR/21.68

Conservative betting(below)
10K
RoR/19.23

The decrease in EV is significant. Later I will run a sim that does full spread on wins and NO more than double on losses and see what that does. Stay tuned Doc.

 

[bjcount]

Here we go. 6D, 75% , 1-15 spread

This is the Betting Strategy I used. Never more than Double your bet, on a win, and NO increases on lose.

Example: If you bet 5 units(lastbet) and TC jumps +5. Bet 10units on Win and stay 5units if lose. (conservative)


[
The decrease in EV is significant. Later I will run a sim that does full spread on wins and NO more than double on losses and see what that does. Stay tuned Doc.

JJ
Those are some awfully poor SCORES your coming up with. What's the rule set your using?

BJC

Edit: Here's a set of 6d,s17, das, Zen RA indices. See if they improve your results.

 

[jack.jackson]

JJ
Those are some awfully poor SCORES your coming up with. What's the rule set your using?

BJC

Edit: Here's a set of 6d,s17, das, Zen RA indices. See if they improve your results.

Thanks, but its the H17 rule I was using. I just ran another sim(complete zen) w/S17,DAS,RSA, verses the above sims, H17,Das,Rsa and it spit-out a whimisical 29.54 score with full betting, just from changing H17(21.68) to S17(29.54).

 

[bjcount]

Thanks, but its the H17 rule I was using. I just ran another sim(complete zen) w/S17,DAS,RSA, verses the above sims, H17,Das,Rsa and it spit-out a whimisical 29.54 score with full betting, just from changing H17(21.68) to S17(29.54).

After eliminating some of the extreme indices which rarely ever get hit, I have found that the RA indices produce better results than the complete set. Have you tried any sims with RA?

BJC

 

[jack.jackson]

After eliminating some of the extreme indices which rarely ever get hit, I have found that the RA indices produce better results than the complete set. Have you tried any sims with RA?

BJC

Not yet, but I know what you mean by eliminating some of those extreme indices. I kept trying to get CVdata to generate A2-A5 vs 2 for my Level 3, like the ones Ho2 has for these hands, but then i realized it wouldnt generate them because my level 3 reckoned the 9, where as Ho2 does not. Once I removed these extreme indices, after manually inputting them for my count, my score shot up pretty nicely. I kept trying to tweak these indices, when I probably should of been using RA's the whole time. Do they really improve your Score? I wouldnt think they would, thats why i never tried them.

 

[bjcount]

Not yet, but I know what you mean by eliminating some of those extreme indices. I kept trying to get CVdata to generate A2-A5 vs 2 for my Level 3, like the ones Ho2 has for these hands, but then i realized it wouldnt generate them because my level 3 reckoned the 9, where as Ho2 does not. Once I removed these extreme indices, after manually inputting them for my count, my score shot up pretty nicely. I kept trying to tweak these indices, when I probably should of been using RA's the whole time. Do they really improve your Score? I wouldnt think they would, thats why i never tried them.

I have run a few sets or RA indices for different strats and different games. All have outperformed their max ev indices.

Using RPC tags I ran a "beat to death" sim (3 days long on my centrino duo, t5500 @1.666 GHz, 980MHz)
Instead of trying to learn 160+ indices for one set I used the typical TC range encountered for RPC, TC+8 to TC-6 (6d-8d), changed all the extreme one's to BS, and still was left with 90+ indices which I use in real games. The SCORE improved 10%-22% depending upon the rule set and opt betting config of a 1-12 spread.

No comments please about playing at negative counts.

With the set of Zen RA indices in my previous post the SCORE came in 20% higher than the complete set in CVData, same game, same rules, same bet spread.


BJC

 

[jack.jackson]

I have run a few sets or RA indices for different strats and different games. All have outperformed their max ev indices.

Using RPC tags I ran a "beat to death" sim (3 days long on my centrino duo, t5500 @1.666 GHz, 980MHz)
Instead of trying to learn 160+ indices for one set I used the typical TC range encountered for RPC, TC+8 to TC-6 (6d-8d), changed all the extreme one's to BS, and still was left with 90+ indices which I use in real games. The SCORE improved 10%-22% depending upon the rule set and opt betting config of a 1-12 spread.

No comments please about playing at negative counts.

With the set of Zen RA indices in my previous post the SCORE came in 20% higher than the complete set in CVData, same game, same rules, same bet spread.


BJC

Are you sure about that? I just used your RA vs CVdatas(maximum ev) for 6D,DAS,S17,RSA,75%,1p,1-15(reset bets after shuffle/CHECKED)

Maximum EVs(below)

I figured this would be the other way around. Wow Im shocked!