Yesterday in the game thread I suggested that, given Josh Towers' struggles this season to pitch effectively as a starter, the Jays might be better suited moving Casey "Tek" Janssen to the rotation and attempting to move Towers. That made me think as to what the overall implications and effect on the team would be if they were to pursue such a course.
Janssen has thrown 38 Innings thusfar this season and has an ERA of 2.37. Now, that's inflated by his last appearance in which he surrendered 6 runs without recording an out, but Casey was doing ridiculously well before that, and much of that was a product of luck. Casey's 18/7 K/BB ratio in his 38 innings don't suggest dominance as a reliever, so I'll keep his ERA where it is, and extrapolate him as a 2.37 ERA pitcher through, say, 90 innings. Now, a reliever's innings are not equal to a starter's, especially a reliever used in high-leverage situations like Casey has been. Janssen's leverage index ("LI") for the season so far is 1.54, which means that each run he prevents actually has 1.54 times as much impact on the game as a run prevented in the 1st inning. A starting pitcher presumably possesses a LI of 1.
If Janssen is moved to the bullpen, Jason Frasor will presumably inherit his 8th inning role. Frasor presently has a 3.09 ERA and much better peripherals than Janssen (33/13 K/BB in 32 Innings). Frasor has not been leveraged as much as Casey, posting a LI of 1.23 this season.
So let's say that Frasor replaces Janssen's innings. Given the two's expected productions, which I think is more than fair to Casey, Janssen would only save 7.2 runs over Frasor, which leverages to about 12 runs. If we expect Janssen to have a 2 ERA and Frasor's to climb to 3.50, we'd be talking about 15 runs, 23.1 leveraged. These stats actually overstate the difference between the two b/c I am extrapolating over the course of the entire season, when the season is almost half over.
Of course, that wouldn't be the only impact to the bullpen. Let's say Frasor's innings are replaced by an average reliever, ERA of 4.00. That would cost the Jays about 8 runs, including leverage, over the course of the season. This is because someone in Frasor's role is expected to pitch less innings, and with less leverage, than someone in Janssen's role. Let's add 12+8 and say that the overall cost to the Jays in moving Casey to the rotation would be 20 runs over the course of a full season. Of course, that person replacing Frasor would now have to be replaced, but since we're already assuming that pitcher is essentially replacement level, that is not likely to have much of an impact. Let's add another 4 runs for margain of error, and say 24 runs.
Now let's look at the benefit to having Tek in the rotation. Of course, that largely depends on what type of starter he would be expected to be. According to this Baseball Prospectus article, the average pitcher can expect to have about a 25% higher ERA as a starter than as a reliever. There are some reasons to suspect that Casey might enjoy better long-term success as a starter, but we'll have to ignore those for now because he has been so effective in relief. Using that method, Casey would be expected to have a 3.00 ERA in the rotation. Now, I don't think that's realistic, because I don't think he'll keep up an ERA in the low 2's in the bullpen. But that really cuts both ways, because it is tough to downgrade his starter performance too much without also lessening the effect that moving him out of the bullpen will have, because if you assume he won't be effective as a starter, it means you probably think he won't be able to keep up his performance in the bullpen.
Anyway, let's assume Casey does significantly worse than the expected Baseball Prospectus reliever-to-starter transition and puts up an ERA of 3.70. Let's also assume that Towers has been unlucky and can get his ERA down to 5.00. Right now, it is 5.75, but I don't think 5 is unreasonable. Still, even with these very conservative assumptions (conservative in the sense that they in all cases assume against moving Casey to the rotation), the difference between Tek and Towers would be over 29 runs, again, over the course of a full season.
Even with conservative assumptions, there is a net gain to moving Tek to the rotation and replacing his innings with Frasor. This will become even more true if Brandon League is able to return and contribute positively to the bullpen. This is because 1) generally speaking, starters are more valuable to their teams because they pitch so many more innings; and 2) the difference between Janssen and Towers > the difference between Janssen and Frasor. Both these things together make moving Casey to the rotation a good idea.
And there are other factors as well: 1) If AJ stays hurt, than the Jays could use Towers as a spot-starter instead of one of the AAA crew, which would presumably help the team; 2)If AJ is not hurt, having a more reliable 5th starter (Tek) would allow the team to give Doc, AJ, and the others more rest, which might keep them fresher and healthier; 3) Tek is 25 and has shown quite a bit, including 4 above average pitches. If his future is in the rotation, it might make sense to give him some experience doing that this season, so he is able to put in a full season as a starter next year. It is unreasonable to expect him to go from 80-90 relief innings this year to 200 innings next season without getting hurt.
Tek to the rotation!