Production of ‘Slight’ Over-Slot Signees

Earlier this year, The New York Times’ Economix blog had a discussion of a study conducted by the economists Stacy Dale and Alan Krueger (Princeton). The researchers looked at the value students derive from what they termed ‘elite’ colleges. Their study’s twist was that instead of only looking at the schools students attended, the researchers also looked at the schools students applied to. Interestingly, the researchers found that the average SAT score at the most selective school a student applied to was more predictive of that student’s future earnings than the average SAT score at the school the student attended. The researchers hypothesized that applying to selective schools may be an important indicator of future success since the decision to apply to a selective school may be an indication that the student has characteristics that will help them later in life beyond measures such as high school GPA, SAT score, etc… which typically factor heavily into college admissions decision.

I bring up this piece of research because as I was analyzing the data for the return on over-slot signees I was struck by the number of players who signed for just a few thousand dollars over MLB’s slot recommendation but with the benefit of hindsight would have been drafted much higher. Mike Stanton signed for $55,000 above slot in the second round of the 2007 draft; Brett Lawrie signed for $20,000 above slot in 2008; Danny Duffy signed for a mere 500 dollars above slot in 2007; Brandon Belt signed for $25,000 above slot last year, and the list goes on.

Signing for even a few thousand dollars above MLB’s slot recommendation is a big deal. Scouting directors and front offices are loath to exceed MLB’s slot recommendation whenever they can avoid doing so, as going over slot usually draws the ire of the Commissioner’s office. On the other side of the negotiating table, securing over-slot deals for a client can be a boon for agents, as they can use each successful negotiation as a tool to recruit players in next year’s draft class. All told, a player asking for even a couple thousand dollars more than slot can cause negotiations to go close to the August deadline. With such an emphasis on the slot cap, I wondered if players who went to the effort of holding out for an extra couple thousand dollars may outperform the typical draftee from that round, above and beyond what we would expect from the elevated bonus total.

To investigate, I took players who signed above-slot deals, but whose bonuses did not exceed 118 percent of MLB’s slot recommendation and compared these players’ expected value to the expected value of the average player who signed for that amount. I excluded players whose bonus was over $80,000 more than MLB’s slot recommendation regardless of whether the player’s bonus was less than 18% more than slot. Unfortunately, I only had detailed slot data dating back to 2007. Still, within these parameters, I had a sample of 38 players who fit the criteria.

Based on how players of the same draft position were ranked on Baseball America and John Sickels prospect lists, we would expect these 38 players to be worth just under 124 million dollars. Surprisingly, these ‘slightly over-slot’ signees were expected to provide nearly 190 million dollars worth of value. The increase in expected value was statistically significant. Making these findings even more surprising, the slight over-slot group actually received less in total bonus money than the comparison group. I compared second rounders to second rounds, early first rounders to early first rounders, etc… and several players slotted in at the back end of these comparison groups (ex- the Giants took Wendell Fairley with the 29th overall pick in 2007 and gave him a 1 million dollar bonus, close to $80,000 over slot, but because I compared him to other players in the back third of the first round, the group he was compared to received, on average, a bonus worth 1.3 million dollars. So, Fairley’s bonus was actually $300,000 less than the group he was compared to).

So what to make of these results? Maybe players who believe they are worth more than MLB’s slot recommendation have characteristics that lend themselves to having a productive baseball careers beyond what statistics and talent evaluators capture.

If we believed that this ‘slightly above-slot’ effect was real, would we expect players who hold out from signing completely to show the same increased level of production? That belief doesn’t seem to necessarily square with a quick a priori look at some of the most notable players who have refused to sign and gone to college or played independent ball.

Maybe there is a certain sweet spot where a player has the optimal combination of a high sense of self worth as well as a strong desire to play professional baseball. Maybe this finding is just a statistical blip and with more data we would see this finding go away. There are a number of possible explanations. And even if with more data the effect persisted, it’s tough to find a real takeaway message. If teams began to act on this information pretty soon every player would be asking for a couple thousand dollars more than slot. In some, I think this is a cool finding, albeit one without much in the way of a practical implication.




Print This Post





13 Responses to “Production of ‘Slight’ Over-Slot Signees”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Statliners says:

    Do you think this could be caused by a player who is demanding more money than other players? An extreme example would be Andrew Miller in the 2006 draft. He was considered by many to be the best player in the draft, but slipped to the Tigers at #6 because most of the top 5 teams were not willing to pay him the money Miller was demanding (about 2 mil/year and a 3.6 mil signing bonus). Also, maybe some drafts the talent level is considered more “deep” causing more valuable players to be taken later.

    Vote -1 Vote +1

    • DJG2111 says:

      This seems like it would have a statistically significant effect. I would think a lot of players who sign above slot slide down the draft due to signability issues. I seem to remember some making it clear that they didn’t want to sign with the teams that had the highest draft picks (perceived as the worst teams). How do these players signings compare to the slot recommendation for their projected draft positions?

