Since it’s creation, the August 15th deadline for teams to sign drafted players has become one of the most important days on the baseball calendar. Interspersed among the headlines of which early picks have signed are the reports of players drafted in later rounds, sometimes even outside of the first ten rounds, signing six and seven-figure deals. These ‘over-slot’ signees are a bit of a mystery. Few outside of the scouting industry know much of anything about these players, and it’s difficult to judge the value of an over-slot signee relative to a team’s other draft picks. Is the 12th rounder who signed for $500,000 a better prospect than the third rounder who signed for $400,000? What type of premium do teams pay to sign players to over-slot deals? How much does it cost to sign a ‘second-round’ talent late in the draft? Can it even be done? These are the questions I sought to shed light on.
Unfortunately, reliable bonus data on players who sign late in the draft is tough to find before 2005, so our sample will be restricted to the 2005-2009 drafts. Because many of these players are just starting their major league careers or are still playing in the minors, we can’t know precisely how much value each draftee will ultimately provide, but using Erik Manning’s translation of Victor Wang’s research, we can come up with a fairly accurate approximation of each player’s value based on how they ranked on Baseball America’s Top 100 lists and in John Sickels’ team-by-team rankings.
For each draftee I looked at the bonus they received, the round they were picked, and the value corresponding to where they ranked on Baseball America’s Top 100 list or, if they did not make BA’s top 100 list, how John Sickels ranked them. I did not look at the prospect rankings for the most immediate season after each player was drafted, because I was concerned that a player’s initial draft position could bias how they were ranked (so the first prospect rankings I looked at for players drafted in 2005 was BA’s 2007 Top 100 list). I recorded players who had reached the majors before appearing on a relevant prospect list as being ranked 26th to 50th overall on BA’s Top 100 list. I gave extra weight to more recent prospect rankings, because a player’s ranking on a prospect list is likely a better reflection of that player’s true ability as he gets closer to the majors and develops a track record of performance. For players in the sample who appeared on four different prospect lists, I rated the fourth season as being worth 50 percent of the player’s ‘Total Value’ estimate, the third season as 25 percent, the second as 15 percent, and the fourth as only 10 percent. For players who appeared on three different prospect lists, I used a 50-30-20 weighting system. For players with only 2 seasons in the minors, I weighted the most recent as being worth 60 percent. Ideally, the weighting system would be consistent across groups, but for this type of broad-brush analysis, small differences in the how different seasons were weighted should not cause big problems.
The next step was to identify over-slot signees. I qualified players who signed for $150,000 or more after the tenth round as being an over-slot signing. For players drafted in the first ten rounds, I calculated the standard deviation of the bonuses awarded in each round and qualified players whose bonuses were 1 z-score above average as being an over-slot signing. This method worked for rounds 4-10, but in the first three rounds and the supplemental first round, the difference in the bonus received by players drafted at the beginning of the round is substantially higher than players drafted in the middle and end of those rounds. To overcome this problem, I looked at players drafted 10 spots ahead and behind each player in these rounds to determine their z-score.
To come up with an operational definition of the value of a typical player from each round, I repeated this process with players 1 z-score below average. Once I removed the over-slot signees and the money savers, I took the average of the remaining bonuses to find an estimate of the typical bonus awarded in each round.
Below is the typical bonus for each round and the average return that the ‘typical’ draftee of that round brought based on how players ranked on the relevant prospect lists:
* Note that these bonus figures are based on the 2005-2009 drafts, to apply these findings to current drafts you would need to account for any inflation bonuses have undergone over the past two drafts.
With less than ideal sample sizes, there were no statistically significant differences between the return on players signed to over-slot deals who were signed in the 11th through 20th rounds and those players signed to over-slot deals after the 20th round across bonus pools. With larger sample sizes, I would expect that players drafted earlier would perform better, but I’d wouldn’t predict that the magnitude of the difference between the two groups would be all that substantial as by the 11th round we’ve gotten into the thicker part of the bell curve of talent distribution, and the difference between being taken in the 15th and 25th round could be based on one team having a more optimistic view of a player’s signability than the rest of the industry.
With no vast differences in predicted outcomes between the two groups, I combined both to come up with an overall estimate of a player’s expected value based on the signing bonus the player received. After combining the two groups, I sent a best-fit line through the data to give an estimate of expected value based on the signing bonus a player signed for (R-squared .69), and compared this equation to the best fit line through the data for the ‘typical’ draftee from each of the first ten rounds (R-Squared > .99).
Below is the comparison of the two groups, with the final column representing the percent of the return from signing a player to that particular bonus in the later rounds as an over-slot signee compared to the return of signing a typical player from the round corresponding to that bonus (click for a clear view, returns listed in millions of dollars):
* Corresponding round in parentheses. Click for clear and enlarged view.
With a less than ideal sample size, these results need to be taken with a grain of salt, but the results provide an interesting foray into the topic. Somewhat unsurprisingly, it appears that teams do indeed have to pay a premium when signing over-slot draftees. From 2005-2009, a typical second round pick signed for just under $600,000 and was expected to provide 2.6 million dollars of value. The data suggests an over-slot signee who signed for $850,000 over the same time period would provide approximately the same value. To acquire a third round talent, the above data suggest that teams need to pay approximately $200,000 more than the bonus given to a player taken in the third round. Interestingly, as the singing bonus of the over-slot player increases, the gap between what that money buys you on the ‘over-slot’ market and the return on a similar bonus in the early rounds of the draft widens.
An important thing to note is that just because teams pay a premium to sign over-slot players in the later rounds, this analysis does not suggest that over-slot deals are bad investments for teams. On the contrary, over-slot signees still project to provide surplus value, teams just have to pay more to acquire the player, thus cutting into the surplus value the player provides.
In a subsequent post, we’ll look at over-slot signing in the first ten rounds, and some differences between the return on over-slot pitchers and position players.