Does EV ever Trump Game Quality?

blackjack avenger

Well-Known Member
#5
Going Till Broke

SleightOfHand said:
EV
% advantage
Not sure if these measure quality of game, perhaps I needed to add "in relation to risk"

Chasing raw EV is not necessarily the best thing, if it were the case one would bet all on any advantage.

:joker::whip:
good cards
 

Sucker

Well-Known Member
#6
Does the bear sh*t in the woods?

EV is the DEFINITION of game quality! It matters not which tools you use to get there; the higher the EV, the better the quality of the game. And the ONLY time you need to worry about your EV "in relation to risk" is in the case of long shot bets, which simply do not occur in standard BJ bets.
 

Sucker

Well-Known Member
#7
blackjack avenger said:
Chasing raw EV is not necessarily the best thing, if it were the case one would bet all on any advantage.
Chasing raw EV is not just the best thing - it's the ONLY thing! And of COURSE you would not bet all on any advantage. Use your "Search BlackjackInfo" feature & read up on as much as you can about the Kelly Criteria. This is VERY important stuff for a professional gambler to know.
 

blackjack avenger

Well-Known Member
#8
Kelly & I Go Way Back

Sucker
I think you need to rethink your thoughts on EV is everything.

Simple ?
Do you use EV or RA indices?

It's a math fact to max. EV you bet all on any advantage, but we don't do this because of risk. So EV is not the only thing.

Kelly is not about max EV
 

aslan

Well-Known Member
#9
Sucker said:
Chasing raw EV is not just the best thing - it's the ONLY thing! And of COURSE you would not bet all on any advantage. Use your "Search BlackjackInfo" feature & read up on as much as you can about the Kelly Criteria. This is VERY important stuff for a professional gambler to know.
The only thing I bet it all on are those rare "two headed" coin props. :joker:
 

Sucker

Well-Known Member
#10
blackjack avenger said:
Chasing raw EV is not necessarily the best thing
If EV is not the very first thing you consider when determining proper bet size, then please educate me. How do YOU determine how much to bet?

blackjack avenger said:
Kelly and I go way back
Your question made it appear otherwise. So I wholeheartedly apologize for sounding pedantic. :eek:

blackjack avenger said:
Simple ?
Do you use EV or RA indices?
In other than certain circumstances, using RA indices is a chicken's way to play; and it "leaves money on the table".
 

aslan

Well-Known Member
#11
In my life, there is no such thing as an unreplenishible bankroll, so you might see me betting it all with 20% the best of it. In truth, I've bet it all with less than 20% many a time in games of skill. But given a truly unreplenishible bankroll, you would not see me betting it all with even 99% the best of it. (Did anyone say, "Locksmith"?)

Sorry, Sucker, I guess I'm the sucker; I often leave some on the table. :(:cry:

"He who fights and runs away, lives to right another day."
 

Southpaw

Well-Known Member
#12
Some of the assertions in this thread are very misinformed.

Condolences to the new AP's that may, too, become misinformed by reading this thread.

Just so that everyone is on the same page, let's reassure ourselves of what "EV" means. When determining bet sizes, the IBA is typically used, but if you are going to be citing the overall percentage return that a game is "expected" to yield, then the TBA is usually cited.

The only time I could ever imagine tossing SCORE, NO, and DI to the wind, erring to the preference of EV--a statistic that, for all intents and purposes, should be disregarded when judging game quality--is if one had a bankroll of infinity dollars.

But before we enter the world of Philosophy and all of its nonsensical, imaginary cases, let's get back to reality. Game desirability must be risk-adjusted, because there is no such thing as a bankroll of infinity. Not even the very most replenishable bankroll.

Assume Andrew Provant (AP) desires to play at an RoR of 5%. If he plays to maximize his EV, he will be placing his max-bet whenever he sees the advantage. However, to keep his RoR at 5%, he will have to use a tiny unit size, and therefore will be making less money than he could be. On the other hand, if he played to maximize his SCORE--by betting optimally and using RA indices--rather than EV, he would be able to use larger units, thus making more money, all while still at an RoR of 5%.

