Players with Abnormally Strong Walk Years Get Paid More

There is enough literature out there to debunk the theory that players generally play better in the year prior to free agency. Anecdotally speaking, this season, we saw top pitchers like Johnny Cueto, Zack Greinke, and David Price have solid years in line with their established levels. Jason Heyward and Justin Upton also produced seasons that resembled their career numbers. Meanwhile, Ian Desmond, Jeff Samardzija — and, to a lesser extent, Alex Gordon and Jordan Zimmermann — did not quite live up to prior years. The two biggest examples of players with out-of-the-ordinary walk years, Yoenis Cespedes and Chris Davis, remain unsigned into the middle of January. Out of the top 11 free agents, just two had abnormally strong walk years. Just because the walk-year performance is a myth, that doesn’t mean that players who do perform extraordinarily well receive less in the way of compensation than their more consistent counterparts.

Examining recent contracts, we can attempt to determine if those players who had big jumps in their walk-year performance were paid more than those with more consistently strong performances. Over the past ten offseasons, 39 hitters have received contracts in excess of $50 million. While a broader look at all free agents might reveal a few more interesting players, I set a floor to examine only those free agents who could have benefited substantially from big walk years, as well as similar contracts for those without the same leap in performance during the final season of their contracts.

To establish those players with big walk years, I performed a simple, Marcel-like calculation (weighting seasons by multipliers of 5, 4, and 3) of the WAR of a player’s three previous seasons to establish a base of expectations for their walk-year performance. Then, I looked at the player’s walk-year performance as a comparison. Due to survivorship bias, this group does not represent an appropriate dataset to debunk the walk-year myth, given that most players who would perform poorly (e.g. Ian Desmond) in their walk year have already been purposefully removed. It is interesting to note, however, that only 15 of 39 (38%) players had big jumps in their walk years. Eighteen players (46%) were fairly close to their expected WAR, while another six (15%) had down years (defined here as a 40% drop or worse) ahead of free agency. On average, the group of 39 players was just 0.7 WAR (18%) better in their walk year despite the survivorship bias issue.

Below is a chart with the players who exceeded their expected WAR by 50% or more in their walk years, including Yoenis Cespedes and Chris Davis. Players are sorted in order of percentage by which they exceeded their expected win totals (denoted as % Over Exp. WAR).

Abnormally High Walk Year WAR: 2006-2015
WAR FA-3 Year WAR FA-2 Year WAR FA-1 Year wAVG WAR Walk Year WAR Diff in WAR %Over Exp. WAR
Victor Martinez 2.1 0.0 0.8 0.9 4.3 3.4 401.0%
Nelson Cruz 1.3 1.1 1.3 1.2 3.7 2.5 200.0%
Yoenis Cespedes 2.9 2.4 3.3 2.9 6.7 3.8 131.0%
Adrian Beltre 2.9 3.9 2.1 2.9 6.4 3.5 120.7%
Adam Dunn 2.9 0.6 1.1 1.4 3.0 1.6 116.9%
Jose Reyes 5.9 0.7 2.5 2.8 5.9 3.2 114.5%
Jason Bay 5.2 -1.0 3.0 2.2 4.6 2.4 107.5%
Shin-Soo Choo 6.0 1.5 2.3 3.0 5.5 2.5 85.9%
Carl Crawford 3.3 2.7 5.9 4.2 7.7 3.5 84.1%
Gary Matthews 1.6 2.2 2.4 2.1 3.9 1.8 82.8%
Alfonso Soriano 5.1 2.2 2.4 3.0 5.4 2.4 79.5%
Chris Davis 2.1 7.0 0.8 3.2 5.6 2.4 75.5%
Russell Martin 2.4 2.0 4.1 3.0 5.0 2.0 68.1%
Aaron Rowand 5.8 3.8 1.2 3.2 5.4 2.2 67.9%
Jacoby Ellsbury -0.2 9.4 1.2 3.6 5.6 2.0 56.3%
Mark Teixeira 5.9 3.5 4.4 4.5 6.9 2.4 54.2%
Alex Rodriguez 6.6 9.1 3.8 6.3 9.6 3.3 53.2%

Given the nature of big free agents, it would be easy to suggest that teams ought to avoid paying big money to players who have huge walk years. On the other hand, that wisdom is relevant to all free agents, as many fail to return complete value for the signing team. Many of the deals above are still ongoing. The Adrian Beltre contract has been fantastic, while Nelson Cruz and Russell Martin have been solid deals thus far.

