FT (R U KIDDING ME?)

1 for my last 47 FTOPS.

I really think Fu(ck)ll Tilt has it out for me. I've started tracking my allins there in the NLH tourneys (just for info purposes to see if I'm crazy). My results there over time are ridiculously bad.

Tournies
ROI (courtesy of official poker rankings)

2009 FT -59% Stars 70%
2008 FT -12% Stars 8%
2007 FT 48% Stars 105%
Overall FT -7% Stars 32%
My stats (which go back 1 yr further) FT 5% Stars 32%
other sites Party 4% Live 67% UB 83% Cake 68%

Cash Games

2009 FT -6k Stars -10k (stars double the hrs & higher limits)
2008 FT -56k Stars +75k (stars about 25% more)
2007 FT -4k Stars +24k (stars more than doublte the time)
Overall (includes some pre 2007) FT -58K Stars +94K
other sites UB +44K Party +81k (this one's dubious pre-uigea) Bodog -10k Cake +9k

The disparity among Tilt and pretty much every other site is crazy. The fact that cash games and tournies pretty much look the same is odd too. I first started to notice this during my short -term short stacking phase. I played Tilt and Stars during it, and I used Poker Tracker which you could load into some software to measure your allin luck (to see how you ran during allins, this was all that mattered really for a short stacker). On Tilt I ran so far below expectations that I was a slight loser. On Stars, I ran slightly above expectations, but this meant I absolutely crushed the games.

I know the disparity on Tilt could be mathematically explained through standard deviations from the mean. At the same time, eventually reversion to the mean is expected if your sample size gets large enough. As time continues to go by, I am really beginning to wonder............................