Last week, I took a look at player projection systems, specifically focusing on the Dodger offense. The purpose of the post was to describe how recent research has shown that revised in-season projections were correct on average. The post went on to examine which Dodgers were over and under-performing their Rest of Season projections.
There’s another angle which needs to be examined on this topic. The projection systems update to take into account things which have happened so far this season, and at this point players have enough plate appearances to make measurable impacts on future expectations. Below is a table showing pre-season and rest-of-season Steamer projections for all of the Dodger offense regulars, taken after the conclusion of Sunday’s game. The offensive numbers are shown as wOBA (weighted on-base average).
These projections are supposed to be “sticky,” since baseball players tend to regress towards their individual means, but who has left the biggest impression on the Steamer projection system so far this season?
|Player||Pre-season projected wOBA||RoS projected wOBA||Difference|
|Scott Van Slyke||.327||.329||+.002|
The largest improvement upon a player’s pre-season projection belongs to Dee Gordon, whose projected wOBA has increased by .015. To give an idea of the significance of the revision, it’s the difference between 77 wRC+ (23% worse hitter than league average) and 88 wRC+ (just 12% worse than league average). It still hasn’t fully bought into the level of play Gordon has shown thus far this year, but it’s about as much movement as you could reasonably expect. Drew Butera‘s projection has improved from terrible to slightly less terrible, and Yasiel Puig has improved his projection significantly as well.
Some of the players without a significant impact to their projections might be a bit surprising. Scott Van Slyke is crushing the ball so far this year, but since a significant portion is with the platoon advantage, a high BABIP, and a slightly high HR/FB ratio, his projection has not moved very much. Matt Kemp‘s projection hasn’t budged, which seems counter to how good he has been recently. However, as of Sunday, Kemp’s 2014 wOBA was exactly the same as his pre-season projected level of .347.
Andre Ethier leads the “decliners,” which seems valid given how much he has under-performed expectations so far. Adrian Gonzalez falls into the same grouping, though even after the downward revision he’s still projected to have an excellent 132 wRC+. Justin Turner‘s appearance on the decliner’s list is probably the biggest surprise, but his great numbers this season have been fueled by a very high BABIP which the projection does not expect him to maintain.
Even though most players have moved their projections a little bit, as a whole, the magnitude of adjustment has been around zero. Most “break-downs” have been matched by another player overplaying his pre-season expectations. Without Dee Gordon’s break-out and Yasiel Puig’s sustained success, the Dodgers would be projected to be in a very different place.