Projections accounting for relief v. starting role

awinters  5/9/2014 8:47 AM

I noticed possible room for improvement in the daily/weekly projections for dual role pitchers scheduled to take a start.  It's well documented that, on average, a pitcher projected to pitch in relief will have better numbers than if he were to pitch as a starter.  I've done some studies on this and have some adjustment figures for innings pitched as starter v. innings pitched as reliever.  For pitchers projected to pitch most of their season innings in relief but to start occasionally, I feel that the daily projections page does not adequately account for the reduced performance likely to result from the starting role.  For example, I believe  this results in the daily projections overrating pitchers like Tyler Lyons or Cesar Ramos when they have a start.  I'd be curious if you can comment on how the Monster factors takes into account a changed role.  

ken_admin  5/9/2014 9:54 AM

Currently, we only have one set of projection parameters for each pitcher so there's no automatic adjustment (other than innings) when a middle guy pitchers as a starter. It's something we can look into to see if there's a generic adjustment can be made or if it requires a unique adjustment for each pitcher. What are your thoughts?

awinters  5/9/2014 10:35 AM

My personal thought is that a general adjustment would be far easier to implement and probably more accurate.  If you want I can share with you the research that I did.  Basically I took about ten seasons and, for each season, filtered only pitchers who had pitched a minimum of 10 IP innings pitched as both a reliever and starter.  For each player-season I compared the difference WHIP, ERA and K/IP and then took the median differential between how the pitcher performed in each role.

ken_admin  5/9/2014 10:40 AM

Yes, anything you can send would be appreciated (kslight@baseballmonster.com). Thanks