One easy way to determine if the members of the above group are being paid more than they would have been in the absence of their giant walk year, would be to simply add up the years and salaries and compare the totals to those produced by the player who didn’t produce unexpectedly strong walk years. The above group’s average contract went for 5.9 years and $109.5 million, for an average of $18.6 million per year. Removing the six walk-year underperformers, we are left with 18 players who were within expectations in their walk year. Those 18 players received contracts averaging 5.6 years and $102.8 million, for an average of $18.4 million per year. A slightly higher guarantee in a group of players this small could lead to the conclusion that the big walk year provides some extra benefit. However, this could also be mitigated by the fact that the big-walk-year group had an average WAR of 5.5 during the walk year, while the other group’s average WAR was 3.8 in the walk year. Given that number, it might be fair to say that the walk year is of no benefit.

The evidence so far would lead you to believe that players who have unexpectedly big seasons in their walk year do not see a benefit from that great year. However, we can delve a bit further. Just because they receive the same amount of money, and just because their WAR was higher in their respective walk years, that does not necessarily mean their talent level and expected production over the course of the contract was necessarily higher. By creating an expected contract based on their present talent level, we can better compare who is getting paid more relative to expected production.

To calculate these expected-contract terms, I first set about determining the level of production one could expect from each player in the first year of his new contract. To do this, I used the same method as above, taking each player’s previous three years and weighting them (5, 4, 3 with normal aging). Then, using the first year WAR total, I projected WAR totals over the life of the contract using standard aging curves (-0.5 WAR/year age-31 to age-36, -0.75 WAR/year at age-37 and beyond). That WAR was translated into dollar figures by using $8 million per WAR this season and subtracting $250,000 every year for the value of a win at contract start date. Over the course of the contract, 5% inflation per year was used when valuing a win. Note that for contract years I’ve used the same figures as each player actually received. While it is possible to create an expectation for more or fewer years, using the same length for each expected contract allows for easier overall comparison.

The projected dollar figures overshot the actual dollar figures by about 15%, and there are certainly a few valid reasons for this: (1) it is possible that high market hitters are slightly undervalued when compared to all of free agency, especially pitchers and those who take mid-sized deals, (2) it is possible that the dollar-per-win estimates used are a bit high, and (3) some regression toward the mean might be necessary. While not necessarily solving the above problems, but helping to more easily show the difference between the big walk year group and the rest, the expected figures were downsized to meet the totals actually handed out.

The average expected value of the contract for the average walk year players was higher than those that had a big walk year, but as we know from above, the players with the big walk year got paid a bit more. To be clear, the average walk year players are not players who are average relative to Major League Baseball, but put up a performance that was average for them in their walk years. The chart below shows the difference between the groups.

Effect of a Big Walk Year on Free Agent Contract
Group (No.) Expected $ (in M) Actual $ (in M) Difference
Walk Year Bump (15) 101.6 109.5 $7.9 M
Walk Year AVG (18) 112.8 102.8 -$10.0 M
Walk Year Decline (6) 86.3 96.5 $10.2 M

The walk year decliners have been included mainly to separate themselves from the players who had average years. With just six players who had surprisingly poor walk years, not much can be shown, but the decline did not appear to have had much of an effect on the individual contracts in that group. As for the players with the big walk year, the difference in expectations between those players who produced a typical year is nearly $18 million over the life of a contract.

In looking at the individual numbers, which I will show in full below, there is a major outlier in Jason Heyward, whose projection had him getting a massive $300 million payday. A few other players are out there, but not quite to that level. Even taking Heyward away, there is still a difference between groups.

Effect of a Big Walk Year on Free Agent Contract (w/o Heyward)
Group (No.) Expected $ (in M) Actual $ (in M) Difference
Walk Year Bump (15) 101.6 109.5 $7.9 M
Walk Year AVG (17) 101.8 98.0 -$3.8 M
Walk Year Decline (6) 86.3 96.5 $10.2 M

Admittedly, 39 players is not a lot of data — nor is the addition of Cespedes, Davis, or Upton likely to move the needle much this offseason unless they end up taking less than $100 million. Looking into the individual numbers, over the last ten years, the players who had a big walk year were more likely to be overpaid relative to expected production and highly unlikely to be underpaid.

While it is not definitive, there is some evidence to suggest that players benefit from having big walk years compared to similarly situated free agents.

*****

Here’s the full chart of free agents in their walk years. Note that Victor Martinez appears twice; he signed two contracts.

