Do the best free-agent deals come to teams that wait?

ByBradford Doolittle ESPN logo
Thursday, January 18, 2018

Everyone is getting nervous. By everyone, I mean those with a vested interested in the baseball free-agent market. Agents. Teams. Players. Fans. And, yes, writers, too. After all, the bugle sounds for spring training in less than a month, and we're not accustomed to this much late hot stove drama.

Day by day, we wait for the dam to break. Meanwhile, the theories for the free-agent freeze-out abound and somehow grow more ominous by the day. Yet, theory is theory: We don't really know what's at the root of all this. We also don't know if this is a one-year blip or the new reality for baseball free agency.

With so much uncertainty in the air, let's try to veer the conversation toward objective territory. Here's a question that we can at least approach with a little data analysis: Does a slow market favor the teams or the players/agents?

It's a simple question. Answering it is not simple. Here's what I did:

-- I compiled a database of free-agent signings that encompass the wild-card era (since 1994), based on data from I picked that data source because most of the signings listed there include the reported length of each contract, though there are gaps in the information. I plugged as many of those gaps as I could for multiyear deals. Each transaction is dated and flagged for contract length.

-- Not included: Option years, contract extensions and international signings. The aim is to capture deals signed during the winter free-agent process, so in-season signings also were omitted. I ended up with 2,130 signings in the database. Each signing that occurred in October, November or December was flagged as "early," while signings dated in January, February and March were marked as "late."

-- The signings database was merged with yearly WAR data from and salary data from the Lahman Baseball Database. There were a few holes in the salary data, which largely appeared to be related to deals for the yearly minimum. Minimum salary figures were autoplugged into those gaps. Also, the Lahman data did not include 2017 salaries, so I added them manually using information from Baseball Prospectus.

-- Finally, I calculated the annual surplus value of each season of each contract, based on research from Matt Swartz at FanGraphs. In a nutshell, the concept of surplus value is based on an estimate of the cost to purchase a win (or one WAR) via free agency for a given season. With that number in hand, you can then see how much value a player produced -- that season's cost of a win times his actual WAR. You then subtract his salary and the remaining difference between value produced and value paid is what we refer to as surplus value, which I'll abbreviate as "SV" through the rest of this piece.

With all of these SV estimates at our disposal -- and these are indeed estimates -- we can address all sorts of questions about the free-agent marketplace. So let's do so with an impromptu Q&A, though in this case, the questioner and the answerer are both me.

1. Is it better to sign a player early or late?

It depends on how you define better, but in general, this seems like a case of good things coming to those teams who wait. In the aggregate, the clubs get better value with late signings. The average SV per contract season is $0.9 million for late signings compared to minus-$0.3 million for early signings. The average salary paid out is $5.97 million for early and $4.0 million for late.

Of note: While the value might be better for late signers, you do tend to get better players if you move early. The average annual WAR produced by late-signing teams is 0.81, while early signers get 0.98 WAR per contract season.

2. How does contract length play into this?

When it comes to the ultimate SV produced by a free-agent contract, length is everything. One-year deals account for nearly two-thirds of the signings in the study, and they collectively produce an average SV of $0.83 million. Two-year deals are break-even. Only two other contract lengths produce positive SVs: six-year and 10-year deals. We'll dive into those in a bit, though, keep in mind, both categories are of the small-sample variety.

Grouping the lengths together gives you this:

Again, the longer you go on a deal, the better the player you're signing, but your spending efficiency goes down accordingly.

3. What is the interplay between signing date and spending efficiency?

Despite what we've learned in these first two questions, it's not a simple choice between on-field value or spending efficiency. The longer you wait, the more you can generally narrow the gap between the two.

In all three length categories, you get better value by waiting until later in the signing period. On the shortest deals, you still get better players by moving early in terms of WAR. However, medium- and longer-duration deals combine better performance and better value if you sign them in the late period. Bingo!

