startling difference in results between $30 and $50 on PP

[eastbay]

I know these sample sizes are small, but...

Last month, I played about 50 $50 SnGs on party with a dismal ROI of about 20%. Over about the same time period, starting a little earlier, I played about 50 $30 SnGs with an ROI of nearly 70%.

Can this result be "real"? Is the play really that different at $50 as it is at $30? Is my 50 tournament sample really so small that I could be around, say, 40% for both in the "long term"?

My $30 result has been nearly constant for the 50 tourney sample. My $50 result started much higher, around 60%, and slowly went down, down, down. I eventually went back to the $30 level just to make sure my game hadn't fallen apart and I could still get 50%+ ROI there. It looks like I still can (but again, I've only played another dozen there, with a 72% ROI, which nearly matches my previous results.)

Has anyone else seen this kind of jump in performance when moving between buy-ins like this that held up as "real"?

I've seen people say that $50 is a "bad level" because of the mix of fish and strong players. I was highly skeptical of this, but I'm wondering (hoping?) now if there might be something to it. Does anyone want to support that idea or shoot it down? I could take a crack at the $100 level, but I'd feel much better about it if I was hitting 50% at the $50 level first.

eastbay