Free Agent Performance in Walk Years
Exp Walk WAR Walk Year WAR Diff %Over Exp. WAR AGE at k Exp. WAR First Year of k Years of k Exp WAR Exp. $ Actual $ DIff $
Victor Martinez 0.9 4.3 3.4 401.0% 36 1.6 4 3.1 21.3 68.0 -46.7
Nelson Cruz 1.2 3.7 2.5 200.0% 34 1.8 4 4.2 28.8 58.0 -29.2
Yoenis Cespedes 2.9 6.7 3.8 131.0% 30 4.5 6 19.5 145.3
Adrian Beltre 2.9 6.4 3.5 120.7% 31 3.8 5 14.2 87.2 80.0 7.2
Adam Dunn 1.4 3.0 1.6 116.9% 31 1.3 4 2.4 14.3 56.0 -41.7
Jose Reyes 2.8 5.9 3.2 114.5% 28 3.5 6 18.0 117.7 106.0 11.7
Jason Bay 2.2 4.6 2.4 107.5% 31 2.2 4 5.8 33.4 66.0 -32.6
Shin-Soo Choo 3.0 5.5 2.5 85.9% 31 2.9 7 9.9 71.3 130.0 -58.7
Carl Crawford 4.2 7.7 3.5 84.1% 29 5.9 7 33.8 219.1 142.0 77.1
Gary Matthews 2.1 3.9 1.8 82.8% 32 2.5 5 7.5 44.3 50.0 -5.7
Alfonso Soriano 3.0 5.4 2.4 79.5% 31 3.1 8 11.2 71.9 136.0 -64.1
Chris Davis 3.2 5.6 2.4 75.5% 30 4.4 6 18.9 141.2
Russell Martin 3.0 5.0 2.0 68.1% 32 3.5 5 12.5 88.1 82.0 6.1
Aaron Rowand 3.2 5.4 2.2 67.9% 30 3.6 5 13.0 70.9 60.0 10.9
Jacoby Ellsbury 3.6 5.6 2.0 56.3% 30 5.1 7 25.2 181.5 153.0 28.5
Mark Teixeira 4.5 6.9 2.4 54.2% 29 5.2 8 31.1 192.0 180.0 12.0
Alex Rodriguez 6.3 9.6 3.3 53.2% 32 7.0 10 45.0 282.6 275.0 7.6
Jorge Posada 3.9 5.6 1.7 43.0% 36 4.1 4 12.6 66.9 52.4 14.5
Prince Fielder 3.4 4.7 1.3 37.6% 28 4.3 9 28.2 200.7 214.0 -13.3
Torii Hunter 2.4 3.2 0.8 32.9% 32 2.2 5 6.0 32.7 90.0 -57.3
Jhonny Peralta 2.9 3.8 0.9 31.4% 32 3.1 4 9.4 62.4 53.0 9.4
Aramis Ramirez 3.0 3.7 0.7 24.7% 29 3.4 5 14.0 82.7 75.0 7.7
Jason Heyward 4.9 6.0 1.1 21.8% 26 5.1 8 37.8 298.7 184.0 114.7
Jayson Werth 4.5 5.1 0.6 14.4% 32 4.5 7 20.8 134.8 126.0 8.8
Justin Upton 3.2 3.6 0.4 11.9% 28 3.1 6 15.6 117.0
Nick Swisher 3.6 4.0 0.4 11.9% 31 3.9 4 12.6 80.9 56.0 24.9
Victor Martinez 3.2 3.5 0.3 10.5% 32 2.4 4 6.6 39.4 50.0 -10.6
Hanley Ramirez 3.2 3.3 0.1 2.6% 31 3.2 4 9.8 67.2 88.0 -20.8
Pablo Sandoval 3.0 3.1 0.1 2.5% 28 2.6 5 11.5 81.0 95.0 -14.0
B.J. Upton 3.4 3.3 -0.1 -2.5% 28 3.1 5 14.0 92.5 75.3 17.3
Matt Holliday 5.7 5.4 -0.3 -5.8% 30 5.4 7 27.3 170.4 120.0 50.4
Josh Hamilton 4.8 4.4 -0.4 -7.4% 32 4.7 5 18.5 121.9 125.0 -3.1
Chase Headley 4.6 4.2 -0.4 -8.7% 31 4.3 4 14.2 97.4 52.0 45.4
Robinson Cano 6.4 5.8 -0.6 -10.0% 31 5.7 10 33.0 259.1 240.0 19.1
J.D. Drew 4.7 4.2 -0.5 -10.0% 31 4.3 5 16.5 86.2 70.0 16.2
Brian McCann 3.1 2.8 -0.3 -10.2% 30 2.6 5 8.0 54.5 85.0 -30.5
Carlos Lee 3.5 1.9 -1.6 -45.8% 31 2.3 6 6.5 39.5 100.0 -60.5
Alex Gordon 5.4 2.8 -2.6 -47.7% 32 3.8 4 12.2 86.2 72.0 14.2
Albert Pujols 7.8 4.0 -3.8 -48.8% 32 5.5 10 30.0 219.8 240.0 -20.2
Derek Jeter 4.6 2.3 -2.3 -50.3% 37 3.1 3.0 7.1 41.3 51.0 -9.7
Ben Zobrist 5.4 2.1 -3.3 -61.2% 35 3.5 4 10.8 76.5 56.0 20.5
Curtis Granderson 4.4 1.6 -2.8 -63.4% 33 2.8 4 8.2 54.4 60.0 -5.6





Craig Edwards can be found on twitter @craigjedwards.

6 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Original Greaser Bob
6 years ago

So Theo wrecked it. Again.