C.lupus wrote:floppymoose wrote:C.lupus wrote:Just out of curiosity, are you using some kind of mathematical algorithm for these w-l predictions?
Not what I would call an algorithm, but an approach. The rough idea is that I take the projected rapm values and projected number of possessions for every player in the league that I think will get at least a certain minimum number of possessions, and use that to calculate a team scoring differential, and then use that to get an overall win loss. There are numerous complications along the way. I use a replacement player value for teams who are going to be forced into playing rookies or unknowns. I have to adjust the possession count from last year because players change teams, come back from injury (Kobe), get older, grow into larger roles, etc, etc. I also hand adjust some of the rapm values if the player appears to be an outlier (Beverly probably not really a top 10 player in the league). And I also try to account for the conference splits and uneven schedules.
I'm not finished combing all the data and there are some win totals I don't believe, like 15 for the Lakers. But I do think they will likely be last in the west.
Seems like a reasonable approach.
I've seen teh list for all 30 teams, and by and large that list looks pretty good. There are a couple teams that seem way too low, like teh Lakers and Hornets, but, the fact that the overall list passes the sniff test leads me to believe that the methodology is sound. So, there is no salt involved in the rankings at all as far as I know, I guess maybe he fudged on a few players, but, it's really a ground up approach based on data and projections.
Personally, I think the Lakers and Jazz are going to be fighting for that last spot, both around 25 wins (I'll say plus or minus 3) but, that's just me guessing. I feel a lot more comfortable posting projections based on Floppy's work than I do just my guesses. And who knows? Maybe the Lakers will be a 15 win team and it was everyone of us that was wrong and the system that was right.
For me, I'm going to give Floppy's rankings the benefit of doubt and not be too harsh until we get to the end of the year and can see how they really were.
Anyone that thinks otherwise, I challenge you to predict the record for all 30 teams yourself, and see if you were closer at the end of the year. It's pretty difficult to do. Injuries and trades really throw off early estimates. You might be surprised how off you end up.
"Thibs called back and wanted more picks," said Jorge Sedano. "And Pat Riley, literally, I was told, called him a mother-bleeper and hung up the phone."