Better to Sign out of HS or College? Part 1

With the advent of the August 15th signing deadline, an increasing amount of attention each summer is devoted to which players choose to sign professional contracts and which high school players decide to go to college. With hundreds of thousands of dollars- and sometimes millions- hanging in the balance, the decision of whether to sign or go to college is a monumental one for players and their families. Not only do players have to choose between realizing the dream of playing professional baseball or going to college- two good options to be sure- there is also a pressure to get the best deal possible. The stark reality is that for many players the bonus they receive after signing is the most money they will ever get from playing the game of baseball, so it’s important to get the best deal possible.

In this study, I tried to answer whether players are better off financially by signing out of high school or going to college. In trying to answer this question, I was forced to make several assumptions, and, in some cases, engage in some flat-out guesswork. Therefore, the findings that follow need to be taken with the methodological shortcomings in mind. In this post and the ones to follow, I’ll provide an outline of my methodology along with the results. I’ll let you be the judge of whether or not there is simply too much guesswork to draw a meaningful conclusion. If nothing else, the study should provide a solid groundwork for the types of issues that need to be dealt with in the future.

I looked at all of the players drafted out of high school in the first 45 rounds of the 2003-2007 drafts who chose not to sign (a total of 1,341 players), and compared the round they were taken out of HS to the round they were first taken out of college (e.g., Matt LaPorta is recorded as being drafted in the 14th round of the 2006 draft even though the Brewers took him in the first round of the 2007 draft after he returned to college for his senior season). LaPorta was a notable exception. For most players, recording the first year they were drafted did not drastically alter the analysis.

Some of you might have noticed that including the 2007 draft in the sample creates a small problem. Players drafted out of high school in 2007 have not yet graduated, thus players who were not picked after their junior season (the 2010 draft) are recorded as going undrafted, although several of these players will likely be drafted following their senior season (the 2011 draft). I’m not particularly worried about this, as players drafted as college seniors- especially those not drafted as juniors- typically receive very small bonuses, so including 2007 draftees in the analysis should not have a notable effect on the overall results.

To estimate the bonus a player coming out of HS was likely offered, I calculated the average bonus given out to high school players for each round in the last two drafts. Unfortunately, the samples were small enough that the data was not representative of what a typical player was offered in each round. For example, players in the 6th round averaged a bonus of $508,000 compared to $458,000 for players drafted in the 4th round. Furthermore, there was very little data beyond the first ten rounds.

To smooth the data and create an estimate for bonuses outside the first ten rounds, I put two power-law trendlines through the data. One through all 16 rounds of bonus data I collected and one through only the first ten rounds. Due to the fact that there was likely a selection bias in the bonus data I collected- players who signed may have been offered more money than the average player in that round- the trendline through the entire data set likely overestimated bonuses in the later rounds, while the trendline through only the first ten rounds likely underestimated bonus totals in the later rounds. To overcome the problem I simply used the observed bonus averages for first rounds (where there were few unsigned players and large enough samples to provide a realistic estimate), and then an average of the two trendlines beyond the tenth round.

It’s worth noting that I don’t believe that potential bias made the data unusable, there are plenty of players like Scott Frazier (5th Round Phillies), Kevin Gausman (6th Round Dodgers), and Austin Wilson (12th Round Cardinals) who all turned down a lot of money to go to college in the past draft.

With larger samples and no selection bias to worry about, calculating bonuses for players drafted out of college was much easier. For the first ten rounds, when the samples were large, I used the average bonuses from 2009 and 2010. From the 11th round on, I used the average of the trendlines derived from the total data set and only the first ten rounds (as outlined above).

Below is the model I created to predict average bonus offered by round (click on the image to see a readable version).

With estimates of the average bonus offered to both HS and college draftees by round, I was able to estimate both what players likely turned down out of high school and what they were offered out of college. Because so few players went unsigned in the early rounds, I grouped players drafted in the 2nd and 3rd rounds, 4th and 5th, 6th through 8th, and 9th and 10th together to create larger sample sizes. Beyond the tenth round I analyzed groups of five rounds together (e.g., 21-25).

The results- which you can see below- were surprising. Except for players drafted in the 4th and 5th rounds (which was likely the product of a small sample size), every group of draftees received a higher mean bonus out of college. For some groups of players, the difference was significant. Players drafted in the 11th through 20th rounds received an average bonus of close to $250,000 more after going to college than what they were likely offered out of HS, an increase of close to 250 percent!

Click below for a full listing of the results.

Interestingly, despite receiving more money out of college, in every single group, players were drafted, on average, in a later round out of college than they were in HS. These two seemingly conflicting findings can be explained by the fact that the distribution of bonuses forms a power-law distribution. Players drafted in the first round receive millions of dollars, while college players drafted in rounds 8-10 often receive less than $100,000. The result of this very uneven distribution is that getting even one or two players into the first round more than compensates for many players going in the later rounds or even undrafted. Thus, while a player who decides not to sign out of high school receives, on a average, a higher bonus out of college, chances are that the player will actually go in a later round.

Below is a more visual way to interpret the data. The blue line is an estimate of the bonus a high school player would receive in a given round, and the red line is an estimate of what a player drafted in that round would receive after going to college.

By only looking at bonuses, the data suggests that going to college may be in a player’s best interest. However, as we’ll see over the next couple of days, there are other aspects that need to be considered.