Standard Deviation Question for Norm/QFIT

[Pelerus]

According to CVCX, the standard deviation for my game (S17, DAS, 6.4/8 pen, spread $10-$120 at +5, $0 at <= -1) is 35.796 units per hour, and 4.185 units per hand. My question is how CVCX derives the standard deviation per hour given the standard deviation per hand. When I use the method of multiplying the SD per hand by the square root of the number of hands per hour, I get 4.185 * sqrt(100) = 41.85. This number, although reasonably close to the CVCX estimation, is off by enough to warrant the question.

Thank you for your input!

 

[Sonny]

When I use the method of multiplying the SD per hand by the square root of the number of hands per hour, I get 4.185 * sqrt(100) = 41.85. This number, although reasonably close to the CVCX estimation, is off by enough to warrant the question.

That's because you are not playing 100 hands per hour, you are only watching 100 hands. You are probably only playing about 73 hands per hour.

Make sure you are looking at the “SD per Hour” section, not the “SD per 100“ section. If you are Wonging out then the two will not be the same because it will take you longer than one hour to play 100 hands. The numbers above look like you are playing about 73.16% of the hands, which would be consistent with the strategy you described.

If that is not the case, check the Setup page. Look at the bottom left section where it says "Define Data Collection." If the “Hands per Hour” is not set to 100 then it may explain the discrepancy.

-Sonny-

 

[zengrifter]

I don't think he would be playing 73% of the hands if he is avoiding counts of -1 or less. zg

 

[Sonny]

I don't think he would be playing 73% of the hands if he is avoiding counts of -1 or less. zg

Probably not, but if he is leaving at -2 (playing through a -1 but nothing less)then it would be about right. Hopefully we can get some more info from him.

-Sonny-

 

[Pelerus]

Followup

If the information Sonny provides is correct, then how would I go about figuring my standard deviation for any length of time, such as, for example, 20 hours of play?

Since CVCX gives the SD per hour as 35.796, could I divide by sqrt(100) to get 35.796 / sqrt(100) = 3.58? Could I then multiply the 3.58 by the square root of any number of hands to get the standard deviation for that number?

 

[zengrifter]

I'm not fully conversant with CVSim, but based on this thread I think you can simply use the 100/hr factor and multiply that times #hrs, then use the SqRt of that times the CVSim 1hand StDv. zg

 

[Sonny]

Just substitute the fact that your're only playing 73 hands per hour into the formula (instead of the usual 100). You end up with this:

SD per Hand * sqrt(73) = Hourly SD

= 4.185 * 8.544 = 35.76 Units per Hour

The result above is slightly different than the one from the software because I rounded the hands/hour and the sqrt result, but it will give you the general idea of how to use the formula.

-Sonny-

 

[Pelerus]

Methods

I believe that both methods arrive at the same conclusion. For example,
for two hours of play:
method 1: 3.58 * sqrt(200) = 50.628 units.
method 2: 4.185 * sqrt(73*2) = 50.567 units.
for 18 hours of play:
method 1: 3.58 * sqrt(1800) = 151.885 units.
method 2: 4.185 * sqrt(73*18) = 151.702 units.

 

[Sonny]

Yeah, both methods will give you the same answer. The second method just uses the SD/hand right from CVCX without having to convert it. Just make sure to use the SD/100 numbers for your ROR calculations, not the SD/hr.

-Sonny-