In "The Hidden Game of Baseball," Pete Palmer and John Thorn developed a linear weights system for valuing players. Linear Weights assigns a weight to each stat, based on that stat's likelihood of producing runs. Although calculated differently, it works similarly to Bill James' Runs Created.
Runs Created, however, did not address pitchers. Linear Weights did. The oversimplified LWTS for pitchers is to divide League ERA by the pitcher's ERA and multiply it times innings pitched. The average pitcher scores 0, since if the pitcher had a league average ERA you would be multiplying 0 times the innings pitched. Negative scores occur if the pitcher's ERA exceeded the league ERA. There's a park adjustment in there too.
I calculate linear weights for each pitching season. I calculated the career LWTS totals for all pitchers in the Hall of Fame and developed a grade scale.
I also wanted to take into account the effect of defense behind the pitcher. Baseball Prospectus includes a pitcher's DERA, which is what the pitcher's ERA would be if he had an average defense behind him. After doing the initial raw LWTS career grade above, I separately grade the pitcher on what I call his DERA LWTS. I the 3-year peak, 5-year consecutive peak, 7-year peak, LWTS per 100 innings pitched, and career linear weights, much like I do for Win Shares and WARP1. I did the same LWTS calculations for all Hall of Fame pitchers to develop a scale for measuring whether a pitcher fits within the Hall based on these DERA LWTS scores. I then average the LWTS career grade with the DERA LWTS grade.
I did a separate calculation for Hall of Fame (or near Hall of Fame relievers), because their LWTS and DERA LWTS totals are lower than starters. Therefore, relievers get a different grade scale.
This test gets normal weight in a pitcher's GPA.
