2015 Pitching Projections — Interactive
Last week at FanGraphs, we looked at our Positional Power Rankings for position players. As a companion to our rankings, we had our offensive projections built into an interactive data visualization. This week we are ranking pitching staffs. In the interest of synergy and symmetry, I constructed an analogous pitching projections data visualization using FIP as the primary comparison metric.
This interactive chart is similar to the one I constructed for the batters’ OPS, except there are a few more options to choose from regarding metrics. The chart defaults to FIP, but you are able to view the data by either ERA or WAR as well. Since relief pitchers have a fundamentally different role and stats than starting pitchers, you are also able to filter the players based on those two positions.
To give you a summary of what you are looking at, each circle represents a pitcher: green is for a starter and blue is for a reliever. The size of the circle represents the number of projected innings pitched. The gray bar illustrates the 25th-75th percentile for each team, and the yellow line is a measurement of the combined performance of the players which are visible on the chart.
Since we have multiple metrics on the same chart, the yellow line, which I’m using as a general measure of central tendency, changes for each metric and set of pitchers displayed. For WAR, the meaning is simple; it is the average of the WAR for the players displayed on the chart.
FIP and ERA are a different story. The average of players’ FIP and ERA doesn’t represent the team’s performance; this would overvalue poor performing pitcher who only pitches a handful of innings. To correct for this, I used the rotation’s or bullpen’s combined FIP and ERA from the team projected stats which can be found using our Depth Charts tool. For a view which has both starters and relievers, projected innings pitched were used to combine the position’s metrics into a single metric for the entire team. One complication of using the team performance is that it can fall outside the 25th-75th percentile range I have displayed. This is indicative of the distribution being skewed in one direction or the other.
For example, Cincinnati’s relief pitcher FIP falls outside the 25th-75th percentile. Aroldis Chapman is so far beyond the rest of the bullpen that he single-handedly pulls the team metric up.
All projections current as of 3/31/2015.
I build things here.
Wouldn’t it make more sense to do this graph in FIP-, rather than unadjusted FIP? Park factors obviously give an advantage to many of the teams at the top of the chart.
That’s basically what the WAR sort is, no? Given that it’s FIP-based, basically the only changes in rank or grade would come from park/league adjustments, I think.