By Justin Merry (justin at basement-dwellers dot com) for RotoGraphs/ottoneu
What are the categories and points used in the ottoneu FanGraphs Points system?
Why don’t you award points for traditional fantasy categories, like RBI, wins, earned runs, etc?
The main reason is that these simply aren’t good statistics for measuring player performance. They are all strongly influenced by one’s team, and in some cases are based on such convoluted rules that they really aren’t worth tracking--not when we have much better alternatives. Don’t agree? Take a look at Joe Posnanski’s rants about batting average, earned runs, and wins, and on RBI’s.
Why can a pitcher give up hits all day long and not have negative value?
We do not track hits allowed in this system. The reason is they are substantially dependent on the quality of the fielders behind a pitcher. We instead follow the lead of statistics like FIP and track things that a pitcher can directly control: strikeouts, walks, hit batters, and home runs. The point values actually come straight from the FIP calculation, and which are based on linear weights and are adjusted to account for average values of balls in play.
For most pitchers who accumulate a decent number of innings in a season, the total points accumulated in the FanGraphs Points system will closely approximate the total points in most other fantasy baseball systems. However, there will be pitchers every year (and it’s rarely the same ones in consecutive years) who will see disparities. These will typically be due to unusually good or bad fielding, good/bad “luck” on how balls are hit into play, good/bad “luck” on the sequencing of events (HR, 1B, BB is better than BB, 1B, HR), or good/bad “luck” on run support. Generally speaking, how well a pitcher actually pitched (as opposed to the result of his performance) should be more closely measured by the FanGraphs Points system than most other fantasy points systems.
Why do pitchers get so much credit for an IP?
A pitcher’s primary job is to generate outs, and that’s what we measure with innings pitched. While we cannot get completely away from the impact of a pitcher’s fielders when recording outs generated (they are certainly influenced by a pitcher’s teammates!), they are the major product of a pitcher’s performance. We are effectively giving 5 total “runs” (~45 points) worth of credit per 9 innings to pitchers, as that will be enough to win roughly half of a pitcher’s games. The other statistics we gather are used to modify the strength of that performance: more strikeouts would lead to more wins, while more walks and home runs would lead to fewer walks.
IP is the main measure of playing time for pitchers in our system, because we do not track other measures like wins, quality starts, etc. The points we use for IP keeps our pitcher scoring calibrated with our hitting system (which is based on absolute runs): the top hitters and the top pitchers will both top out at around 1250-1350 points in most seasons.
Why -1 point per at bat?
When an AB doesn’t result in a hit, that means it is an out. We’re penalizing hitters for making outs, just as is done in linear weights (on the absolute runs scale--the penalty would be larger if we were using linear weights above average!).
Why is [insert name of any fantasy player with lots of SBs here] not ranked as highly in this system as he is in another fantasy league?
In most fantasy baseball scoring systems, steals are overvalued compared to reality. In many cases, you will see a steal given the same point value as a single. And often, getting caught stealing has no effect--or, at worst, it cancels out one stolen base in the scoring. Studies of real baseball, however, when the data are expressed as linear weights, show that a steal is worth just a small fraction of the value of a single, and getting caught stealing is far worse than the extra benefit of stealing a base.
Simply put, our scoring system better reflects the true value of stolen bases than most other fantasy baseball scoring systems, be they category leagues or points leagues.