clock menu more-arrow no yes

Filed under:

Valuing the Braves’ penalties

New, 120 comments

Spoiler alert: it’s kind of a big deal

MLB: Spring Training-Media Day Mark J. Rebilas-USA TODAY Sports

By now, Braves fans know that the Braves were smacked with the punishment stick last week, and that the blow was a heavy one. Eric Longenhagen at Fangraphs described the punishment as “harsh.” Our own Eric Cole described it as “brutal.” I’m sure one can find other sources with similar unkind words about the punishment. It was a punishment, after all -- and I think intended to be punitive more than corrective or instructive.

There are a few questions that the whole penalty thing has raised in my mind, but I think the most important one has been the overall value of the penalty. In part, that’s because it seems like no one has really taken the initiative on valuing it. When there’s a trade, or a signing, you hear all sorts of commentary on the quality of the move, the value it provides, and so on. Yet, here we are, with the statement from the Commissioner’s Office about a week in the rearview mirror, and I haven’t seen anything trying to affix any kind of value to the punishment as a whole.

Part of the reason I think valuation of the punishment is interesting is because I would have actually preferred that the punishment be done in “equivalent dollar” terms than anything else; it would have felt far cleaner than that to me. But, that’s clearly not MLB’s style, for better or worse, so that ship didn’t just sail -- it never left the dock. In any case, I can see why no one has really tried to value the punishment yet: it’s really hard to do. To that end, the various thoughts below are less coherent than I would like, and I’m not sure how additive some of the different components are. In the end, though, you can take the different parameters and come up with your own answer. I think you’ll find that the penalties are substantial, but not crippling.

Before anything else, a quick disclaimer that I understand there’s a continued temptation to discuss the fairness or appropriateness of the penalties. That’s fine, but that’s not really what I’m trying to do here. I have my own separate thoughts on that, but the valuation presented here is meant to be quite separate, and does not challenge the actual scope or depth of the penalties on their face.

First, let’s lay out the penalties themselves:

  • The Braves lost 12 players, as follows:
  • The Braves are limited to making international free agent signings of $10,000 or less in the 2019-2020 signing period. They also have been knocked down to essentially zero “pool” for this period, so it’s not like they can trade their slot values for that pool to other teams.
  • The Braves have had their “pool” for the 2020-2021 signing period halved.
  • Loss of a third-round pick in the 2018 amateur (June, Rule 4) draft.

There were also other penalties that negated the Braves’ ability to sign Robert Puason when he became eligible, and voided the potential deal the Braves could have signed with Ji-Hwan Bae. I’m not directly considering these in this analysis, because they are essentially a voiding of things that hadn’t yet happened. The effects of these two negations are also likely to be marginal.

Let’s go through these, one-by-one.


Loss of Signing Bonus

This one is actually pretty simple. You can see from the table above that the Braves spent $16.1 million on signing bonuses for players that are no longer in the organization. The penalty, therefore, effectively stripped the team from being able to use that $16.1 million for any other purpose.

In addition, the tax also needs to be considered. The Braves had a 2016-2017 IFA pool amount of about $4.8 million. (Thanks to @KnockahomaNTN for correcting me, I have no idea how I previously got the much higher original pool number but clearly I added wrong.) They ended up spending about $15.2 million in that period. Since the tax essentially doubles the cost of any overage on top of the pool amount, the tax was an extra $10.4 million ($15.2 million less $4.8 million = $10.4 million).

  • Therefore, the total monetary outlay the Braves lost was $26.5 million ($16.1 million plus $10.4 million).

Note: I found somewhat different signing bonus numbers from what @KnockahomaNTN did, so my numbers differ from his a bit; he had $26.9 million. Not a huge difference, and it’s highly likely that I’m wrong and he isn’t, so feel free to adjust upward by $400,000 if you think it matters. It rounds to $27 million, anyway.

Loss of Player Value

This is where it starts getting complex, because many of the removed players were essentially unknowns with very high risk, given their age and distance from the majors.