      Vote -1 Vote +1

  2. Dan in Philly says:

    Did you remove the 4 players you mention above who inspired the study from you consideration to avoid the obvious fallacy including them would have with such a small sample size?

    Vote -1 Vote +1

    • Dan in Philly says:

      Also, did you control the study for outliers which would make such a (again) small sample size swing so wildly? Compare any small sample and you might get such a finding, which would be a fun april fools study to do one day. Did you know that left handed infielder draftees who played highschool ball in teams from cities containing the letter “m” signifigantly outperform other draftees? How will this knowledge adjust drafting strategies for future teams???

      Vote -1 Vote +1

      • Maybe, but says:

        I wouldn’t necessarily exclude those players for reasons of selection bias. If the author first defined what constitutes a “slightly over-slot” bonus and then applied that rule to every player in the draft regardless of how each player eventually performed, then those four players above should be included without an issue. What would be problematic is if the 118% cut-off was created with players like the four initially mentioned in mind (it does seem kind of subjective, no?), in which case yes, there would be some cherry-picking going on and a possible bias.

        Vote -1 Vote +1

      • Dan in Philly says:

        @Maybe, but
        There is a huge problem with including those players in the study. You already know the outcome of those players before selecting them! You already know those players have significantly outperformed their signing bonus, and therefore their performance will skew any study.

        Take for example the idea of studying all players based on their initials. Since you already know a certain first baseman in St. Louis will be included in any such study, you can say “I randomly pulled all players with the initials “AP” and learned they are totally outperforming the rest of the league!” Well, duh!

        Since the inspiration for the study is included in the results, the findings are not meaningful.

        Vote -1 Vote +1

    • Jack says:

      Why would you exclude them?

      Vote -1 Vote +1

    • No, I did not. But I came up with the parameters (max 118% of slot and max $80,000 over slot, whichever comes first, before finding those players. Although I can see how that is confusing the way that I wrote it in the article.

      The first ‘slightly’ over-slot player that caught my attention was Danny Espinosa, and it jogged my memory of Dale and Krueger’s study. When I saw Jason Knapp signed for a ‘slight’ over-slot bonus just a few spots before that, I ran the query using the above parameters. Now, I didn’t remove Espinosa and Knapp from the sample. Maybe I should have. But I don’t think anyone but the biggest methodological nitpickers would dub the inclusion of Espinosa and Knapp grounds for a selection bias.

      Vote -1 Vote +1

  3. AustinRHL says:

    I think Statliner’s hypotheses are more plausible than yours, Reed. The other possibility I’d suggest is that if teams are willing to acquiesce to a slightly over-slot demand, then it’s likely that the player has impressed the team or improved his stock beyond what the team determined when they drafted him.

    Vote -1 Vote +1

    • I think the alternative you suggest is more plausible than Statliner’s hypothesis, although that’s not to say that his suggestion couldn’t have an effect.

      Guys who let it be known that they are seeking big money, don’t sign for bonuses within the parameters I set up (118% of slot and within $80,000 of slot). Statliner brings up Andrew Miller, but he signed for well over a million more than slot. Now, it’s certainly possible that some players who come into the draft seeking much more than slot end up settling for a small amount over slot, but I think it’s rare many of those cases would make it into the sample given the fairly strict parameters.

      I think your suggestion that a lot of the players in the sample would be guys who have made strides over the summer, prompting teams to be willing to invest more money in them, is a good one and is likely at work to some extent. The one qualification I’d like to point out is that a sizable number of draftees, particularly players drafted out of college, don’t play in summer leagues that summer, so it’s not like there are scouts bearing down on all of these players over the summer.

      As I said in the article, there are a number of possible explanations. The finding could just be a statistical blip due to a small sample size. But I think it’s cool to try to think about some of the ways psychological traits could influence performance.

      Vote -1 Vote +1

  4. Mr Punch says:

    Seems to me that Statliner’s hypothesis applies primarily to draftees who sign for well above slot – more than 18%. For the “slight” group, it might be that the teams just think they’re worth the money. (We don’t know if they’re the only ones who asked for more than slot, do we?) Teams draft when they have a pick, not when the slot money is right for a player they want, and must often (after the very top of Round 1) choose players they like better than some of those taken higher by other clubs. They may well be thinking, say, “This second-rounder is worth more than some of the guys who went in the lower half of the first round.”

    Vote -1 Vote +1

  5. Franco says:

    I always hear that certain such and such team don’t go overslot because they’ll catch a lashing from the MLB. What does that even mean? Literally it’s just Selig leaving you a nasty voicemail or is it something tangible?

    Vote -1 Vote +1

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>