Trust by now you've already picked up your copy of the Hardball Times pre-season annual, and are finding it as fascinating a read as we are. Though the real test, of course, will come at the end of the year, when we blow the dust of the projections and see how well they stood up! In the meantime, however, we had a chance to ask some questions of co-author Chris Constancio. He's a statistician who has written for the HardBall Times since 2006, and developed firstinning.com in 2005 as the first web site to provide split statistics, batted ball information, and three-year projections for minor league players. He also works to provide data for a major league organization. We quizzed him both about the projections, and Arizona'a farm system...
Do we really need another set of projections? What makes yours different?
I can honestly say that I believe our projections are the most unique set you can get for less than $10 because we're using information that most other people don't have access to. We accounted for batted ball information at a very detailed level (even for minor leaguers), we individualized the projections based on physical characteristics and performance attributes, and we used growth curve modeling to project each players' performance over the next three seasons. We used detailed park factors so that we can get an accurate take on how, for example, a right-handed hitter's number of doubles would be affected by playing at Triple-A Tucson. We also offer projected fielding values and three different wins above replacement estimates for each player. There's just a lot of information there. And even if you don't need another set of projections, the book is a fun read. The team summaries were written by dedicated and knowledgeable fans like yourself, so readers are bound to find out a lot about teams and players they don't follow very closely.
Some of the fielding figures go significantly against what we'd expect. For example. in AZ, Stephen Drew and Chris Young are seen as among the worst fielders on the team, Scott Hairston the best. How were these worked out, and how does it fit with your description of, say, Young as "capable of playing a fine center field".
There are no publicly-available fielding metrics for minor league players that are any good, so many young players' fielding runs saved projections are heavily influenced by small samples of performance in the major leagues. Those fielding estimates are helpful for most players, but I wouldn't assume much precision for the rookies. My observations and my early attempts at play-by-play fielding metric for minor leaguers suggest Drew will be about average at shortstop while Young will be about 5 runs above-average in centerfield.
Do you think that projections for 2009 are anything more than "finger in the air"?
Yes, they are. I used a pretty rigorous method to model growth rates for walk, strikeouts, hit outcomes, and more for each player. These rates are based on each individuals' age, physical characteristics, batted ball tendencies, and whatever other variables were statistically significant predictors of change over time.
The big problems will occur if the baseline projections are wrong. If Chris Young really turns out to be a .280 hitter rather than a .240 hitter, then you would need to account for that when looking at the long-term forecast. But the general direction and size of the growth rates are very sound. Three years from now, we will be able to look back and see that about 50% of the players will be doing better than their 2009 projection while about 50% will be doing worse than their 2009 projection.
Your prediction for Randy Johnson's ERA is 3.43, which is lower even than the most optimistic fan in our Snakepit Community Projections. What led you to this conclusion?
He was very unlucky when he posted his 5.00 ERA in 2006, and I worry that many fans are considering that number when thinking about how Johnson will do in 2007. His component statistics suggest he pitched much better than that last year, and if you keep that in mind while also accounting for his performances over the past four years and a move to a more pitcher-friendly league and division, it's not difficult to see why he projects to be one of the best pitchers in the National League this year. Johnson is a high-risk player due to his advanced age, but he could outperform our projection if he's healthy this year.
The system also seemed to love Chris Snyder, projecting his OPS to match Chad Tracy's in 2009. What does it like about him, compared to Miguel Montero?
Snyder has maintained an above-average walk rate at every stop of his career, so he's a safe bet to have at least an average OBP going forward. And I think our system does a pretty good job of recognizing when a hitter has an unlucky batting average of balls in play the way Snyder did in 2005. Montero is projected for more growth over the next few years, but he has only experienced one outstanding season in his career and that occurred in a very favorable hitting environment for left-handed hitters. His baseline projection is merely average for a catcher. It isn't easy to interpret projections for players who are likely to be used in platoon roles, however. Montero could easily outperform his projection if he only starts against right-handed pitching, for example.
The "first wave" of AZ prospects reaches the majors full-time this year, in Drew, Young, Quentin and Montero. Which did you think will perform best a) this year, and b) in their career?
Great question. I think Quentin will be the most productive hitter right now and during his career, but Drew and Young are not far behind and either of those two players might have more value because they can play premium defensive positions.
How do you see Justin Upton projecting in the majors?
It's too soon for a meaningful projection, but he clearly has All-Star potential. His 2006 season was interesting; he began the season as a productive hitter but didn't do a good job of handling pitchers' adjustments the second and third time around the league. He also probably tired under the grind of his first full season, but he flashed some power at the end of the season. I suspect he'll develop like Hanley Ramirez; his minor league performances might be very inconsistent over the next couple years but he won't disappoint when he reaches the major leagues.
Pitching still seems to be an issue for Arizona in the minors; which of their arms have the most potential in your eyes? And will Jason Neighborgall ever be anything more than a 102 mph loose cannon?
Three months from now, I suspect Max Scherzer will be the Diamondbacks best pitching prospect. Nobody else stands out as a possible future ace in a rotation, but I think there's reasonable depth among pitchers in the farm system. Micah Owings is going to be a solid mid-rotation starter, and wait til you see him hit! I like Dana Eveland's chances of being a productive big league starter, and I think recent draftee Brett Anderson might develop into a similar type of pitcher in a few years.
Eduardo Baeza is flying under the radar and could emerge as a top ten prospect in the system this year. Baeza was promoted to Lancaster as a result of the Micah Owings and Greg Smith promotions last summer, and in retrospect it looks like a bad decision for him. Baeza immediately lost his control; he walked 11 batters, hit two others, and threw three wild pitches in his first three starts in the California League. His track record suggests he's better than that, and if healthy he could put up some very impressive numbers this year.
I really can't comment on Jason Neighborall's future. A lot of my analytical strategies rely on looking at developmental trajectories of minor league players in the past, and there is no precendent for a pitcher like Jason Neighborgall. He's really that unique.
Tell us about one prospect in the AZ system, whose potential is bigger than the media coverage they get.
Gerardo Parra is a 19-year-old outfielder from Venezuela who probably deserves a lot more attention than he is getting right now, but I guess that's bound to happen to teenagers in a deep farm system. From a performance perspective, there's a lot to like; Parra had a solid Dominican Summer League campaign in 2005 and he demonstrated good plate discipline and excellent contact skills during his first season in the U.S. last year. When I look at his physical characteristics and early performance indicators, however, I don't imagine him evolving into a noteworthy power threat. I see him more as a good fielder and well-rounded offensive player. He'll be in the Class A Midwest League this year, and I wouldn't be surprised to see production comparable to what Carlos Gonzalez did at that level two years ago.
Put you on the spot time. Having analyzed every player in the majors, who do you predict to win the NL West, the NL and the World Series?
I honestly like the Diamondback's chances of winning the National League West. Some folks at THT are using the projections to predict final standings within each division, and you might be pleasantly surprised at the results. Its tough enough to predict how any team is going to perform over a full season, so predicting how teams will do over 5- or 7-game series is impossible right now. The Yankees and the Mets have two of the best teams on paper, so I'll guess that the Mets win the National League Championship and the Yankees win the World Series.