It would take some skill by the pitching coach and manager to correctly identify when a pitcher’s stuff is having an off day as opposed to when opposing hitters are simply putting up good plate appearances. That’s a lot to ask from many current managers.

Still, you could anticipate 5 or so pitchers to pitch per game. On an average night you might have three that put up a good two innings each (two good appearances by your normal starters and a good performance by a typical reliever), one pitcher that gets in and out of trouble giving up one run, and one pitcher who was having an off day and wasn’t pulled soon enough and therefore gives up 2 to 3 runs in fewer than anticipated innings.

On an average night then, you might let in 3-4 runs. Of course that was just a hypothetical designed by me to support my assertion, so you can’t draw much from it. However, the idea might work because of how much better hitters get after each time they see a pitcher, and the fact that by not being asked to go deep, good starting pitchers are able to put up great short outings, more frequently. That and the fact that most bad games/innings are luck getting the best of a pitcher as compared to other outings/innings. Luck strikes relief pitchers and starters alike. See Burnett in game 5. The difficulty, as mentioned above, is identifying when it isn’t bad luck, but bad stuff. Then again, bad stuff hits starters just like bad luck (see Wang on short rest): you start down 4 runs and you’re forced to go with the all reliever plan anyway.

Regardless, starters should be on short leashes.