Kevin Maitan, the crown jewel of the class, though, has been highly-ranked, and therefore is easier to value.

  • Based on the prospect valuations done by Point of Pittsburgh, Maitan could be worth $38 million if you think he’s a Top 50 prospect in MLB; or $22 million if you think he’s only a Top 75 prospect.
  • Before the 2017 season, Fangraphs had Maitan as a 55 FV, which they translated to a $38 million surplus value. At midseason, he had slipped to 67th, but it’s unclear whether his FV slipped, and the Fangraphs’ valuations are based on FV rather than rank. So, you could say he would still be valued at $38 million, or you could say he fell into the next “bucket” of value, which would be a 50 FV, or $20 million in surplus value for a position player.
  • The table below summarizes my thoughts on Maitan’s value -- the “medium” reflects an average of $20 million, $22 million, and two instances of $38 million that are captured in the bullets above. Note that this is surplus value, which is meant to be comparable to the straight dollar figures used elsewhere in this analysis. It is not really a measure of Maitan’s expected production, but rather a measure of his expected production after accounting for any salary owed him. (If this concept is foreign to you, it can be explained in the comments, but it’s a pretty standard way of thinking about baseball value.)

Note: As people have pointed out to me, the consensus on Maitan has become more negative over time. To that end, perhaps the best “medium” value is actually $20 million, or at least closer to $20 million than $30 million. In any case, you can pick and choose from a range of values here. The fact that you can quibble with how exactly to value the loss of Maitan is very separate from the fact that regardless which value in this range you pick, his removal from the system is a very substantial chunk of the overall penalty.

The other players are where it starts to get more complex. There are 11 of them, and I don’t really know much of anything about them as a group. None of them made a Baseball America or Fangraphs Top 100 list, and did not appear to make any “honorable mentions” or fringe lists that I could locate, either. Things like the Point of Pittsburgh article focus only the top 100 prospects in MLB, so this poses a challenge to us. What’s more, though, is that in preseason aggregate analysis of prospect rankings that examined 219 total prospects by Ben Markham of Viva El Birdos, none of the names came up. So perhaps these guys aren’t even “Top 200” prospects, but somewhere below that. Where specifically below that? I don’t really know.

I don’t have any great rules of thumb here. But, some variants are below. I don’t have a real preference here, they’re just thoughts:

  • Use the value for Top 100 guys. This is essentially $18.1 million per Point of Pittsburgh, or about $12 million per Fangraphs.
  • Use half of the value for Top 100 guys, meaning something like $6 million to $9 million.
  • Use one third of value for Top 100 guys, under the assumption that these guys aren’t even making it to the Top 200 so they really aren’t that valuable. That indicates something like $4 million to $6 million.
  • Map a logarithmic function through the Point of Pittsburgh values for the top 100 prospects, and see what that function predicts for “top X” prospects. This function predicts about $8 million for the 150th-ranked prospect, about $3 million for the 200th-ranked prospect, and about $2 million for the 219th-ranked prospect.

The table below tries to aggregate these. The low value is the lowest indicated above ($2 million, logarithmic prediction for 219th-ranked prospect), the high value is the $18.1 million for a Top 100 player which is the highest value indicated above, and the average is simply the average of the different values here.

To be honest, I actually think the $8 million figure, and the $91 million in total, is a bit high, because there are 12 prospects here, and all 12 aren’t running the gamut from “fringe top 100” to “no value.” So, perhaps something akin to $50 million or $60 million for these prospects is more appropriate? Anyway, you have the tools to do your own estimations given the numbers here, so it’s easy to disagree and slap your own estimate on the surplus value for these 12 lost players.

Loss of IFA Market Participation

This is by far the trickiest piece. Despite the relative importance of the IFA market, there does not appear to have been (or, I have not been able to find, anyway) any kind of comprehensive, or even partial, attempt to value participation in the IFA market as a whole. What I’ve done here is very speculative, and I’m actually not really all too sure how to translate it into major league surplus value. But, bear with me, and see if the results make sense to you.