Well, now some of the more bold may say, "Ah, that's yellow. Maximize the EV and deal with the higher RoR, you chicken!" So, let's say AP decides to increases his unit size, while playing the EV maximizing strategy to the point that he now has the same return that would have been provided by the SCORE-maximizing strategy at 5% RoR. This will unequivocally raise AP's RoR; for simplicity, let's say that doing this raises AP's RoR to 13.53%. AP may now be very pleased that he now has now fudged the EV maximizing strategy in a way that it has the same return as the SCORE-maximizing strategy. However, if he were to play the SCORE-maximizing strategy at a unit-size such that RoR would become 13.53%, then the SCORE-maximizing strategy would be again victorious.

While the case of AP dealt with strategies, the same can be said about games. Let's assume that we have a DD game and an 8-deck, each of which we will play a strategy such that the EV is 2%. The EV is the same, but which game is better? Before we answer this, recall that the hourly Standard Deviation tends to be way lower when playing DD games. Therefore, for any given RoR, the return will be much better on the DD game than for the 8-deck game.

Conclusions:

1. Games cannot be described on the basis of EV alone.
2. Increased Variance is synonymous with lower return, because of the variable unit-size concept. They are one of the same, rather than two distinct entities of differing importance.

And on the subject of RA indices. I am not quite sure why exactly they are the butt of all jokes around here when they are statistically superior to EV-maximizing ones. They may not be a lot better, but they are better. I have come to the conclusion that they must be misunderstood.

The EV-maximizing indices do just what their name implies. Risk adverse indices maximize SCORE; they are not "risk-adverse" or "risk-scared," they are risk-smart.

Think of it this way. It is well-known that the advantage gained for making a departure when the index is close is absolutely minimal. So, assume we have just placed our max-bet and have been dealt a hand that we may want to double-down. We calculate the TC and find it to be just above the index. Granted the EV gained for doubling here is extremely small (as opposed to not doubling), why would you be willing to slide out another max bet here to do so? After all, you would not slide out a max-bet for your max-bet when the TC is only marginally good; if you are like most of us, you'd wait for an edge of 2-3%.

As putting out a second large bet for such a marginal advantage will increase our RoR, we will then have to decrease our unit-size, thus leading to less profit. It all comes down to this question:

"Am I losing more EV by not doing the risky play or by having to use smaller units in order to maintain a constant RoR?"

The RA index is found at the first TC where you'd be losing more EV by not making the risky play.

SP
 

Sucker

Well-Known Member
#13
I was coming to the same conclusion - that we're not on the same page together. Therefore I'll revise my original statement:

Does EV ever trump game quality?
Yes - when deciding how much to bet. Suppose you're playing a game with a 2% overall advantage, and the count shoots up to the point where you have a 10% advantage on the next hand. Obviously, you're not going to play the hand as though your edge was 2%.
 
#14
Well Said

Southpaw said:
Some of the assertions in this thread are very misinformed.

Condolences to the new AP's that may, too, become misinformed by reading this thread.

Just so that everyone is on the same page, let's reassure ourselves of what "EV" means. When determining bet sizes, the IBA is typically used, but if you are going to be citing the overall percentage return that a game is "expected" to yield, then the TBA is usually cited.

The only time I could ever imagine tossing SCORE, NO, and DI to the wind, erring to the preference of EV--a statistic that, for all intents and purposes, should be disregarded when judging game quality--is if one had a bankroll of infinity dollars.

But before we enter the world of Philosophy and all of its nonsensical, imaginary cases, let's get back to reality. Game desirability must be risk-adjusted, because there is no such thing as a bankroll of infinity. Not even the very most replenishable bankroll.

Assume Andrew Provant (AP) desires to play at an RoR of 5%. If he plays to maximize his EV, he will be placing his max-bet whenever he sees the advantage. However, to keep his RoR at 5%, he will have to use a tiny unit size, and therefore will be making less money than he could be. On the other hand, if he played to maximize his SCORE, rather than EV, he would be able to use larger units, thus making more money, all while still at an RoR of 5%.

Well, now some of the more bold may say, "Ah, that's yellow. Maximize the EV and deal with the higher RoR, you chicken!" So, let's say AP decides to increases his unit size, while playing the EV maximizing strategy to the point that he now has the same return that would have been provided by the SCORE-maximizing strategy at 5% RoR. This will unequivocally raise AP's RoR; for simplicity, let's say that doing this raises AP's RoR to 13.53%. AP may now be very pleased that he now has now fudged the EV maximizing strategy in a way that it has the same return as the SCORE-maximizing strategy. However, if he were to play the SCORE-maximizing strategy at a unit-size such that RoR would become 13.53%, then the SCORE-maximizing strategy would be again victorious.

