Well, block payments just dropped below 11 and difficulty passed 9.54
It's amazing how much a software update and better parameters can get you.
On that note, I'm starting to think the chains/day measurement is becoming less accurate. Either that or I've been quite unlucky.
It seems to not be accurate for me - I haven't gotten any blocks since the 25th (other than an orphan on the 25th) on my desktop though so I feel it's hard to tell. Then again my sister got three blocks on saturday with the same cpu... It seems like it's a rough average that was accurate when it was added but is slowly becoming less useful for judging how often you might get a block from my experiences.
I wonder if there's a way to use stats like jhPrimeminer (ypool client) does but in this build. Basically counts how many chains per hour of different lengths. For instance, my AMD Phenom II X4 is currenlty 17 6-length chains per hour, 193 5 length/hour, and 2046 4-length/hour. There are more stats of course, but that gives you an idea. If you let it run for a day or so, it is quite accurate.
You are proposing to build a model to predict how many blocks can be found per day. As I understandd that was basicly why Sunny added the chains per day gauge. However Sunny pointed out somewhere that chain/d is not block/day. They are just closely related.