Filling Out the Orioles Rotation with Bench Players

Yesterday, I posted this:

Last year, Chris Tillman, Ubaldo Jimenez, and Jeremy Hellickson threw 262 innings with an ERA of 7.45.

Since 1991, positional players have logged 284 innings with an ERA of 6.84.

— Camden Depot (@CamdenDepot) January 2, 2018

The brutality of that unholy trinity is put into a very cruel perspective in that the club would have been expected to have given up 17 fewer runs if they would have trotted out guys like Hyun Soo Kim,  Caleb Joseph, and Ruben Tejada onto the mound instead of roughly 30 MM spent on the arms above.  The club would have won about two more games.

So, what is the overall line of a position player pitching since 1991?

n = 181
284 IP
3.0 K/9
5.8 BB/9
6.84 ERA

However, the above might be a poor group to look at when considering what a pitching inclined bench could do.  What if we restricted the players to those who appeared in three or more games.  This might show players who are more capable:

n = 64
79.2 IP
3.3 K/9
5.2 BB/9
5.76 ERA

What is interesting about this group is that with a 5.76 ERA, they are basically replacement level.  If the Orioles had received replacement level pitching from those starters, then the club would have been expected to give up 58 fewer runs and would up with a 78-84 record instead of their 72-90.

Once you go to 4 games or more started, you are only left with nine players and an ERA of 5.74, so it may be difficult to discern ability once you hit that level.

One more angle to look: fastball velocity.
Moving from a professional pitcher to a professional hitter may be pretty disruptive to a batter.  How well do they adjust?

Active	IP	K/9	BB/9	ERA
>85	31	5.8	7.3	11.03
76-85	28	1.6	2.9	6.43
<76	20.1	0.4	2.2	9.3

What we seem to see here is that pitchers in the 76-85 MPH zone for their fastball seem to be significantly more productive than those who throw slower than that or faster.  It may be that the reduced speed is disruptive, but too much of a difference and the batter can adjust to it.

Now, if we stretch the math to find experienced pitchers (as represented above in the 3 GP or more group whose ERA was 77% of the overall group) who can throw in the 76-85 range, then we might expect a one inning average of a 4.95 ERA.  Maybe, but there is a lot of noise in there.

So let us say you can go once through the lineup with a league replacement reliever and then an ideal bench player, you have an ERA of 5.00 over five innings.  If you can do that with 50 starts, you see the 217 runs from that trio above drop to 146 runs, which is an improvement of 7 wins.

Jeez.