Baseball’s Middle Class Remains Embattled

One of the most remarkable news items of the offseason — is it still the offseason? — emerged towards the end of last week, as the calendar melted from February to March.

Neil Walker has produced seven straight seasons of 2.1 WAR or greater and ranks 61st in position-player WAR (11.7) since 2014. He also remains among the notable unsigned free agents in this slowest of offseasons. Last week, the Royals offered Walker what has to be the most insulting of deals for an unemployed player of his stature: a non-roster invitation to spring training. Our depth charts project Walker for a 2.6 WAR campaign and a 111 wRC+ — if he finds a job. Jay is trying to find him. He could help a number of teams. That he’s not employed as of March 6 is somewhat remarkable.

With Yu Darvish and Eric Hosmer having finally signed for relatively handsome sums, with spring camps opening and games being played, some of the claims of collusion and talk of the direst of offseasons for the MLBPA have quieted.

Craig Edwards found that the six free agents predicted to sign deals greater than $50 million have exceeded the FanGraphs’ crowd estimates by 6.2% this offseason. On the one hand, the crowd typically underestimates those deals; on the other, this is roughly the amount by which the crowd underestimates them. One could conclude, then, that the result has actually been typical.

And, in a sense, it has been a typical offseason — for the top free agents. But the top free agents are always going to do well.

We can debate whether Hosmer is a “top” free agent, but most outlets placed him within the top 10 of their offseason rankings and called for him to earn a nine-figure contract. Next winter, Bryce Harper, Manny Machado, and probably Clayton Kershaw are going to be immune from any devaluation of free agency that might be occurring.

For me, Walker is a symbol of this offseason. He is the face of a trend I examined last January — namely, that the middle class of free agents (which also represents the vast majority of free agents) is the demographic being pinched. The middle class of free agent is still a productive major-league player. But he is typically on the wrong side of 30 — Walker is 32 — at a time when front offices prefer younger talent and maximum payroll efficiency.

Walker, once the Pirates’ player rep, did not seem to foresee this problem — like much of the MLBPA — when he spoke to this then newspapermen back in 2015 about the players’ share of revenue perhaps showing indications of decline.

Are players concerned by the current revenue split? Not yet, said Pirates union rep, Neil Walker.

“We haven’t had much discussion about it,” he said, “and frankly at this point, we don’t see it as much of a concern.”

Just over a year ago, I examined how three groups of players were faring: the “wealthy,” the “middle class,” and essentially the pre-arb player. I defined as “wealthy” those players earning at least three times the average player salary for that year ($4.5 million last season). I considered the “minimum” earners to be compensated within 10% of the league minimum to cover most players with pre-arbitration status. All other players I combined into the “middle class” group. I employed USA Today’s salary database of Opening Day payrolls as the research source.

What I found is that the middle-class group had shrunk from including 68% of players in 1995, to 60% in 2006, to 57% in 2011, to 52% in 2016, while the “minimum” player population grew from 18% in 1995 to 39% in 2016.

Applying the same methodology to last season, the middle class shrunk by another point, to 51.0%, while cheap labor inched up to nearly 40% of the MLB workforce. We’re getting closer to a future — one in which the rest of the American workforce has already found itself — where the majority of the players will not occupy the middle tier of compensation.

The middle class of free agent has been particularly embattled this offseason.

Craig found that the crowdsource estimates in recent offseasons (2013-17) have tended to overestimate the contracts for the middle tier of free agents. For instance, for projected contracts between $10-$40 million guaranteed, the forecasts were 19.3% above the actual deals signed.

For this offseason, I looked at all the free-agent deals predicted to be worth less than $45 million among our top 50 free agents. (Players who beat the crowdsourced estimates are highlighted.)

