Fantasy Baseball Notes

Regular readers of this blog will probably want to skip this one. This time of year is exciting for me because fantasy baseball uses numbers, research and strategy to win. Around this time magazines start coming out with projections of how well real-life major league baseball players will do this year, and in my league you rank a couple hundred players before letting the computer simulate a draft to form twelve teams that square off all summer. Below describes notes on my ranking system this year so that I can refer to it quickly a year from now rather than reinventing the wheel.

Projections came from the Fantasy Baseball Index (FBI) which hit the shelves approximately the first week of February. I reserved the right to adjust some of their projections. For example, some young players had "breakout" years nearly double any previous output, and FBI projected that same level of output, so I'd tone it down. Some (like Kerry Wood's 2.05 ERA when he's never been below 3.00) just plain didn't make sense. As I typed the projections into my spreadsheet's "Data" tab, I took some creative license to adjust those numbers accordingly.

In the "Draft" tab, I jotted down the level of performance required to earn 10 points in each category. The five usual categories for hitters (runs, home runs, runs batted in, stolen bases, batting average) applied. For pitchers, I allow a higher ERA for starters than relievers, so that the average comes out to reach the 10 point level (assuming 200 innings pitched x 5 starters, and another 250 from relievers). This year, a starter's ERA of 3.60 and reliever's 3.00 would hit the 3.48 team mark. I do likewise for SO%, asking 80% for starters and 100% for relievers to hit 83% overall. For walks+hits per inning pitched (WHIP), the difference between starters and relievers would've been so small that I just set the levels equal. For starters, categories are wins (15, every year), strikeout percentage, ERA and WHIP. For relievers: saves, strikeout percentage, ERA and WHIP. For starting pitchers, no points are awarded for reaching 200 innings but 1 point is docked for failure to reach at least 180 innings; for relievers, the point total is reduced by two. This is to limit the chance of pitchers' totals from exceeding hitters' totals, and likewise to keep relievers at or below starters.

I grant one point for each category in which the 10 point level is projected to be reached. If someone is way below that level, I grant a negative point, and if way above then two points. That keeps people like Juan Pierre from being overvalued... his runs and stolen bases might be high (1 R point + 2 SB points) but his power numbers are so abysmal (-1 HR point -1 RBI point) that he stays close to zero in value.

Here's what "way above" and "way below" are:

Runs (target 90): +/- 20
Home runs (22): +/- 10
RBI (85): +/- 20
SB (14): +/- 10
AVG (.286): +/- .020
W (15): +/- 5
SV (28): +/- 10
SO%: +/- .10
ERA: +/- .25
WHIP (1.24): +/- .10
IP* (200): - 20

* For innings pitched I wanted to get a measure of durability. So for purposes of strikeout percentage I used the projected value, but in terms of granting points for innings pitched I used their three year average. That way, if a pitcher's only in his third year or less (like Tim Lincecum), he likely ends up with -1 point. In the case of known injury risks like Rich Harden and Joba Chamberlain, I arbitrarily subtracted an additional point.

Then I sorted them according to their points, before making further rankings within each rating (e.g. sorting ratings of "5" relative to each other).