So I noticed that Curt Schilling's RA9-WAR was actually slightly higher than his fWAR, despite his career ERA being 0.23 runs worse than his career FIP: https://www.fangraphs.com/players/curt-schilling/73/stats?position=P#value
That seemed very odd so I looked at the differential between his career RA and ERA. I guess this anomaly is due to him having only a 0.18 difference between his total run and earned run averages?
For comparisons sake Roger Clemens has a 0.33 differential, Pedro Martinez a 0.27 and Randy Johnson had a whopping 0.44 higher RA than ERA. Greg Maddux's was 0.40
Is there another reason that Schilling's RA9-WAR would be a touch better despite a rather significantly worse ERA than FIP?