Visualizing Draft Strategies

Oakland’s Moneyball draft received a good deal of criticism and inspired countless “statheads versus scouts” debates concerning player development. Teams like the Dodgers used their first ten picks of the 2002 draft on high school players like James Loney, Greg Miller, and Jonathan Broxton. Oakland, on the other hand, used all of their early picks to draft college players like Nick Swisher, Joe Blanton, and Jeremy Brown.

What kind of draft strategies are teams using today? Are some organizations still relying on just high school or college players? I’m going to attempt to answer these questions via an exploratory analysis of 18 organizations’ drafting strategies.

The Method

Correspondence analysis is an analytical technique that reveals correlations between categorical variables. For our purposes, the categorical variables will be major league organizations and drafted player types. Correspondence analysis also facilitates visual representations of this relationship, so we can look at a scatter plot and find out which teams are similar to one another in drafting strategy and also what variables account for the similarity.

Each drafted player is categorized as one of four player types; hitter with college experience, pitcher with college experience, hitter with high school experience, or pitcher with high school experience. I decided to only include draft picks from the first five rounds of the past three drafts, and only included 18 organizations that have maintained some consistency in their front office during that time. If your favorite team is not listed below, it’s probably because they have a new scouting director.

The Results

Three years and 295 draft picks result in the following plot:

Interpretation of this plot is fairly straightforward. Teams’ proximity to one another reflects similarity in drafting strategies. The proximity to variables such as “High School Hitters” or “College Pitchers” reflects the team’s preference for those types of players relative to the other organizations included in this study.

The first dimension (the x-axis) explains most of the variance in organizational draft tendencies and simply represents the prevalence of high school or college players among the team’s early draft picks. On the far left we have teams like the Angels, Braves, and Dodgers. Those organizations draft an above-average proportion of high school players in the first five rounds of the draft. On the far right we have teams that almost exclusively draft college players at this stage of the draft: the Blue Jays, Giants, Diamondbacks, and Padres.

The second dimension (the y-axis) gets at more subtle differences in positional preferences, particularly among teams selecting high school players. For example, both Milwaukee and Kansas City slightly lean towards drafting high school players relative to the rest of the organizations in this study, but Kansas City has focused on high school hitters more than Milwaukee and is therefore located closer to the “High School Hitters” point in the upper left quadrant of the plot.

The beauty of descriptive analytic techniques is that you, the reader, are free to interpret the results in a way that makes sense to you. I look forward to hearing about your own questions and analyses of this information over the next week. For now, I’ve decided to categorize the teams in three ways:

”We like to raise our own kids.”

That’s how Atlanta Braves scouting director Roy Clark once explained his drafting strategy. The Braves, Angels, and Dodgers all clearly prefer to draft high school players in the early rounds of the draft. The Angels director of scouting Eddie Bane explains, “If Jered Weaver falls in your lap, you take it. But my thing with college players is they’re older. By time you sign him, they’re 23 and the clock is ticking.” While the Angels draft more high school hitters than most teams, the Braves and Dodgers have focused on drafting high school pitchers.

None of these organizations are dogmatic in their approach. The Angels are the most extreme; they drafted five high school players with their first five picks of last year’s draft. They also drafted a few college pitchers two years ago, however. The Dodgers and Braves both selected more college players than usual in last year’s draft, but it’s not clear if this was due to circumstances of the draft class or if these selections suggest a trend towards a more balanced draft strategy in the future.

An Attraction to Older Guys

We can identify more extreme drafting strategies at the other end of the spectrum. Some organizations completely avoid high school players in the early rounds of the draft. Under the direction of John LaLonde, the Blue Jays have only selected college players in the first five rounds of the draft. The Giants are notorious for their skepticism regarding the draft and Craig Whitaker is their only high school player drafted out of high school in the early rounds of the past three drafts.

Oakland once belonged to this group, but last year they surprised everyone by selecting three high school pitchers before the fifth round of the draft. In the correspondence analysis plot, Oakland now occupies an unusual space between the “High School Pitchers” point and the cluster of teams that focus on college players. The Diamondbacks are also remarkably consistent in drafting college players in the first five rounds of the draft, but they could not resist taking Justin Upton out of high school with the first pick of last year’s draft.

The Middle Ground

The “statheads versus scouts” debate regarding player development continues to persist, but most teams appear to prefer a balanced approach to drafting college and high school players. Teams in the middle ground certainly do demonstrate tendencies in their drafts; the Marlins and Twins draft more high school players than the White Sox and Nationals, for example. The most important finding here, however, is how many teams occupy the central space of the plot. In the middle ground there are no neat clusters of rigid draft strategies. Instead, most of these teams certainly appear to be looking all over the population of amateur players for the best fit for their team.

One observation that I find interesting is how three “large market” organizations are most extreme in their pursuit of high school players while “small-” or “mid-market” teams make up most of the middle ground in this plot. I’m not sure if this is a coincidence or if there’s some good explanation for this relationship.

What can we learn from this?

The 2006 first year player draft begins tomorrow. By now, you are probably wondering how any of this information can help you understand what will happen later this week.

The predictive utility of these results are admittedly questionable because an organization’s draft position and the makeup of the particular draft class can greatly affect draft day decisions. For example, Kansas City has a history of drafting high school hitters more often than other teams, but they will almost certainly take a college pitcher with the first pick of tomorrow’s draft because nearly everyone agrees that there are no high school hitters worth taking with the first pick of this particular draft.

I think there are two things we can learn from this analysis, however. First, some organizations are indeed dogmatic in their approach to the draft. Nearly all teams will claim they are taking the “best player available”, but a few organizations will in fact focus on high school players in the first few rounds of the draft while other will exclusively draft college players. It’s nice to be aware of these exceptions, but the second finding here is that the majority of teams will not focus on one type of player in the draft. Instead, most teams select a range of players in the first five rounds of the first year player draft.

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: THT Daily: June 5
Next: Crucial Situations, Part 2 »

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>