why i love BR management

[FlyingSaucy]

Excel is the greatest software of its kind.

If you play strictly SnGs you really don't need pokertracker to do the kind of bankroll trending analysis that is needed to instill confidence in your play. And of course confidence is key to improving your game, even if it is abstracted quite far from every day hand to hand combat.

I still think PT is the best for ring game analysis, and for MTTs you can throw this out the window.... but...

Here you go. It's really quite simple enough to just DIY.

column A: Date
column B: End of day bankroll

You could stop here and that would be a good start. If you're not doing this already, you really should start doing it religiously.

column C: $ change from previous day (simple subtraction)
column D: % change from previous day (simple division)
column E: $ change from 7 days ago
column F: % change from 7 days ago
column G: $ change from 30 days ago
column H: % change from 30 days ago


After you've collected at least 90 days worth of data, somewhere else, either on the same sheet on the top, or another sheet, do your statistical analysis.

R = current roll
S = Std Dev of the entire % change column (week or month)
Ar (Average expected growth rate) = Avg of the entire % change column - F for weeks, H for months
Xr (maX expected growth rate) = A+S
Nr (miN expected growth rate) = A-S

Avg expected roll for x months (or weeks) in future: R*(Ar+1)^x
Min expected roll for x months (or weeks) in future: R*(Nr+1)^x
Max expected roll for x months (or weeks) in future: R*(Xr+1)^x


Start collecting the data today, in a few months you'll have enough for the numbers to mean something, no pun intended. I'm thinking x becomes unreliable any higher than 3 or 4.


So here is an example of what is conjured up - my numbers right now.

Current Roll: 1685.42
Weeks Std dev Avg ch High Low
1 11.62% 6.07% 17.69% -5.56%
1787.66 1983.56 1591.77

Months Std dev Avg ch High Low
1 14.55% 28.15% 42.70% 13.60%
2159.89 2405.13 1914.65

So with the data I have right now, in 1 month I can expect my bankroll to be between 1914.65 and 2405.13, with the average expectation to be around 2159.89. Let's see what happens. Check back in a month i guess. I'm recording these numbers as they change from day to day, to see the accuracy of the projection once I get to that point.

Also note the variance on my projected weekly values is much higher when compared to the average change value, meaning, in a given week, I can easily expect to be down 5.56% (6.07 minus 11.62) of my roll and it would not be out of the ordinary when looking back at previous data.

[Bailey]

I think the real question is.. why you love little boys?..

[a500lbgorilla]

I think the real question is.. why you love little boys?..

and bigred said I was the only one!

[FlyingSaucy]

I think the real question is.. why you love little boys?..

Excellent point you made about how taking the mean of differences between a given range of dates, for example 7 days, can potentially distort the average because each day's worth of bankroll change really factors into 7 different numbers, not just one. Thanks for the insightful response.

[flomo]

rilla that sig is hillarious

[FlyingSaucy]

in case anyone is reading this and was actually interested in my method - BR currently at $2190, so my system appears to be pretty accurate in prediction after 1 month of saving the projections.