Crowd vs. Actual on Contracts $45 Mil. and Less
Top 50 Rank Player Crowd Actual Difference
12 CC Sabathia $26.0 $10.0 -$16.0
13 Todd Frazier $42.0 $17.0 -$25.0
14 Jay Bruce $39.0 $39.0 $0.0
16 Brandon Morrow $18.0 $21.0 $3.0
19 Addison Reed $27.0 $16.8 -$10.3
20 Jarrod Dyson $16.0 $7.5 -$8.5
21 Eduardo Nunez $33.0 $8.0 -$25.0
24 Bryan Shaw $21.0 $27.0 $6.0
25 Logan Morrison $20.0 $6.5 -$13.5
26 Joe Smith $12.0 $15.0 $3.0
27 Anthony Swarzak $14.0 $14.0 $0.0
28 Jaime Garcia $36.0 $10.0 -$26.0
29 Yonder Alonso $20.0 $16.0 -$4.0
30 Jason Vargas $20.0 $16.0 -$4.0
31 Juan Nicasio $16.0 $17.0 $1.0
32 Pat Neshek $16.0 $16.3 $0.3
34 Carlos Gomez $24.0 $4.0 -$20.0
35 Andrew Cashner $22.0 $16.0 -$6.0
36 Tony Watson $12.0 $8.5 -$3.5
37 Michael Pineda $16.0 $10.0 -$6.0
38 Jake McGee $24.0 $27.0 $3.0
39 Tommy Hunter $12.0 $18.0 $6.0
40 Steve Cishek $12.0 $13.0 $1.0
41 Jhoulys Chacin $20.0 $15.5 -$4.5
42 Lucas Duda $18.0 $3.5 -$14.5
43 Alex Avila $16.0 $8.3 -$7.8
44 Cameron Maybin $16.0 $3.5 -$12.5
46 Howie Kendrick $16.0 $7.0 -$9.0
47 Brandon Kintzler $14.0 $10.0 -$4.0
48 Welington Castillo $18.0 $15.0 -$3.0
49 Doug Fister $8.0 $4.0 -$4.0
50 Jon Jay $8.0 $3.0 -$5.0
Totals $632.0 $423.3 -$208.8

The crowd predicted this middle tier of free agents — those who have signed — would earn $632 million guaranteed dollars on the open market this winter. Those players, meanwhile, have actually signed deals for just over $423 million guaranteed. Overall, the middle tier of free agents has signed for roughly 33% less than the crowd estimated

Moreover, mid-tier free agents like Walker, Jeremy Hellickson, and Jonathan Lucroy remained unsigned.

The top of the class did fairly well and is expected do very well next year, but the middle class remains embattled.

Darvish and Hosmer got paid. It is a player like Walker, however — one on the wrong side of 30, one who is useful but deemed highly fungible — that is affected. It is a player like Walker, who is stuck in the middle.

We hoped you liked reading Baseball’s Middle Class Remains Embattled by Travis Sawchik!

Please support FanGraphs by becoming a member. We publish thousands of articles a year, host multiple podcasts, and have an ever growing database of baseball stats.

FanGraphs does not have a paywall. With your membership, we can continue to offer the content you've come to rely on and add to our unique baseball coverage.

Support FanGraphs

A Cleveland native, FanGraphs writer Travis Sawchik is the author of the New York Times bestselling book, Big Data Baseball. He also contributes to The Athletic Cleveland, and has written for the Pittsburgh Tribune-Review, among other outlets. Follow him on Twitter @Travis_Sawchik.

newest oldest most voted

Interesting article Travis.

Couple of possible takeaways:

1) The contract values did not include any value for various opt-outs which have considerable value to the player. The actual value and the crowd expectations did not include and do not value these.

2) Perhaps going forward FG authors and articles that use $X per WAR will need to model/discount for where a player lies in the player spectrum? A “high end” player would be $X per WAR and a “middle-class” player would be $.7X per WAR?


Or better yet, we could stop using $/WAR as a fixed entity and instead ask: “What is $/WAR in this year’s free agent market?” And then update the number based on who signs and their Depth Charts projections. (you could also stratify or otherwise test for non-linearities too)


Its important to remember that $/WAR is descriptive and not predictive, its just what teams have spent per WAR in the past, not necessarily what they think a WAR is worth. That figure is gonna get inflated by bad contracts like, say, Pujols, we just don’t have any way to determine what the optimal $/WAR should be.

It isn’t intended to be used as a predictive tool for value, but more as a predictive tool for market behavior.

We don’t know if this means the market is getting more efficient or if the market is being artificially deflated.


Its actually what the value of an extra win is to the average team. FA can command that and not just 50% because their salary is subsidized by young stars making far less than that and that there is free and competitive market for their services (questionable today)

On average, players of all classes are paid collectively at half the $/WAR stated.


I agree, but if you want to ask what the market rate is at a given time, perhaps it makes sense to see how much teams are actually paying for expected production at a given time? Instead of continually throwing out $9million/win because there are enough guys like Pujols who unexpectedly sucked? (although I think he’s out of the time-frame Swartz uses now, and other lousy contracts are in instead).

It’s not a very good predictive tool for market behavior unless teams believe that number, and there’s not any real evidence that they do based on the projections (rather than the outcomes).