L28

OK, I've been doing a lot of maintenance. If you read yesterday's post, you know about the new columns in the FNFT stats. Well, I've changed them a bit. Instead of "L30" for the last 30 days, I've changed to "L28" for the last 28 days. The way the MLB schedule is, it make more sense to use some number of full weeks in this summary.

So, the columns are:

FP = FNFT points in all games this year
FP/G = FNFT points per game in all games this year
L28 = FNFT points in the last 28 calendar days
L28G = FNFT points per game in the last 28 calendar days

Note that L28G is not just "L28/28". It's based on the actual number of games that the player has played in the last 28 days.

Here's another tweak: Ever since I created this thing, the FP column has been calculated by applying a formula to the player's year-to-date stats. Because I can't exactly account for some of the bonuses we use, this number has always been a rough approximation of the player's total points but not an exact count. I've changed that. The FP statistic is now the exact number that you would get if you added each players FNFT points in each individual game they have played. This makes the FP/G and L28G columns directly comparable. If the L28G number is greater than the FP/G number, then the player has been hotter recently than his average for the year - and vice-versa.

Give me some feedback on this if you have any. Does the 28-day window seem reasonable? I could change it to 21 or even 14. The caveat is that I have to use the same window for both hitters and pitchers (well, I wouldn't have to but it would be a trillion times harder to make them different). Since starting pitchers only play every five days, I think you need a big enough window to capture at least 3-4 starts.