I've just seen I became gold member again, and I donated my first 1000$ to poker platforms in rake. I think a good idea to review my winnings/losses.

heres my graph:

Uploaded with ImageShack.us

The Y-Axis is BR, not BI!
it's after i lost 43$ of the starting capital and redeposited. but still i've never been down more than -7 :-) To bad I cant get the hand histories loaded for the smaller rooms...

all in all i played
- about 20.000 cash hands (whenever i hit a downer and tried something else - but i don't like it too much)
- about 1700 DONs (1,2 and 5 $ on different sites)
- about 1600 Turbo SNGs ( " " )
- aout 130 MTTs

I played on cake, party, heaven, everest and betfair. at the beginning of the red line i won 2 MTTs in a row:-)
Btw in the last 2 days i came back from 227$ to more than 400, but yesterday evening it still felt like losing when i lost 50$. Quite strange, isn't it? I win 100$ but still i feel unlucky^^ maybe it's because I got all my premiums cracked, I even lost a huge MTT where i was 1st out of 1500 for more than half of the game (4 hands in a row: AQ, AA, KK, AQ; I all played them to my best, and lost vs hands like JQs preflop)

Well, back to the topic: Because I moved between rooms, and limits, I don't have sample sizes. The only "bigger" samples are 500 in the 2$ DONs at cake, and 400 for the 1$ turbos at betfair. In the 2DON it seems like I'm winning(~ 80% confidence if i calculated correctly). for all other formats I have less than 300 games played, so i can't tell whether im losing or winning there. And I did not want to lump all games of 1 limit together (i.e. 1$ of cake with 1$ of betfair etc) because I don't think I'd have the same winrate at everest and party (due to different rake, rb, boni and player levels). Or should I?

Now, the linear regression doesn't tell me much neither (winnings of 11ct/game, R² = 0.3), because of my MTT winnings, and because I expect to win more in the 5$ games and less in the 1$'s . should be more of an exponential function anyway^^
If I plot my total cumulative Buy-in vs BR I get an overall ROI of 2.5% with the same R² (lumping all limits and rooms together). But I still should have different ROIs at diff. limits.
If I split my graph at the 3 color sections, sample size gets even smaller.
If i dont include the MTTs, I have winnings of 2ct/game, R² = -0.27. That graph looks creepy! *erase*

So, how can I proceed to see whether I'm winning or not? IS there anyway except increasing sample size? Should I consider staying at one room even though Im not playing with a bonus anymore? What's ur opinion?
And one more thing, I tried some 10$ DONs once. I felt they are a lot harder than the 5$'s...Do you think they are still beatable at 10% rake + rb, or are they a waste of time?