Probability of a (bad) run to happen

[SongheJe]

Hello all.

Premise: I don't want to complain on how unlucky I am, I want simply to give all the meaning I can to some numbers.
I just posted this in BBV and gotta lot of insults on how a sucker tilter fish moron loser etc I am.
Don't tell me about this. I just need an answer to a specific and detailed question. Please read carefully, I try to be as clear as I can (excuse me for my english but it's pretty scholastic). If something is not clear please tell me.
So, the point is:

In the last 153 sit & goes I played I went allin 371 times in 1v1 scenario with showdown.

My EV from pots was 627,813 chips. I obtain this number using a program that I developed and called 'luckometer' that sums all the shares of the pots in preflop 1v1 allins.
Example: I go allin AA vs KK on a 5000 chips (total after allins) pot, my share is ~80% of 5,000. Luckometer calculate the odds of a hand to win vs another enumerating all possible boards.
In this bad run, I had a result of 502,848 chips effectively won. It means I lost 124,965 chips on 'bad luck'.
The only data I want to know is
What are the odds it can happen ?
So, what are the odds that going 371 times allin I have a -24.85% in chips won from expected value ?
I appreciate if someone can give me this answer.
More, if someone could give me the 'how' to calculate this would be great. Also having some reference (where to search for) should be great.

Thank you very much.
SongheJe

[Troll_Inc]

I think you are going to need the result of each hand to calculate what you want to calculate.

If you get me that information, then I will tell you by showing with real data.

[SongheJe]

Hello Troll_Inc.
Thanks for your answer.
Result for each hand is a text file long to paste here.
I attach a link.
http://www.sngoev.com/processed.log
Thanks again.
SJ

[sixhigh]

Don't expect any sane responses in BBV.

Your results seem pretty gross. But it's a mess to calculate the exact probability of running this bad. But you're halfway through already. You have a program that gives the equities, just let it run some hundred thousand simulations, et viola!

If you assume every pot has the same size and your equity in every pot is 50% (or some other fixed number) the probabilities would be distributed binomially. In this case your question would translate into something like: How likely is it to win 148 (or less) of 371 pots when the probability of winning one pot is 50%. The answer to this question is 1/17207. Pretty unlikely.

Obviously that's just an approximation, but the result should be in the right ballpark.

[SongheJe]

Hello sixhigh.
Thanks for your answer.
I understand that difference size of pots make impossible to retrieve the data I needed.
But the luckometer have been done to measure luck relative to weight of pot (I apologize again for my english but it's not my language. I hope I am able to explain, however).
So I modified the program to know the number of win/lose/split.
Number are these:
I was on edge 51.48% of times.
I had average edge 50.50%. So we can assume I was 50/50 in all pots, is it ?
I Won 142 pots, lost 218 and split 11.
I think that for our number we could (with approssimation) say that I won 147 and lost 224, is it ?
So the 1 on ~17,207 maybe possible ?
Cheers
SJ

[Troll_Inc]

Song.,

I think what we need at a minimum is just 3 columns.

1st column = size of pot
2nd column = your equity (i.e. chance to win)
3rd column = actual equity
This analysis will be somewhat crude but probably close enough.

Can you put that in a comma delimited format with your luck program? or better yet cut and paste into and excel file?