Disregard Proper Basic Strategy?

[mushin]

In another thread a statement was made to the effect that it would not be worth doubling down for the slight advantage it gives in the interest of bankroll preservation.

I find this notion extremely crazy but I would like to put this out there to the community. All things aside is there ever a time not to make a proper Basic Strategy play or even index play in the interest of "preserving your bankroll?"

The only bankroll that I see you preserving by not making a the proper play according to BS or the count is that of the House.

[Bodarc]

If you are not going to follow basic strategy or your counting strategy then the best strategy is to not play at all unless you are making cover plays where the difference in two choices is miniscule, but then if no one is watching, why even do that.

Variance is built into your ROR computation so as long as your ROR is not excessive, why vary your play?

[MJGolf]

Which thread Mushin? I'm surprised that was an actual experienced comment, when all math shows that you should double down when you have the opportunity and don't ever "double for less". Maybe it was in a different context? The only time I could see that being a possibility is if you were at a particularly large count; had a very large bet out; and were reaching some bankroll "goal" or all time high, and didn't want to take the chance to hurt that opportunity. But the "numbers" or "correct" decision, would always say double when you had that opportunity, I would believe.

[ohbehave]

Correct me if I'm wrong, it was in the context of an index deviation, example doubling 10v10/A. So the decision is between making the index play or making the BS play. Both have a player advantage, you are not giving up an advantage by staying with BS in this case. What you are giving up is a very slight additional advantage in exchange for lower variance. Both decisions would be correct in as far as having an advantage.

[mushin]

Correct me if I'm wrong, it was in the context of an index deviation, example doubling 10v10/A. So the decision is between making the index play or making the BS play. Both have a player advantage, you are not giving up an advantage by staying with BS in this case. What you are giving up is a very slight additional advantage in exchange for lower variance. Both decisions would be correct in as far as having an advantage.

By not doubling you are indeed giving up an advantage that has been precalculated into the house edge for the particular game you are playing... much in the same was that the house edge changes for H17, S17 or surrender. Lets say you are on a table where you can surrender... but you never take advantage. You are giving up the advantage that rule provides you. The same would be true for a double... by not doubling your give up whatever advantage is pre-calcualted into the house edge according to the BS for that game.

If we take that further and choose not to double a hand according to your index plays would you not still be giving up the advantage associated with it? While you may be trying to limit your perceived risk it is just that perceived risk. Would it not be better even for someone with a limited bankroll to get as much money on the table as possible when they have the advantage? When we have our max bets out... Isn't a double opportunity simply just a way to temporarily increase our bet spread anyway?

[ohbehave]

By not doubling you are indeed giving up an advantage that has been precalculated into the house edge for the particular game you are playing... much in the same was that the house edge changes for H17, S17 or surrender. Lets say you are on a table where you can surrender... but you never take advantage. You are giving up the advantage that rule provides you.

If we take that further and choose not to double a hand according to your index plays would you not still be giving up the advantage associated with it? While you may be trying to limit your perceived risk it is just that perceived risk. Would it not be better even for someone with a limited bankroll to get as much money on the table as possible when they have the advantage? When we have our max bets out... Isn't a double opportunity simply just a way to temporarily increase our bet spread anyway?

I get all that, but in the end both decisions are correct from an advantage point of view. It comes down to the players' risk tolerance, and that is what Bosox was trying to convey.

The same would be true for a double... by not doubling your give up whatever advantage is pre-calcualted into the house edge according to the BS for that game.

Actually, here if it is an index play the house edge is for the BS decision, not the index decision.

[The Suburbs]

By not doubling you are indeed giving up an advantage that has been precalculated into the house edge for the particular game you are playing... much in the same was that the house edge changes for H17, S17 or surrender. Lets say you are on a table where you can surrender... but you never take advantage. You are giving up the advantage that rule provides you. The same would be true for a double... by not doubling your give up whatever advantage is pre-calcualted into the house edge according to the BS for that game.

If we take that further and choose not to double a hand according to your index plays would you not still be giving up the advantage associated with it? While you may be trying to limit your perceived risk it is just that perceived risk. Would it not be better even for someone with a limited bankroll to get as much money on the table as possible when they have the advantage? When we have our max bets out... Isn't a double opportunity simply just a way to temporarily increase our bet spread anyway?

This is not the correct way of thinking. If you had a $10,000 bankroll and I told you that I would be willing flip a coin with you for $10,000 where you get paid $10,001 if you get it right, would you take that bet? No, of course not. While taking the bet is the best way to maximize EV, one needs to consider risk.

To use a blackjack analogy, let's say you knew you had a 10% advantage (through whatever means) for a given hand. Also, let's say you're playing approximately full kelly with a $10,000 bankroll. Thus, you're betting around $1,000 on the hand. If you're dealt a pair of 10's vs. dealer's 6 and the TC is 4, would you split them? What about if you got another 10 after splitting them (assuming the TC was still 4 after)? What about another? Get the point?

[mushin]

This is not the correct way of thinking. If you had a $10,000 bankroll and I told you that I would be willing flip a coin with you for $10,000 where you get paid $10,001 if you get it right, would you take that bet? No, of course not. While taking the bet is the best way to maximize EV, one needs to consider risk.

To use a blackjack analogy, let's say you knew you had a 10% advantage (through whatever means) for a given hand. Also, let's say you're playing approximately full kelly with a $10,000 bankroll. Thus, you're betting around $1,000 on the hand. If you're dealt a pair of 10's vs. dealer's 6 and the TC is 4, would you split them? What about if you got another 10 after splitting them (assuming the TC was still 4 after)? What about another? Get the point?

