Thursday, February 25, 2010
Projecting individual prospects
Posted by Alex PediciniBryan Smith wrote an interesting and thought-provoking article earlier today over at Fangraphs. Basically, Smith is trying to formulate a method to predict a median future performance for an individual prospect (he used Cubs shortstop Starlin Castro as his example).
This is certainly an worthwhile idea and one that could potentially revolutionize the way prospects are viewed. As we know there is a sizable risk involved in almost any prospect, but if we could somehow quantify this risk we could more accurately project future performance.
One way to do this would be to develop some type of confidence intervals for a prospect's future performance during his first six major league seasons. However, to do this would require estimates of injury risk and potential "bust factor".
I'm not sure how probable or realistic this would be to do but it is still an exciting idea nonetheless. The guys over at Sons of Sam Horn have been putting together something along these lines. They are compiling a "community success probabilities" for Boston prospects. This is a good start to trying to predict a prospect's future role and having a variety of opinions may yield better results, although obviously the younger and further away a player is from the majors the more difficult this becomes.
I'm sure there is plenty of proprietary work done by clubs regarding this type of analysis and for good reason. If a team can narrow their estimates of a player's future ceiling they will be better prepared to develop or possibly trade this player. I expect there to be major strides in this area of prospect valuation and projection as the decade progresses.
Alex is a student in Boston. He welcomes all questions and comments here
We do projections based on the average performances of all those players who have gone before, playing in the same situations/contexts.
What we can also do is look at the future career path of the most similar players from the past. Castro is a SS who will be 2o years old in 2010, and who highest level last year was AA. I rate his MLEs as +8 batting runs above replacement for SS, and +7 glove. Using aging curves derived from season stats, I have a peak for him in 2014 at a mean projection of 305/343/442, good for 22 BRAR. A consistent 3 WAR player, good but not spectacular.
So look up (in recent years) all the other 20 yeard old shortstops in AA who had similar offensive and defensive profiles. How many never made the majors? Became regulars? Became stars?