New Standard Deviation Stat

[brone]

Hi,

I don't think the SD/Hr and SD/100 stats that were added in beta 25 (in the Details - Details section of the Sessions Tab) are correct.
They're giving me 3-4x the amount that PT2 does.

I thought maybe that there was a problem combing the BB values when mutliple stakes are in the DB, but when I filtered to just ONE stake, the problem was still there.

-Brone

[kraada]

We'll look into this and get back to you.

[CyaET]

I think this stat is wrong too , I have a standard deviation of 200BB/100 which seems impossible

[kraada]

We are looking into this, though prima facie that value isn't impossible. You'd just have to have very swingy sessions and play at deep tables, I'd imagine. Though, I agree that sounds very large, and hopefully our further investigations will clear things up.

[CyaET]

I imported a stat in pt3 : standard deviation and it gives extacly 4x less than your standard deviation stats

[kraada]

Was that stat from the repository? If so, could you tell me which one it is?

[iv8an]

For me to is 3-4x higher.

[kraada]

Could someone please confirm for me where they got this other stat so that we can get the two formulas compared? That is the only way to figure out what is causing the discrepancy.

[iv8an]

PT2: Ring game player staticstics => Session notes => More Detail => there is Standard Deviation/100 Hands
PT3: Texas Holdem => Cash game => General => Player Summary => Std Dev BB/100 (when you "Configure This Report")
In PT3 numbers are 100-300 BB/100 and in PT2 25-80 BB/100 for same sessions (hands). Std Deviation 200BB/100 is just too much (for 10k hands).

[WhiteRider]

I've just been doing a bit of investigation of this and initially it seems that PT3 is correct.

PT3 is calculating the standard deviation of the BB per 100 hands of each session.
So for each session we calculate the BB/100 hands, then we calculate the standard deviation of those values.
I have tested this by copying a set of session BB/100 values to Excel and using the STDEV function and the values match PT3.
I don't know how PT2 was calculating the standard deviation, but I'm going to try to find out.

Is this what you would expect Standard Deviation/100 Hands to be calculating?
It seems reasonable to me..