projections

i was showing my daughter for the first time the epic "a charlie brown christmas," and i was struck by a line that was delivered so poignantly, so humanly, so utterly childlike. the line is appropriate to both the holiday season and the baseball free agency season. the line called to mind matt holliday and everybody else gunning for the bigger dollars and the bigger years, often demanding sums that have nothing to do with their own needs or wants, but to demonstrate that they are more important than the next player in line (e.g., the rumor that linececum will demand a symbolic $23,000,001 in arbitration to show that he is more important than the best paid pitcher in baseball).

what matt is saying is: all i want is what i have coming to me; all i want is my fair share.*

and just as sally is right and yet so wrong at the same time, because looking out for her own self-interest (something that we all do and must do) clashes badly with the themes of the season, which is supposed to be about selflessness and giving, matt is both right and oh so wrong in baseball. right, in that the money may just go into the steinbrenners' pockets or the legendary dewallet should he agree to less than the ultimate penny he can squeeze out of his negotiating partner/opponent. wrong in that baseball, like christmas, is supposed to be above pettiness and above greed. baseball is about allegiances to a team, a team the player is supposed to sweat and bleed and struggle for; how do we reconcile that with the vision of a player pimping himself out to the highest bidder before the last out of the world series has settled in the glove?

baseball is supposed to be wholesome and american and all those things that wake adam dunn up in a cold sweat in the middle of the night. is it an accident that the two most iconic peanuts associations are christmas and baseball? 

mind you, i am not saying that matt holliday is a bad person because he likes money. i look at free agency and baseball contracts and know that that is how it will be, inevitably. you can call it "wrong" if you like; i won't tell you that you're wrong in return. but one might as well tell a five-year old that she's "wrong" to wish for toys and candy at christmas. it is simply the way the world is, and the best one can say is that we can look at it and wish it were different, and we are fools to devote any more effort to it than wishing.

if i die and go to the afterlife and discover that any deity resembles ayn rand in any fashion, i will surely have to answer for the sin that is this post.

* i read on the internet and therefore it must be true that the actress portraying sally in the christmas special was so young she could not read or memorize her lines, and therefore her lines were often read orally to her, leading to the endearingly choppy delivery of that iconic line. 

Something common to many anti-saberites is the objection to projections. if a player is projected to have an OPS of .750 and he has an OPS of .700, was the projection "wrong"? well, no. or maybe yes, depending on your expectation. but people who design, develop, and promote projections would say "no." a projection is not a prediction of what will actually occur. it's a statement of the most likely outcome. nobody makes - or nobody should make - actual predictions about what will occur on the diamond over the course of a season. too much is left to chance in any one season to claim knowledge of an outcome. were it possible to play the same season over and over again with the same player, you would likely see widely varying outcomes from one season to the next.

to give a better analogy, a projection looks at a bellcurve of the likely outcomes and picks the outcome at the peak of that curve. a player could do better than the projection, or worse, in any given season. the claim is that the individual projection is the most likely outcome. take a look at this 2009 projection for albert pujols from baseballprojection.com.

i like the way this depicts the bell curve -- there's a top and a bottom percentile. in an unlucky 2009, albert could have racked up a .922 OPS. in a lucky 2009, albert could have managed a 1.155 OPS. 

a lot of teams would take unlucky albert. 

but the most likely outcome is something within a relatively narrow range on either side of the 50th percentile1.042 projection.  when a projection is cited, they're not saying albert will hit 1.042, but that 1.042 OPS should be more or less in the middle of the field of possibilities, will be at the tip of that bell curve. remember also that albert is well known as one of the most consistent players in baseball, so a less ridiculous hitter might show more than the .230 variation in OPS between 10th percentile and 90th percentile. 

below is a table of projected 2009 results and actual 2009 results. i could try to do a statistical analysis of the accuracy of the projections, but those already exist.  if you're statistically inclined, you'll probably already have seen the statistical discussions of projections. if not, you probably would rather check out the raw data  than have math on a saturday morning.

what's striking is that a lot of the actual numbers end up looking pretty good, relative to their projections. albert came in  at his projection but for a 50 point bump in SLG (es aun mas ridiculoso!). colin rollinford underplayed his projection by 30 points of OBP, but otherwise did about what was projected. skippy picked up about 20 points of BA and OBP but otherwise matched the prediction perfectly. brendan ryan and yadi were better than predicted; chris duncan, rick ankiel, troy glaus, and khalil greene far worse. it's worth noting that the worst correspondences are between the projections and actual results of injured players. obviously, nobody projects rick ankiel to run into a wall. the correspondences for other uninjured players may be rough in some cases, but workable. 

it's most interesting to me to look at the projections for the rookies. i would have expected them to be far less "accurate" than the MLers' projections since the MLers had more data in their datasets and extrapolating minor league data would seem more challenging. tyler greene's projection was eerily on point; brian barden's was not too far off; thurston's was too optimistic. but on the whole the projections were not obviously worse than those with substantial ML time. some like shane robinson or hoffpauir or matt pagnozzi just didn't get enough playing time to really draw any conclusion. the good results reassure me somewhat on projecting, say, david freese or allen craig in 2010. if freese actually puts up the ~.760 ops predicted for 2009 (here i assume the 2010 projection may be worse where it relies on an injury distorted 2009 season) in 2010 and plays +10 run defense, i'll be very pleased.   

i present these numbers more for consideration and discussion than with any clear end in mind.

* * *

i am pleased to see mudflap will rejoin the team in 2010. the gotay signing is unexciting but harmless at worst.

i do want to take a moment to wish aaron miles well. coming off a tough - though in keeping with the theme of this post well-recompensed - year of injury, divorce, and ineffectiveness in a new and suddenly cold baseball town, aaron has landed on his feet, near his california home, with a rebuilding team in a low-stress environment. no one there will hate him for not being mark derosa. while aaron was a passable second baseman on his best days, he seemed like a legitimately nice guy. good luck, grit.

[edit - if people are still having trouble seeing the second half of the chart, here it is]