While the case of AP dealt with strategies, the same can be said about games. Let's assume that we have a DD game and an 8-deck, each of which we will play a strategy such that the EV is 2%. The EV is the same, but which game is better? Before we answer this, recall that the hourly Standard Deviation tends to be way lower when playing DD games. Therefore, for any given RoR, the return will be much better on the DD game than for the 8-deck game.

Conclusions:

1. Games cannot be described on the basis of EV alone.
2. Increased Variance is synonymous with lower return, because of the variable unit-size concept. They are one of the same, rather than two distinct entities of differing importance.

And on the subject of RA indices. I am not quite sure why exactly they are the butt of all jokes around here when they are statistically superior to EV-maximizing ones. They may not be a lot better, but they are better. I have come to the conclusion that they must be misunderstood.

The EV-maximizing indices do just what their name implies. Risk adverse indices maximize SCORE; they are not "risk-adverse" or "risk-scared," they are risk smart.

Think of it this way. It is well-known that the advantage gained for making a departure when the index is close is absolutely minimal. So, assume we have just placed our max-bet and have been dealt a hand that we may want to double-down. We calculate the TC and find it to be just above the index. Granted the EV gained for doubling here is probably well below 0.1% (as opposed to not doubling), why would you be willing to slide out another max bet here to do so? After all, you would not slide out a max-bet for your max-bet when you only had an advantage of 0.1%; if you are like most of us, you'd wait for an edge of 2-3%.

As putting out a second large bet for such a marginal advantage will increase our RoR, we will then have to decrease our unit-size, thus leading to less profit. It all comes down to this question:

"Am I losing more EV by not doing the risky play or by having to use smaller units in order to maintain a constant RoR?"

The RA index is found at the first TC where you'd be losing more EV by not making the risky play.

SP
I agree
 

psyduck

Well-Known Member
#16
Southpaw said:
Think of it this way. It is well-known that the advantage gained for making a departure when the index is close is absolutely minimal. So, assume we have just placed our max-bet and have been dealt a hand that we may want to double-down. We calculate the TC and find it to be just above the index. Granted the EV gained for doubling here is probably well below 0.1% (as opposed to not doubling), why would you be willing to slide out another max bet here to do so? After all, you would not slide out a max-bet for your max-bet when you only had an advantage of 0.1%; if you are like most of us, you'd wait for an edge of 2-3%.

SP
Are you saying if you have your minimum bet out, you will double under the low advantage conditions? If so, wouldn't your cumulative combined small bets be equivalent to having a large bet out?
 
#17
Sucker said:
...In other than certain circumstances, using RA indices is a chicken's way to play; and it "leaves money on the table".
Unless by "certain circumstances" you mean "having a finite bankroll," RA indices are part of Kelly betting.

Not putting out a table max bet at +0.5% advantage also leaves money on the table and some might call me a chicken for failing to do that. But if (for example) you only put out a $200 bet for a 2% IBA and a $250 bet for a 2.5% IBA, when you have your $200 bet down why the heck would you make a double play that requires you to put down $400 for a 2.5% IBA? That's just plain old overbetting.

I'd think the situations in which you wouldn't use RA indices are very limited, not the other way around.
 
#18
Southpaw said:
Some of the assertions in this thread are very misinformed.

Condolences to the new AP's that may, too, become misinformed by reading this thread.

Just so that everyone is on the same page, let's reassure ourselves of what "EV" means. When determining bet sizes, the IBA is typically used, but if you are going to be citing the overall percentage return that a game is "expected" to yield, then the TBA is usually cited.

The only time I could ever imagine tossing SCORE, NO, and DI to the wind, erring to the preference of EV--a statistic that, for all intents and purposes, should be disregarded when judging game quality--is if one had a bankroll of infinity dollars.

But before we enter the world of Philosophy and all of its nonsensical, imaginary cases, let's get back to reality. Game desirability must be risk-adjusted, because there is no such thing as a bankroll of infinity. Not even the very most replenishable bankroll.