The first and most obvious way to measure the impact of IFA market participation would be to simply total up the number of IFAs an average team signs, and assign them some value, or range of values, like done for the non-Maitan prospects above. But, believe it or not, I could not find any such list anywhere. There are lists published by Baseball America of July 2 signing actions, but these do not appear to capture the full suite of signings over the course of the entire period, and so would seem to be substantial underestimates. There’s also a “complete” transactions roster for each day of the year on Baseball-Reference.com, but this appears to be limited to major leaguers or players who eventually make the majors, because you can find a ton of IFA deals in past years on these pages, but almost no deals from the last few signing periods, presumably because players in the latter have not yet graduated. I still think this is perhaps a better, or if not better, more consistent approach. But without the data, I can’t apply it.

Instead, I chose to leverage something else at Baseball-Reference, which provides a handy list of WAR (albeit, bWAR, but still WAR) per player per season, for all players. It provides this in the same table as how the player’s team acquired that player, one of the choices being “amateur free agent.” Bingo… sort of. There are a few obvious caveats to this approach:

  1. Foreign professionals, including those from Japan, are counted as “amateur free agents” here, regardless of whether or not they are subject to signing pools.
  2. One of the other “acquired” categories is “traded.” While free agency is its own category and therefore removes the team control aspect from complicating this approach, players under team control who get traded would therefore not be captured under any examination of the “amateur free agent” category. The same goes for players acquired via waivers, or the Rule 5 draft; in other words, the “amateur free agent” category captures only contributions by players made to the teams that signed them. In order to adjust for this, it’s important to consider not “amateur free agent WAR vs. total WAR” but “amateur free agent WAR vs. total non-involuntary team change WAR” -- i.e., amateur free agent WAR divided by the sum of [amateur draft WAR, amateur free agent WAR, and free agent WAR].

The first caveat above definitely results in an overstatement of the total WAR contribution from IFAs. The second adds uncertainty, but I’m not sure it’s a particularly biased uncertainty. Sure, some players are excluded, but I don’t think IFAs are more likely to get traded than other players, and IFAs that ink extensions with their original signing teams actually get overrepresented with no easy way to filter them out. So, it gives us some baseline, albeit one that’s probably too high, especially as “foreign professionals” are likely to be far more productive than IFAs subject to the pool/slot system.

I’m not posting data here, because there’s too much of it. I will gladly post any of these data in the comments or provide a link to my data if there’s interest, but in the interest of keeping this streamlined and focused on answers, I’m choosing not to discuss other interesting things about the data here.

For the 2010-2017 period (eight total seasons), the WAR breakdown across baseball was as follows:

Now, as noted, we only really care about the orange-shaded components, because the “traded” chunk is hefty. The math for the contribution of amateur free agents to those components is eight percent divided by 65 percent, or about 12 percent. This suggests that in a given season, 12 percent of total production comes from IFAs.

WAR is predicated on the baseline of 1,000 WAR up for grabs in a season. If 12 percent of that is accrued by amateur free agents, that means 120 WAR, in total, accrues to them. Given 30 teams (I guess you can exclude the Orioles if you want?), that comes out to 4 WAR per team. As a quick sanity check, I also did this same calculation on a team-by-team basis; the median for a team is about 11 percent. The Braves have actually been higher historically than other teams - while they’re pretty much dead average for WAR from position players accruing to IFAs, they’re much higher for pitchers, because of Julio Teheran’s presence in much of the 2010-2017 period, and the lack of substantial pitching production from the rest of the roster in that time (Teheran is first among Braves for RA9-WAR in that period, and the second-place name, Tim Hudson, was acquired via trade and thus not included in this calculus.) So, as an alternative, you could say that a team more dependent on IFAs could be like the Braves and actually derive something like 7 WAR/season (22 divided by 12 multiplied by 4 WAR) from the IFA market. But, perhaps you shouldn’t think of it that way, because remember that the 4 WAR figure includes “foreign professionals.” Up to you.

