SD survey

[Aisthesis]

I honestly don't know the exact formula, but it's a measure of variance from an average. The following is presumably off, but should be moderately close:

Say you have an average win-rate of 10 ptbb over 20k hands. Ok, that's 200 groups of 100 hands each. Now, you were down 50 ptbb on some of those groups of 100 hands, and you were up 100 ptbb on some, and you were close to break even on some.

If you take each of those groups and measure it by the deviation from your normal win-rate, then, you'll get -60 (ptbb) when you were down 50 ptbb, and you'll get +90 when you were up 100, and you'll get -10 for those groups of 100 where you broke even.

Now, if you take the AVERAGE up or down for each group of 100 over your 20k hands, measured according to your "norm" of 10 ptbb, then you get something like standard deviation.

I think the actual formula is a little more complicated, but I'm not a statistician. Anyhow, the above gives you a basic idea of what it means, and that number will be fairly close to SD anyway.

[Silent A]

<u>Everything you ever wanted to know about Standard Deviation, but were afraid to ask ...</u>

[Troll_Inc]

[ QUOTE ]
I honestly don't know the exact formula, but it's a measure of variance from an average. The following is presumably off, but should be moderately close:


[/ QUOTE ]

I'm pretty sure that I asked this question on the Pokertracker forum and got the answer that pokertracker calculates it per hand, not groups of hands.

It's pretty simple and if anyone wants to really understand the winrate and bankroll stuff, you should understand the basic concept.

Say you sit down at a poker table and play 10 hands of PLO100 with $0.50/$1.00 blinds.

Your net win/loss after each hand is
from BB -$1.00
from SB -$1.00
Button $0
Cutoff $25
-$1
$0
$-5
$5
$0
UTG $-1

OK, so the first thing you do is calculate the average net win/loss of the ten hands from this sample.

I would recommend doing this in a spreadsheet like Excel or Googlespreadsheets because it will make calculating your variance and standard deviation easier. Put these values down Column A.

I get a total net win of $21 and therefore when you divide by 10 your net win per hand. $2.1/hand
In Excel, in cell A11 you would type the formula =AVERAGE(A1:A10)

In column B put 2.1 (your average net win per hand) down all ten rows next to each hand value.

In column C, subtract the average net win per hand from the net win of each hand, which is simply A-B.

In column D, square the value of C. In Excel, you can type:
=C*C
In cell D11, sum all the values from D1-D10.

From the formula for variance (see link SilentA provided), you now just divide by n-1, which in this case is 9.

Now that is the VARIANCE of your sample, which is just a measure of how far your individual hand results in your sample vary from the average result of your sample. Standard deviation is just the square root of variance (since the values were squared in the first place).

So in excel, take the square root of the value you calculated for variance and you get the standard deviation. (=sqrt)

All this can easily be done on a piece of paper and a regular calculator.

Statistics is such that it's a use it or lose it skill. If you know it, you'll forget it if you don't use it. And to learn it, it is best to do some calculations yourself. I'll wait to post the answer in case anyone cares to do this themselves.


*******************

For ease of cut and paste, here is column A without the dollar signs.

-1
-1
0
25
-1
0
-5
5
0
-1
2.1

[pete fabrizio]

Why you would want to do this in excel instead of pokertracker, I don't know. But if you did, it would be much easier to use the STDEV() function.

[Aisthesis]

Well, sure, but I think Troll was explaining what it actually is. For just getting the number, the easiest way of all is just to let pt calculate it for you.

[Troll_Inc]

[ QUOTE ]
Well, sure, but I think Troll was explaining what it actually is.

[/ QUOTE ]

[pete fabrizio]

[ QUOTE ]
[ QUOTE ]
Well, sure, but I think Troll was explaining what it actually is.

[/ QUOTE ]

[/ QUOTE ]

Sorry, all those numbers were hurting my eyes.