OK, so my first lottery number picking based on complex mathematical probabilities (complex to me at least) is complete, and the results are not thus far encouraging.

If you're not up to speed, my goal is to pick winning lottery numbers in the Multistate mega Millions Game based on how often or not each number has been chosen in the past. For each drawing, I will predict the winning 5 numbers + winning Mega Ball number using both the most-picked and least-picked numbers from past drawings. Drawings are Tuesday and Friday.

So, for the December 8th drawing, the numbers were 1, 15, 29, 32, 45 and a Mega Ball of 8.

Assuming I spent one dollar on each of the most-picked and least-picked numbers historically, here's how I would have fared:

Most-picked numbers prediction: 5, 14, 24, 25 and 42, with a Mega Ball of #13

RESULTS: Not even one of these numbers was picked. I'm out one dollar.

Least-picked numbers prediction: 3, 11, 19, 31, 49 and a Mega Ball of #8

RESULTS: None of the first 5 numbers was correct; however, this method did successfully predict the Mega Ball number, for which I would be entitled to $2. Therefore, assuming I spent one dollar and got two back, I am now at +$1 for this method, making it the early winner.

If I'd put a dollar down on both methods, I'd be at break-even.

I'll continue this for a while to see if anything can be gleaned from it.

OK, there's another drawing tomorrow night. The predictions:

Most-picked prediction: 5, 14, 24, 25, 42 and Mega Ball of #13. (Despite none of these numbers being picked last time, they are still the most-picked over the last 53 drawings.)

Least-picked prediction: 4, 19, 31, 49, 51 and Mega Ball of #11

I'll report back on Wednesday.