4. So the lesson is to wait until January, then spend like crazy, right?

I wouldn't go that far. However, it is intriguing that teams appear to have had some success by waiting for prices to come down before bringing in players on those medium-duration contracts for four-to-six years of length.

Sample-size caveats apply, of course: 93 percent of all the signings in the study fell between one and three years of length. The long-term deals always generate the headlines, but the bulk of free agency consists of shorter, more modest deals. Given the narrower projection windows needed for those short-term pacts, it's not at all surprising that teams are good at valuing those deals appropriately.

That said, let's consider the 127 deals that fell into the medium-duration category. Question No. 2 told us that in the aggregate, these deals produce negative SV. That's because the 108 signings that occurred during the early period produced an average SV of minus-$1.3 million. However, the 19 medium-length signings from the late period had a positive average SV of $0.6 million, and those 19 players produced a higher average WAR (2.1 to 1.6).

The effect is even more stark for longer-duration contracts. The average SV produced by the 16 early signings in the study of at least seven years in length was a whopping minus-$4.1 million. The names attached to these deals are cautionary tales: Albert Pujols, Alfonso Soriano, Carl Crawford, Barry Zito, Jason Giambi andJacoby Ellsbury.

Nevertheless, Question No. 3 tells us that long contracts signed after Jan. 1 have worked out exceedingly well. Those contracts produced an average SV of $5.2 million and 3.3 WAR per season. Which brings us to ...

5. Is there a Boras factor in play here?

You better believe it, but it's also somewhat of a contradiction. The data presented here tells us that by and large, teams get better value by waiting out the marketplace. Yet, Scott Boras has built his reputation by instructing his clients to wait out the market year after year. In theory, that would play into the teams' hands.

Are Boras' clients the exception to the general rule? Maybe. But, again, we're not talking about large enough buckets to make a definitive statement here.

There are just five late signings flagged in the study that resulted in contracts of at least seven years in duration. Early long contracts have resulted in an ugly minus-$4.1 million average SV figure. However, the five late megasigners noted above have produced the $5.2 average SV noted above. Their average WAR per season is the best performance of any subcategory.

But we're talking about just five contracts, and not all of them have worked out positively:

All of these contracts were negotiated by Boras, though he later had a falling out with Beltran. The numbers are skewed by the massively successful deal Scherzer signed with the Washington Nationals. In the three years before Scherzer signed with Washington, he had produced a total of 15.7 WAR. In the three years since then, he has produced 18.8. If only free agency always worked out like that.

It's important to note that four of these five deals are still in progress, and it's probable that the metrics used here will look worse once the contracts are finished. That includes Fielder's deal, which the Rangers, who will recoup insurance payments, and Tigers will both be paying for three more years even though he hasn't played since the 2016 season.

So, while Boras' clients here have done really well financially by heeding his advice, the teams that signed them have ended up holding a mixed bag. Even with Boras guys, teams have to be careful, just as they do with all long-term signings. That holds true whether it's before or after Jan. 1.

6. What does all this mean for the dangling free-agent class?

I wish I could say. On one hand, it doesn't bode well for the players' hopes of maxing out their dollars from free agency. Teams don't hold back on signings in order to wait for the market to go up -- they are waiting for the prices to drop. Yet so many of the top remaining free agents are Boras clients --Eric Hosmer, J.D. Martinez, Mike Moustakas, Jake Arrieta, etc. -- and history tells us that Boras gets what he wants.

However, we've never quite seen a marketplace like this one. It appears that there are more players left on the market than there are prime salary slots to fill. That suggests that most signings that take place going forward will be team-friendly. And they will eventually look good (for the teams) in the metrics applied here. At the same time, the sheer number of quality players left unsigned could result in a rare late-winter bidding war that flips the typical trends outlined here upside down.

We simply don't know how this odd winter is going to play out. What we do know is that teams have more exact estimates of fair value than I've come up with here. More than ever, they appear to be sticking to those valuations, and it's leading to some anxious days and rare drama for this portion of the calendar.

But the market has to heat up soon, right? After all, some of us have spring training trips to finalize and season previews to write.