Hacking the Scoring System
The Live Vicariously Fantasy Basketball League draft was tonight so I've had hoops on the brain for the past several weeks. I spent hours and hours slaving away at my computer, trying to come up with a draft strategy. I remembered last year that a big flaw in the scoring system was the amount of points awarded for making a free throw; it looked fine at first glance but as we started using it I think everyone realized it was not very well balanced.
This year's scoring system has been changed to account for that flaw and I think it's a lot better. Still, I figured there must be some category that wasn't as balanced as it could be and I was determined to figure out what category that was. And then...let the exploitation begin! =) So here's what I did...
I put together an Excel spreadsheet with every single player in the NBA and their stats for a number of categories (taken from Yahoo Sports). I then entered the formula for our scoring system and ranked the players in order of fantasy points per game (FPG). I made a column for FPG Rank and ranked the players from 1-460 or so. Then I re-sorted the data by each of the major categories: field goals made (FGM), free throws made (FTM), three pointers made (3PM), rebounds, assists, steals and blocks. I copied the "FPG Rank" with the new sort order and put it into a new Excel sheet for each category and made a line graph for each one. The results for FGM graph are shown in the screenshot above.
The way to look at this graph is to figure that the closer the points in the graph are to forming a perfect diagonal line, the closer that high numbers in that category correlate to a high fantasy value. The data points in the graph above are pretty tight, meaning that players with high FGM numbers are really likely to be fantasy stars (which everyone knows). The other categories looked like this:
These are good looking graphs in that none of the categories seem to guarantee a high fantasy value, which is the way I think it should be. I didn't do this for last year's scoring system but if I had, I think the FTM category would be even tighter. I think it's as tight as it is now just because scorers get to the line a lot, so it's going to follow the FGM graph to a certain extent. Anyway, I did find a graph that seemed to be tighter than it should be:
So from this, it looks like steals are valued higher than other categories. Using this information, I tended to choose players that get a lot of steals relative to their fantasy value (which was one of the reasons I drafted Ruben Patterson...just one reason though!). I also looked for players that got a lot of steals last year but didn't get a lot of playing time, which is part of the reason I think Rajon Rando is going to be a monster this year. The sad part is that even if I reach the pinnacle of fantasy basketball and bring home (another) championship, I still can't justify geeking out this much :-\
Good luck to all the teams, it was a fun draft!
