Author: Mike Pastuovic (michaelpastuovic2023@u.northwestern.edu), Weinberg ’23

Intelligent spending in free agency is crucial for constructing rosters that are built for sustained success. Knowing that championship-caliber teams are rarely built through solely internal player development and trades, teams allocate a considerable portion of their payroll to free-agent acquisitions. We are interested in examining how especially large contracts tend to play out.

Figure 1 illustrates all free-agent contracts that a player signed with a new team that was valued at greater than 100M in 2018 league dollars. Instead of adjusting for inflation, we decided to adjust the threshold for inclusion by contract value proportionate to total MLB salaries for the given year.

Teams spend incredible sums of money on free agents. Quickly comes the question, are these contracts worth it? Interesting insights could come from examining players’ production in the years before and after signing these lucrative concepts. Figure 2 explores these players’ production in the three years before signing the contract compares to production in the three years after signing the contract.

As seen, a great majority of players have a lower average WAR in the first three seasons after signing their new contract than they did in the three seasons preceding the contract. WAR is a statistic that includes all facets of a player’s game and combines them into a single statistic to capture how many Wins Above Replacement the player was worth in a given season. We used Baseball Reference’s method of calculating WAR for this project.

37 of the 53 players saw a decrease in average WAR accumulation in the three years after receiving these lucrative free-agent contracts with new teams. The rest of this article will be structured around exploring why these high-end players often perform worse after receiving the large contract and how we can predict the production of future free agents. It is important to note that a signing is not necessarily a failure just because a player is not performing as well as they may have in the past. Teams likely know that achieving up to past performance is unlikely, and still are interested in paying large sums of money for slightly lower projected performance. We will explore how factors like age, contract value, position, and the decade of signing can predict performance relative to past performance.

As we can see in Figure 3, players who sign in their 20s tend to play more closely to their previous production than players in their 30s. This is likely because the primes of careers tend to run from about Age 26-30, as has been thoroughly researched. As players go through their 30s, their production tends to steadily decline. So, our finding that players who sign in their 20s tend to play more closely to their previous production than do those in their 30s makes intuitive sense.

We are also curious if the overall value of the contract can predict a change in performance. It is possible that players crumble under the pressure of a particularly large contract and consequently fall quite short of past performance. However, as we can see in Figure 4 it does not seem like there is any significant trend showing that contract value is predictive of the amount of change in performance. For all contract values, the change in average WAR is relatively similar.

Pitchers and Position Players are the primary positions in baseball and have very different makeups and skill sets. It seems reasonable that they could potentially have significant differences in their respective changes in three-year average WAR after signing a contract. As we see in Figure 5, there is not much difference. In the past, pitchers have performed a little more closely to past performance, but both groups see comparable, significant decreases in WAR after signing these large contracts.

Finally, we look at the decade in which the players signed these free-agent contracts. With so much new data available to front-office decision-makers, in recent years they may have signed players in free agency that are more likely to live up to their past performance. However, as we see in Figure 6, this has not been the case. Players signed 2010-2018 have actually performed worse in comparison to past performance than players signed 2000-2009.

As a group, players have not been the same players that they were before they signed the big contract. But that does not mean that they are not living up to the contract. Front offices have access to a great deal of data and know that for a variety of reasons, production is likely to drop. Our findings indicate that older age exacerbates average production drops. However, we do not find significant differences in expected production decrease when studying contract value brackets and positions. Even as front offices became more data-driven and analytically savvy over the last decade, the players they signed to large contracts saw greater production drops on average than did players who received similar contracts from 2000-2009.

Front offices must look at past free agency signings and outcomes to evaluate similar players and determine fair prices. They need to understand that they should not pay for past performance, nor expect that players will continue to produce at past levels in most cases.

Sources:

Contract Data: Cot’s Baseball Contracts – Baseball Prospectus

Total League Payroll Data: The Baseball Cube

Player Performance Data: Baseball Reference