(*) FANALYTICS: Burning my one replacement
Last week, I participated in a draft for a 4-week league. This is one of those newfangled short-term competitions, though mine is a private one run by a group of friends. Most of these new games are for a single day or a week. I thought that was a little too much like blind dart-throwing for me. Four weeks seemed like a reasonable length of time to show off how random performance is.
Twelve teams. Just 16-man rosters. Nine hitters, one at each position, and seven pitchers. That's just over 25% penetration into the player population. Certainly, it should be pretty easy to skim off the best players in a format like this, right?
However, the one wrinkle in this league was that you could make only one roster move during the four weeks. You could make it at any time and it would take effect within 24 hours. So, with virtually no room for error, I spent the draft looking for players who were the embodiment of good health. As much as possible.
Well anyway, the league runs from June 11 until July 8. Opening Day for me was this past Monday, and on the mound for my team was... Brandon Morrow.
So, nine pitches into my season and one of my five starters is now gone.
The first question is, should I burn my one replacement this soon? A starting pitcher is important, but if one of my stud hitters goes belly up tomorrow, I'd much rather dip into the bountiful free agent pool for him. Passing on this replacement would put wins and strikeouts at risk—everything else could be covered. But a downed hitter could potentially cost me in three or four categories.
Heck, there is no way I could predict a hitter going down. And the sooner I used my replacement, the more time I could squeeze out of him.
Sigh. So much time left.
Actually, "time" was my friend in this instance. With the varying rotation schedules of my pitchers, I had a few days' grace before I would start falling behind the pace of accumulating starts. So I assembled a list of possible replacements and made my decision based on that. Any pitcher who I could squeeze an extra start out of might get an edge. And in the mean-time, I could wait a few days to see if any of my hitters went down.
Research conducted several years ago showed that "once a pitcher enters into a dominating streak of any length, the probability is that his subsequent start is going to be a better-than-average outing. The further a player is into a streak, the higher the likelihood that the subsequent performance will be of high quality. In fact, once a pitcher has posted six dominating starts in a row, there is greater than a 70% probability that the streak will continue."
There is more about this research in the Baseball Forecaster, but the bottom line was that my best candidates were pitchers who had just started a streak of dominating starts.
Looking for these, I narrowed my list of replacements to five: Clay Buchholz, Ryan Dempster, Phil Hughes, Ubaldo Jimenez and Jered Weaver.
Note that in an exercise like this, ERA is meaningless. I might get five or six starts total out of any of these pitchers during the four weeks; volatile ERAs would mean that I had to focus on the underlying skills. That immediately ruled out Buchholz and Jimenez. Dempster has got some upside, but his inconsistency and poor surrounding cast made him the third cut.
That left Hughes and Weaver. Both on good teams. Weaver had a slightly better skills profile but he was coming off the back injury which was worrisome. Phil Hughes... well, he also had the earliest next start of the bunch.
So, there it is. And I am now officially a spectator.