More on fraction of decisions going to starters
Posted by Andy on March 19, 2009
So this graph is very similar to yesterday's:
The dashed line is exactly what was on yesterday's graph, showing the fraction of decisions going to starters. Today's graph has a little more data, breaking it out my wins and losses. So the red line shows fraction of wins by starters (versus the total number of wins going to both starters and relievers) and the blue line shows the same for losses. You can see that starters have historically picked up more of the losses than the victories, but that gap has closed to nothing in recent years. (This is the same conclusion drawn from my post two days ago where the starters' historical win percentage has climbed from 48% to 50%.)
Check out the 3-year period of 1986-1988. Going from 1986 to 1987, starters picked up more losses and fewer victories. I think this was due to the increase in offense that saw more starters get blitzed in games without any significant change in bullpen usage by managers. In 1988, when the baseball returned to normal, the fraction of wins picked up by starters skyrocketed, as did their average innings per start (as you'll see tomorrow.)