Normal or anormal variance

[Alexpoker]

Here's my 3 last all in in MTT (buy in 5 or 10 dollars).

1.Preflop
Hero : AK (90%)
Vilain : AQ (10%)

2. Flop
Hero : KK (90%)
Vilain : Q8 (10%)
flop : QT9

3. Flop
A [img]/images/graemlins/spade.gif[/img] K [img]/images/graemlins/spade.gif[/img] (90%)
AT (10%)
flop : 3 [img]/images/graemlins/spade.gif[/img]5 [img]/images/graemlins/spade.gif[/img]9 [img]/images/graemlins/diamond.gif[/img]

In each case, I wasn't "short stack".
I lost the 3.

Is it normal variance (that is to say lots of player has know this situation) or anormal variance ??

thanks

[PairTheBoard]

Suppose you were spinning a wheel where you had 90% chance of winning and 10% chance of losing. You should average 0.9 wins per spin. For 1 spin the variance for your number of wins is (.9)(.1)=0.09. The standard deviation for your number of wins is Sqrt(0.09)=0.3.

Now, suppose you spin the wheel and lose. That's 0 wins when you had 0.9 expected wins. It's 3 standard deviations below your expected wins. And if you were using this 1 spin to estimate your variance via the sample variance you would compute a sample variance of (0-.9)^2 = .81 which is 9 times greater than the actual variance.

So if you Spin the Wheel 1 time and lose is that abnormal variance? Because you lost the spin?

PairTheBoard

[Kerth]

[ QUOTE ]

1.Preflop
Hero : AK (90%)
Vilain : AQ (10%)

2. Flop
Hero : KK (90%)
Vilain : Q8 (10%)
flop : QT9



[/ QUOTE ]

What are the percentages supposed to represent?

[T50_Omaha8]

Assuming your percentages are correct, the probability you lose all three is (.1)^3 = 1/1000.
The probability you lose 2 of three is 3*.9*(.1)^2 = 27/1000
The probability you win 2 of three is 3*(.9)^2*.1 = 243/1000
The probability you win all three is (.9)^3 = 729/1000

This may seem like an absurd longshot, but every time you have a string of 1000 all-ins as a 9-1 favorite, you can expect to lose three in a row. So, to be honest, nothing is 'normal' or 'abnormal' in retrospect. If this was the first time you've ever played poker, then it would surely be considered improbable, but if you've got time under your belt, you can expect something like this to happen every once in a while. Another way of thinking of this is that if you go 1000 all ins favored 9-1 and never lose three in a row, you got 'lucky' (but would you call someone lucky if their Win/Loss string looks like LLWLLWLLWLLWLLWLLWLLWLL...?).

Just rest assured that after 1000 of these 9-1 favored AIs, it's very very likely you'll be between 88% and 92% overall.

[DWarrior]

[ QUOTE ]
1.Preflop
Hero : AK (75%)
Vilain : AQ (25%)

2. Flop
Hero : KK (81.6%)
Vilain : Q8 (18.4%)
flop : QT9

[/ QUOTE ]

http://www.pokerstove.com/ son

[Alexpoker]

Thanks for the real probability.

In fact, I try to evaluate the luck or bad luck I have during MTT in order to see if I play Ok or not (i'm a beginner so I don't know wether I'm a good or bad player). So, I made a Excel file with all my "all in" and the probability that I win. Then I compute my "variance".

Does someone else do this ??

[DWarrior]

I'm a cash player, but I hear SNG people advise to play 500 games then you'll have a decent estimate of your mean and variance.

You should also check out the SNG section, 2+2 pretty much reduced SNGs to a science.