NBA Black Box

With the start of the NBA season on Tuesday, my life just got a little busier, though I can't wait to see if I can repeat the results of last season. For those who haven't checked it out yet, NBA Black Box is the little sister of NFL Black Box (NFL Black Box is the big brother). I began to write a possession-by-possession NBA simulation, but while I worked, I got sidetracked by a few equations I had come up with. NBA Black Box is not a simulation, it is a set of algorithms that work together to pick the straight up and ATS winners. I've given up on the elusive O/U; for some reason, it just doesn't work. There are no plans to finish the simulation - the algorithms are doing just fine.

Here are the results from when I started tracking, on March 1. They exceed my wildest expectation.

SINCE MARCH 1

WINS			PLAYOFFS WINS
Correct	Games	Win%	Correct	Games	Win %
279	353	79.0%	36	48	75.0%
SPREAD			SPREAD
Correct	Games	Win%	Correct	Games	Win%
242	411	58.9%	32	54	59.3%

Those numbers are correct, check out NBA Black Box to verify. From March 1 to the end of the season, NBA Black Box was 279-74 (79.0%) picking the winner and 242-169 (58.9%) ATS. Similar results for the playoffs. From March 1, using a unit of 100, NBA Black Box returned 5610 units, for a ROI of 12.4%. And that's only after 2 months.

I'm going to track the progress on my own initially (not going to post the results), until I see some consistent performance. It takes time for reliable stats to build up. I have no idea how long that will be. It could be a couple of weeks or it could be a couple of months. The bottom line is I don't want to put unreliable picks up - I don't want people to lose money as a result of bad picks and I don't want to ruin my reputation for making quality predictions.

I'll keep you folks updated as the season continues.

Jonathan