Assume Andrew Provant (AP) desires to play at an RoR of 5%. If he plays to maximize his EV, he will be placing his max-bet whenever he sees the advantage. However, to keep his RoR at 5%, he will have to use a tiny unit size, and therefore will be making less money than he could be. On the other hand, if he played to maximize his SCORE, rather than EV, he would be able to use larger units, thus making more money, all while still at an RoR of 5%.

Well, now some of the more bold may say, "Ah, that's yellow. Maximize the EV and deal with the higher RoR, you chicken!" So, let's say AP decides to increases his unit size, while playing the EV maximizing strategy to the point that he now has the same return that would have been provided by the SCORE-maximizing strategy at 5% RoR. This will unequivocally raise AP's RoR; for simplicity, let's say that doing this raises AP's RoR to 13.53%. AP may now be very pleased that he now has now fudged the EV maximizing strategy in a way that it has the same return as the SCORE-maximizing strategy. However, if he were to play the SCORE-maximizing strategy at a unit-size such that RoR would become 13.53%, then the SCORE-maximizing strategy would be again victorious.

While the case of AP dealt with strategies, the same can be said about games. Let's assume that we have a DD game and an 8-deck, each of which we will play a strategy such that the EV is 2%. The EV is the same, but which game is better? Before we answer this, recall that the hourly Standard Deviation tends to be way lower when playing DD games. Therefore, for any given RoR, the return will be much better on the DD game than for the 8-deck game.

Conclusions:

1. Games cannot be described on the basis of EV alone.
2. Increased Variance is synonymous with lower return, because of the variable unit-size concept. They are one of the same, rather than two distinct entities of differing importance.

And on the subject of RA indices. I am not quite sure why exactly they are the butt of all jokes around here when they are statistically superior to EV-maximizing ones. They may not be a lot better, but they are better. I have come to the conclusion that they must be misunderstood.

The EV-maximizing indices do just what their name implies. Risk adverse indices maximize SCORE; they are not "risk-adverse" or "risk-scared," they are risk smart.

Think of it this way. It is well-known that the advantage gained for making a departure when the index is close is absolutely minimal. So, assume we have just placed our max-bet and have been dealt a hand that we may want to double-down. We calculate the TC and find it to be just above the index. Granted the EV gained for doubling here is probably well below 0.1% (as opposed to not doubling), why would you be willing to slide out another max bet here to do so? After all, you would not slide out a max-bet for your max-bet when you only had an advantage of 0.1%; if you are like most of us, you'd wait for an edge of 2-3%.

As putting out a second large bet for such a marginal advantage will increase our RoR, we will then have to decrease our unit-size, thus leading to less profit. It all comes down to this question:

"Am I losing more EV by not doing the risky play or by having to use smaller units in order to maintain a constant RoR?"

The RA index is found at the first TC where you'd be losing more EV by not making the risky play.

SP
Post of the month goes to the Spaw. zg
 

Southpaw

Well-Known Member
#19
psyduck said:
Are you saying if you have your minimum bet out, you will double under the low advantage conditions? If so, wouldn't your cumulative combined small bets be equivalent to having a large bet out?
Well, here's the deal with RA indices. They take into account the average bet size that you will have out at that particular TC. Face it, it is all but likely that the TC will have been so high as to call for your max-bet, only to have to dropped into the negatives by the time it it's your turn to play your hand (assuming a 6- or 8-deck game). So, the assumption that the TC will be close enough, for all intents and purposes, as to when you placed your initial wager is a pretty good one.

However, to address what I think you're getting at, in an ideal situation, one would have a separate RA index for every possible wager in your spread. This, however, is way too burdensome, and it is perfectly all right to correlate the RA index with the average bet that you would have at the specific TC of the RA departure. A consequence of this is that, for the most part, the only RA indices that will show significant difference from their EV-minimizing counterpart are the ones that are both (1) large and positive and (2) require the player to put up more chips to complete the play (i.e., double-downs, splits and insurance).

And to answer your question, assume the following situation. At the beginning of the round the TC was -1 (thus I placed a minimum bet), and I was dealt 10,10 v. 5. However, by the time it is my turn to play, the TC is higher than the EV-maximizing index, but lower than the RA index. I would probably still want to split here, despite being lower than the RA index, for I remember that the RA index assumes that I have a max-bet out. I am not exposing myself to significant risk here, just for putting up an extra minimum bet to attain the EV from splitting my tens.

The above logic has been discussed in Zg's interviews, I believe.

Best,

SP
 
Top