Here’s the real challenge, though. It’s tempting to say that since the Braves are effectively banned from IFA for one period, and operate at half of their pool for another period, they’re losing 1.5 seasons of IFA participation, and therefore 1.5 “units” of 4 WAR each. That was my first instinct. But, I’m not sure that’s correct. That would require an assumption that signing periods are directly linked to annual production, and I don’t know if that’s sound. We could make it sound by just making that an assumption by fiat. I still don’t know if that’s accurate, but I guess we could keep it in mind.

Another consideration is that IFA signees actually lead to multiple years of production. In this way, IFA acquisition always “stacks up” -- if we assume that an average IFA that makes the majors is productive for three years, then that 4 WAR total is made up of some combination of IFAs signed across three periods. I actually am not too sure what to do with this. I can tell you that over the 2010-2017 period, IFAs tended to contribute to a given team for about three seasons on average (your average IFA is not a long-tenured regular, in other words). I’m not sure what to do with that, either. So maybe we stick with the “one signing period equals one year of IFA production” assumption.

That 4 WAR per season can come at any stage of team control. If it comes during the first few seasons, where the players are being paid league minimum, the value is substantial: 4 WAR at $8 million per WAR equals $32 million of production, offset by just about $0.675 million in league minimum salary. However, arbitration salaries eat into that pretty handily. Using the Point of Pittsburgh arbitration salary estimator rules of thumb, you get the following:

Under the assumption that the Braves lost 1.5 such seasons, then, you could say that the estimated effect is about $35.7 million in surplus value. That’s comparable to losing a second (high-end estimation, anyway) Maitan. It’s certainly not so off that it makes this calculation, or our other calculations, look particularly daft. That’s a good sign.

The other consideration here is that this provides an alternative way to value the loss of the big 2016 signings. The in-season average of about $23.7 million is pretty much identical to the “low” estimate of $24 million, assuming each of the 12 non-Maitan players is worth about $2 million in surplus value. That’s kind of neat. No idea if it’s even remotely right, but it’s neat.

Note that this surplus value is coming well in the future, if at all. Therefore, in today’s dollars, you may want to discount it. But, the choice of a discount rate is very fraught, as is the decision about how many years to discount it by. I won’t elaborate too much here, but these are judgment calls that substantially change the result. I am leaving it non-discounted for now, in part because there’s a lot of uncertainty about what the Braves’ window really is, and whether it’s actually fair to discount this value based on a standard time value of money assumption when the Braves’ contention window may not be “right now.”

Third-Round Pick in the June 2018 Amateur Draft

This one is easy, and it’s small, so it’s actually not a huge deal. The net value (surplus value) of a third-round pick is under $3 million, per the great work done here by Andrew Ball at Beyond the Box Score. Maybe it’s a little higher as teams have improved their amateur domestic scouting, but it’s not worth worrying about.


Putting It All Together

Again, this is all highly speculative. I’m sure this needs revision in both assumptions and parameters. But here’s where we are.

In aggregate, I think the low-end number is probably the safest to use. But still, that’s a pretty large number. That’s basically the Braves’ payroll when they’re not trying to compete, like over the past few seasons. That could buy you about 13 wins on the free agent market in a given season, transforming your .500 team into essentially a sure-fire playoff team. If the Braves gave every team an equal fraction of $110 million going into the next season, that would essentially be giving every team an extra $4 million to play with. It’s not much, but it’s every team, getting to do something the Braves can’t, at the Braves’ expense. It gets somewhat better if you think about this penalty getting cut up across multiple years, or chunks of it being discounted because the damage is going to appear in the future, but it is what it is.

In case contextualizing this still isn’t working for you, think of it this way. Matt Kemp, assuming he produces nothing at all going forward, is a net drain of about $38 million on the team. This penalty, per the table estimated above, is estimated to be at least 2.8 times as bad as Kemp’s remaining contract obligations. Would you want 2.8 more Kemp contracts to deal with? Because that might be akin to what the Braves are being saddled with as a result of this penalty.

Happy Holidays, I guess.