Score

**moses** · 09-20-2016, 09:03 AM

Delete

**Phoebe** · 09-20-2016, 09:23 AM

I think SCORE in CVData is what you name cSCORE. It uses your bet ramp. Not optimal ramp.
For me, I look at winrate and RoR.

**BJGenius007** · 09-20-2016, 09:39 AM

Originally Posted by moses

I've heard it said that SCORE is the most important barometer when evaluating Sims on CV Data. My understanding is cSCORE is also vital but only available on CVCX.

I've also read that n0, Winrate, and other factors are worthy of consideration. How much weight do you give to each one? For instance, SCORE 80%, Winrate 10%, n0 10%.

My understanding is PE is virtually useless, and BC and the EOR are not really all that important. This is a pisser since a great deal of Norm's products are related to BC and PE.

Please clarify in order of importance. I'd like to see the box SCORE

SCORE is a fancy way to say ev with all parameters locked. Without SCORE, we used to say some particular game has ev of $20 if spread $160 to $10, assuming dealer is fast enough to deal 100 hands per hour. With SCORE, all parameters are locked. It has to be 100 hands per hour, bankroll $10000, player spread using one Kelly bet. So SCORE can't be 20%, must be 20, meaning the player hourly earning is $20.

Bankroll $10000 with one Kelly betting means you have to find $9.07 minimal table, but $10 minimal table is close enough.

N0 = 1000000/SCORE

For example, if SCORE is 40, then N0 must be 25000. Players want to find the game with highest SCORE (or lowest N0).

**Phoebe** · 09-20-2016, 10:23 AM

Originally Posted by moses

I'm not seeing an RandR category in CV Data.

CVData don't have BR. You have to call CVCX from summary.

**BJGenius007** · 09-20-2016, 10:33 AM

Originally Posted by moses

Ok. So SCORE and n0 are sort of a reverse of one in the same.

Its fair to say SCORE will be increased with your mind or your money. What is considered a worthwhile increase to jump from a Level 1 to Level 2 count? Or better yet the main topic of endless discussion, hi lo to hi opt ii?

SCORE varies when penetration, counting system changes, even at the same table. In your example, you will have two different SCORE using Hi-Lo and Hi-Opt II. Hi-Opt II should have a higher SCORE at the same table.

**ZenMaster_Flash** · 09-20-2016, 10:49 AM

Phillipe B. ...

Please do not suggest that moses make a purchase.

The results could be close to catastrophic for him.

L.O.L.

**Phoebe** · 09-20-2016, 12:06 PM

Anyway CVCX is free when called from CVData. Of course with limited functions.

**Phoebe** · 09-20-2016, 12:29 PM

SCORE is different because bet ramp is optimized and spread is different.

**Three** · 09-20-2016, 12:33 PM

Originally Posted by moses

I assume the lower the better.

13.5% Risk OF Ruin gives optimal BR growth. Any higher RoR is costing you. Now 13.5% RoR is too high for most AP's so they opt for betting proportional to 13.5% RoR kelly optimal bets at lower bets to decrease risk. This also costs you. Playing the optimal 13.5% you should constantly be adjusting bets to your new current BR. RoR changes every time your BR changes. You adjust bets to maintain the 13.5% RoR. If you do that your are playing to almost 0 chance of busting out.

**Phoebe** · 09-20-2016, 01:00 PM

Originally Posted by moses

That doesn't make sense because I transferred the information. Nonetheless, i assume the SCORE on CV Data is the better choice for comparison purpose.

You transferred the information but by default, CVCX give you his ramp witch is optimized.
SCORE on CVData compare results with your bet ramp and your spread (inside simulation).
Both are useful.

**BJGenius007** · 09-20-2016, 01:02 PM

Originally Posted by moses

The RofR result of the top 3 WINRATE FINISHERS came in at 2.5%, 4.1%, and 5.3%.

SCORE locks on one full Kelly bet means Risk Of Risk is locked at 13.5%. That is what T3 tried to tell you.

**Three** · 09-20-2016, 01:19 PM

SCORE uses a standard setting for BR (10K), spread (1-10 I think), RoR (13.5%), game speed (100 rounds/hour), etc. On the other hand c-SCORE can use anything and gives you (EV/SD)^2 for whatever parameters you set. Don invented SCORE or at least popularized it as he will give credit to at least 1 predecessor with a very similar idea. Don's wish was to have SCORE refer to his concept and c-SCORE (custom SCORE) refer to anything that didn't use the standard inputs he set. Unfortunately many refused to respect Don's request and the term SCORE is often misused which causes the confusion that Don wanted to avoid.

**BJGenius007** · 09-20-2016, 01:33 PM

Originally Posted by Tthree

SCORE uses spread (1-10 I think).

SCORE assumes one full Kelly spread. Meaning if player is using Hi-Lo on 6D S17 shoe game, his spread will be something like this:

$40 when player advantage is 0.5%
$80 when player advantage is 1.0%
$120 when player advantage is 1.5%
$160 when player advantage is 2.0%
$200 when player advantage is 2.5%

Based on how many index plays are used, it is likely that the above conditions occurs at TC +1, +2, +3, +4 and +5.

For players using level 2 counting system, TC could be at +2, +3.5, +5, +6.6, +7 for him to bet $40, $80, $120, $160, $200. Note that level 2 TC+7 is equivalent to level 1 TC +3.5. So level 2 SCORE is higher than level 1 SCORE as his max bet out faster at the same level of risk.