Risk would have already been calculated based on your bet ramp would it not. So if you are willing to play $1000 a hand with only a 10k bankroll should you not take those decisions into account in advance.. But... you are correct I am not willing to bet my 10k on one flip of the coin but the correct play would be to do so. Still that is not the game we are playing... at least I hope no one is thinking they can be a AP and bet their entire stake on a 1% advantage. Still this would not be too much different from being close to tapping out and being forced into the double or not decision or any decision for that matter. Not making the correct play is what increases your risks (Decreased advantage). Making the correct play decrease you risk. Whether or not you win or loose is a matter of fluctuation.

[Nyne]

If you want to get really technical, you can definitely get into situations where the EV maximizing play is not the optimal play, particularly when your bet is a relatively large fraction of your bankroll. But for normal size bets, just use the correct risk averse indices and you'll be fine. Let CVData chug through the computations and try not to second-guess the output. For an in depth discussion of a situation where this could be relevant, read this take from James Grosjean and Previn Mankodi: http://www.blackjackforumonline.com/...t/ACEbjfo1.pdf

[UK-21]

In another thread a statement was made to the effect that it would not be worth doubling down for the slight advantage it gives in the interest of bankroll preservation.

I find this notion extremely crazy but I would like to put this out there to the community. All things aside is there ever a time not to make a proper Basic Strategy play or even index play in the interest of "preserving your bankroll?"

The only bankroll that I see you preserving by not making a the proper play according to BS or the count is that of the House.

I suppose the question is what's the trade off between increasing the house edge by not doubling -v- reduced variance by not staking more. I did read in one of the "bibles" (might have been BJA, but forgive me Ms S if it wasn't) that you really don't give up too much by not doubling soft hands. Does the EV impact of doubling on A4 v 4 at a neutral count justify doubling the bet and increasing the variance?

I use Hi-Lo on the odd occasions now when I play, and in the past I've queried the risk/benefit aspect of doubling against a dealer 2,3,4 with a max bet out at TC+4+. Isn't it a pain when you double 9 v 2 and then are dealt a 2 for a hand of 11? I did some sums a fair while back now that indicated that the total return difference from doubling 2,3,4 added together was less than that of doubling against a dealer 6. The counter arguement would be along the lines that if you're adequately capitalised, this isn't a consideration. I ususally have a bunch of electronic rotten fruit pitched in my direction by the "eek out every last fragment of EV" brigade.



As an afterthought, I should have added that I only play 6 Deck, S17 games (the de-facto standard in the UK).

[NotEnoughHeat]

Correct me if I'm wrong, but this almost sounds like a reference to risk-adverse indices, which only consider situations where you put more money on the table such as splits and doubles. In RA indices it's not that you don't make the index play at all, but you wait till a higher count to make the play. The reasoning behind this is that you still capture most of the EV of the index in general but you leave out the situations that have marginal EV with the same added risk.

While it is true that your calculations will already consider the risk of all your splits and doubles if you use EV-maximizing indices (the standard ones we usually use), if you are playing with just a little more risk than you are comfortable with, using RA indices for that particular spread/ramp may be the way to go.

Not making the correct play is what increases your risks (Decreased advantage). Making the correct play decrease you risk. Whether or not you win or loose is a matter of fluctuation.

If by correct plays you mean the EV-maximizing plays, that is not necessarily correct. I don't think the negative correlation between risk and advantage is as strong as you think it is. Having more risk doesn't automatically mean less advantage (nor does having more of an advantage mean less risk). There can be, and are, lots of plays and moves inside and outside of blackjack which have more risk (volatility) and a higher advantage. For example, you place a max bet out and as the hand is dealt out the count plummets to +2.5 TC (hi-lo). The dealer has an ace, and you decide to take even money (insurance) on your blackjack. This move is not the EV-maximing 'correct play'. However, I think it is very strange to say that this move has more risk than declining the even money.

While it is true that fooling around and just plain ignoring the count will increase your risk, holding off till later to play (rather than completely abandoning) an index can give you a small hit to EV for a decent decrease to variance. Again, all this assumes that we are talking about using RA indices, not just choosing to straight up ignore and never do the index.

[mushin]

Correct me if I'm wrong, but this almost sounds like a reference to risk-adverse indices, which only consider situations where you put more money on the table such as splits and doubles. In RA indices it's not that you don't make the index play at all, but you wait till a higher count to make the play. The reasoning behind this is that you still capture most of the EV of the index in general but you leave out the situations that have marginal EV with the same added risk.

Yes this makes perfect sense to me. In fact... there are time where if the count is rising and it is borderline going into the positive I will wait for it to get a little more positive before I raise my bet. This can also help to compensate for any potential error in TC conversions if you are just a bit off on deck estimation.

If by correct plays you mean the EV-maximizing plays, that is not necessarily correct. I don't think the negative correlation between risk and advantage is as strong as you think it is. Having more risk doesn't automatically mean less advantage (nor does having more of an advantage mean less risk). There can be, and are, lots of plays and moves outside of blackjack which have more risk (volatility) and a higher advantage.

Ok I see what you mean... Im talking more in terms of pure House Edge advantage.

While it is true that fooling around and just plain ignoring the count will increase your risk, holding off till later to play (rather than completely abandoning) an index can give you a small hit to EV for a decent decrease to variance. Again, all this assumes that we are talking about using RA indices, not just choosing to straight up ignore and never do the index.

I have no contention with holding off a bit to ensure a truly good count before pulling the trigger. Again only for indices.

[NotEnoughHeat]

Ok I see what you mean... Im talking more in terms of pure House Edge advantage.

I'm glad I have been helpful. It wasn't too long ago that I joined this forum and it's nice to know I'm making worthwhile contributions.

I'm not really sure what you mean by pure house edge advantage. The house edge is just the players advantage in reverse. Do you mind explaining this point a bit more?