Understanding SCORE Concept

[MercySakesAlive]

In Modern BJ it says "a SCORE of 20 means a long-term win rate twice as great as a SCORE of 10."

Below, Game A has 1/2 the SCORE of Game B but different ROR's and win rates.

Does Game B really have a long-term win rate twice as great as Game A? Or does the statement in Modern BJ only hold true when two games have the same ROR and win rate?

A. 6 deck, DAS RSA/ 1 hand ... pen: 1.0 ... win rate/hour: $30.22 ... ROR: 3.0% ...SCORE: 52.64
B. 6 deck, DAS RSA Ls/ 2 hand ... pen: 1.0 ... win rate/hour: $47.47 ... ROR: 1.2% ...SCORE: 105.28

[Three]

Below, Game A has 1/2 the SCORE of Game B but different ROR's and win rates.

SCORE would mean they have to have the same RoR. c-SCORE would allow RoR to be different. You will get confused about what you read if you don't know the difference. Those that post without using the different nomenclature properly will confuse others with their ambiguous posts. SCORE has a set of standards for BR, RoR etc, while c-SCORE doesn't. The use of SCORE implies that all the SCORE standards were used. Too many people don't get this right when using the terms. I mess it up sometimes and I try hard not to.

Does Game B really have a long-term win rate twice as great as Game A? Or does the statement in Modern BJ only hold true when two games have the same ROR and win rate?

A. 6 deck, DAS RSA/ 1 hand ... pen: 1.0 ... win rate/hour: $30.22 ... ROR: 3.0% ...SCORE ... 52.64
B. 6 deck, DAS RSA Ls/ 2 hand ... pen: 1.0 ... win rate/hour: $47.47 ... ROR: 1.2% ...SCORE ... 105.28

The statement referred to the specificd SCORE not the c-SCORE you quoted in your examples. Neither one of what you listed is the SCORE in A or B above, but it is the c-SCORE. I think Norm tries to use the terms correctly.

[MercySakesAlive]

Can it be said that a c-SCORE of 20 means a long-term win rate twice as great as a c-SCORE of 10 then?

[Freightman]

Can it be said that a c-SCORE of 20 means a long-term win rate twice as great as a c-SCORE of 10 then?

Fair to say

[Three]

You see SCORE keeps most things the same so what can be changed has a proportional effect on win rate. SCORE standards:
1) $10K BR
2) 13.5% ROR
3) penetration
a) 31-32 cards for SD
b) abut 70 cards for DD
c) about 5 decks for 6 deck shoe
d) about 6.5 decks for 8 deck shoe
4) Spread
a) 1-3 for SD
b) 1-6 for DD
c) 1-12 for 6 deck
d) 1-8 for backcounting an 8 deck game

The "S" in SCORE stands for standardized, Standardized Comparison Of Risk and Expectation. It was designed to compare game quality. When people used SCORE to compare approaches to playing a game they use c-SCORE (custom SCORE). As you can see with SCORE standards, if you stick to the standards you will get relationships that don't exist when you use c-SCORE because c-SCORE every standard can differ in the comparison.

[Three]

Can it be said that a c-SCORE of 20 means a long-term win rate twice as great as a c-SCORE of 10 then?

No. That would only hold true if RoR and BR was the same.

[MercySakesAlive]

Just finished reading Chapter 9 in BJA3. It lays out the history of SCORE and c-SCORE, as you probably know. It seems like c-SCORE is not quite as useful as SCORE. SCORE tells you a lot more about two games where c-SCORE is a little less useful I think. It'd be great to have a number that answers the question "when these two games are compared which game will make more money in the long run?" AND allow for different betting schedules, count strategies, ROR, and bankroll.

[Three]

It'd be great to have a number that answers the question "when these two games are compared which game will make more money in the long run?" AND allow for different betting schedules, count strategies, ROR, and bankroll.

I try to maximize several stats as a group when I fine-tune a ramp. I play more games than just BJ and use approaches with varying degrees of complexity that each allow different abilities to fine tune things.

The stats I worry about in no particular order:

1) Maximize c-SCORE (same as minimizing n0). {accomplished by balancing EV gain with increased variance}
2) Maximize monetary EV (the one I place the least emphasis on). {accomplished from betting more when you have an advantage and less when you don't}
3) Maximize W/L% (IBA). {accomplished by reducing average bet with minimal impact to EV. This is done best by lowering negative EV or high frequency low advantage bets. Often frequency is the biggest consideration when trying to increase IBA. Your most frequent advantage bets have a very small advantage so they add little to EV}.
4) Maximize CE {accomplished by reducing variance}
5) Minimize BR for 1% RoR

I try to keep the general tide of all these things headed in the direction specified. Trading some monetary EV or a little c-SCORE to make the other stats better is acceptable. CE and 1% RoR BR take care of themselves if you take care of the first 3 on the list. c-SCORE will try to have a balance across your entire ramp. IBA will try to emphasis the top end of your ramp while decreasing higher frequency advantage bets that add little to EV as a group. EV emphasizes betting more into higher advantages and not at all when at a disadvantage.

Now defining the best ramp is kind of personal (what you seem to be asking about). Some feel EV is all that matters. Others feel c-SCORE is all that matters. Some place the emphasis on CE. I like to get more predictable gains which involve considering all of the above. By including 1% RoR BR as a stat you get a perspective on how everything affects swings. Your BR must be big enough to withstand the swings that are expected. The larger the swings created by your ramp and approach the larger the BR required for the same RoR. If I want to get real sensitive about measuring change I will use 0.01% RoR BR.

[DSchles]

You see SCORE keeps most things the same so what can be changed has a proportional effect on win rate. SCORE standards:
1) $10K BR
2) 13.5% ROR
3) penetration
a) 31-32 cards for SD
b) abut 70 cards for DD
c) about 5 decks for 6 deck shoe
d) about 6.5 decks for 8 deck shoe
4) Spread
a) 1-3 for SD
b) 1-6 for DD
c) 1-12 for 6 deck
d) 1-8 for backcounting an 8 deck game

The "S" in SCORE stands for standardized, Standardized Comparison Of Risk and Expectation. It was designed to compare game quality. When people used SCORE to compare approaches to playing a game they use c-SCORE (custom SCORE). As you can see with SCORE standards, if you stick to the standards you will get relationships that don't exist when you use c-SCORE because c-SCORE every standard can differ in the comparison.

SCORE doesn't stipulate a single level of penetration. Rather, the chapter 10 charts allow for four different levels of pen for each set of rules.

Don

[Three]

SCORE doesn't stipulate a single level of penetration. Rather, the chapter 10 charts allow for four different levels of pen for each set of rules.

Don

Thanks Don. I knew you would be quick to correct on this subject.

[MercySakesAlive]

Thanks for taking the time to write all that great information, Three. That is really helpful. I think I need to consider these other values.

[MercySakesAlive]

Three, I spent some time with the stuff you wrote. Thanks.

Why is it that when I raise some bets the win rate goes up but the win/loss% goes down?

Anyone? Thank you.

[Three]

Why is it that when I raise some bets the win rate goes up but the win/loss% goes down?

W/L% is IBA (initial bet advantage). It divides EV by your average bet. If the average bet goes up too much compared to EV from the change the W/L% will go down. Bet frequency has a lot to do with how things are affected. Raising high frequency low advantage bets will not change EV much but will increase average bet a lot, relatively speaking.