Comparison of actual result to expected result
 

Since I use Eastbays SNG tool to determine if i should push in a certain situation, i thought it would be interesting to see if the EV estimation is close to my actual results.

[img]/images/graemlins/diamond.gif[/img] What did I do?
I broke into my pokertracker database. In each torunament where I was still alive by the beginning of level five, I made a query which gave me: my chipcount, the chipcount of my opponents and the amount of money I ended up winning in the tournament. I then took this data into an Excell sheet and calculated my expected win using conditional probabilities (like Eastbay). I then compared the expected values to my actual winnings.

[img]/images/graemlins/club.gif[/img]What did i find?
As one would expect from a winning player, I found that I was winning more than the conditional probabilities suggects. Overall I was winning 3% more. I was doing best when my expected win was over $200 since that gave me an extra 5 % compared to Eastbays tool. In tournaments where my expected result was less than $200 the edge was only 2% and when less than $100 i was actually at only 97% of the expected.


[img]/images/graemlins/diamond.gif[/img] How does this effect my play?
I guess I should push a litle less with a big stack since i risk more than the tool says. Clearly it cant be much of a correction though, since pushing is what gave me the edge in the first place!

[img]/images/graemlins/heart.gif[/img] Disclamer
Data consist of 4500 $100 tournamnets where i was still alive by level 5.

The expected value was calculated by using conditional probabilities like Eastbay. I did not however take my position in account as opposed to Eastbay. I used the chipcount before the posting of blinds.

[img]/images/graemlins/spade.gif[/img]comments?

Re: Comparison of actual result to expected result
 

When you say "conditional probabilities," do you mean your prize pool equity as determined by ICM?

Re: Comparison of actual result to expected result
 

Cool.