Grading Trades: McCann Test

I recently presented a detailed trade analysis series in which I hypothesized that using harsh depreciation of player value (10 percent depreciation per season) would help to frame trade value for a player. I used the model to grade many of the Brewers’ 2015-2016 trades, and assess current MLB and minor league assets. The model is highly abstract, and encounters potential issues in terms of monetizing WARP (there are real issues with the ~$7 million/WARP market), and also in terms of how to assess potential production. My main assumption about depreciating a player’s production value is that a team will not price a player at a value that is higher than his current performance; this does not mean that each team does not expect their newly acquired roster asset to decline, but rather that even where a team expects a player to improve, they will try to pay a lower price in trade acquisitions.

In terms of assessing prospect trade value, I expected 60 Overall Future Potential (OFP), 55 OFP, 50 OFP, and 45 OFP prospects to have a different combination of WARP over a three season period, reflecting the likely up-and-down performance of a prospect entering the MLB. The issue here is assessing that grade longterm (for example, consider a 60 OFP prospect such as Isan Diaz transcending shortcomings to become the next Robinson Cano, versus encountering mechanical or injury troubles and burning out, versus being traded by the Brewers for another player). Basically, even if a 60 OFP prospect might be expected to become an above average MLB player, it is extremely unrealistic to price them at, say, 4 WARP per season (such a prospect would be valued at least $168 million over their full roster reserve period, and therefore highly unlikely to be traded).

While the Milwaukee front office has been relatively quiet thus far, I will take the opportunity to test my assumptions by assessing the trade between the Yankees and Astros involving catcher Brian McCann. The trade is quite a value play for the Yankees, who are digging deep into the Astros system for young righties that could vastly outplay their current outlook as they develop, or fail to make the next step in their respective minor league careers. This is arguably the benefit of going after extreme low-minors arms. Consider Milwaukee’s own trade involving Adam Lind, Carlos Herrera, Daniel Missaki, and Freddy Peralta; because those arms were so young, Milwaukee could arguably land enough depth to make the gamble pay off (and, indeed, it already has). RHP Albert Abreu and RHP Jorge Guzman were not in the 2016 Top 10 for the Astros system, but BaseballProspectus graded Abreu 60 OFP by the end of the season, suggesting that 2016 was the breakout year for Abreu (who indeed worked from 2015 Rookie ball to Advanced A by the end of the 2016 season).

McCann himself is a relatively stable catcher, posting TAv of .258, .270, and .256 in the Bronx (alongside 22.8 FRAA). The veteran catcher is a good bet to match his remaining two year cost (potentially three due to a vesting option with no buyout), even considering a heavily depreciated WARP outlook. News that the Yankees will kick in $11 million in cash basically makes McCann a wash for the Astros (he will need to average 1.6 WARP over two years to reach that contractual value, a level of performance that is even below a harsh 10 percent depreciation of his current performance level).

To the numbers! The fun part about this trade is considering how MLB teams value cash in trades. Is the cash truly counted against McCann’s contractual value? Or did the Yankees include the cash to purchase an additional prospect?

As you can see, the approximate surplus value for McCann ($27.2 million) is within solid range for a three-year 60 OFP outlook for Abreu ($34.3M), while Guzman’s three-year outlook ($19.6M) is relatively close to the amount of cash that the Yankees surrendered.

Evaluating this model, it is clear that I have not depreciated potential prospect value harshly enough. Basically, over the full six or seven years of reserve control, nearly every prospect OFP value will be greater than an MLB established veteran with a big contract, so I must re-evaluate my prospect depreciation model to account for the risk of a prospect not making the MLB. Obviously prospects are not truly more valuable than MLB established veterans, otherwise teams would rarely trade prospects for veterans if they were attempting to maximize the value of their roster assets. If one assumes that MLB teams each attempt to find an equilibrium price in a transaction (which could be accomplished using game theory), one must evaluate non-depreciation models for veterans, and extreme-depreciation models for prospects.

Exogenous to Game: Astros potential playoff revenue from adding McCann; Astros potential prestige value for winning division, playoff series, pennant, and championship; Yankees $23 million in freed payroll space; Yankees value for keeping Abreu and Guzman seven years each (highly unrealistic / unlikely).

For the brief game above, I assumed that both teams value assets equally, for the sake of simplicity and demonstration. In this case, I revalued McCann at his “actual rate of production” over the life of the contract. For Abreu and Guzman, I added 10 percent depreciation and 10 percent risk to reflect the distance of a prospect from the MLB, and also the general assumption that players will not produce at a consistent level over three seasons. In this case, depreciating 7.0 expected three-year WARP for Abreu by 60 percent revalues the prospect at approximately 2.8 WARP, while depreciating 50 OFP for Guzman over thee-years produces a figure of 1.6 WARP. This hardly accounts for a satisfactory distance between a near-breakout 60 OFP prospect and a basic 50 OFP prospect who throws heat but has a relief role thus far; incidentally, assessing Abreu at a 30 percent three-year depreciation rate and Guzman at a 60 percent three-year depreciation rate gets this game remarkably close to equilibrium (0.4 WARP is basically a rounding error, in the terms of replacement player theory).

From this model, one might conclude that (a) MLB teams do not necessarily “depreciate” an MLB veteran’s production level in order to price him for trade, and (b) different prospect OFP should be depreciated at different levels. The next step would be to advance game theory into more competitive models, in order to judge how teams attempt to receive maximal value from each trade asset. Indeed, one could argue that in a talent-scarce monopoly environment such as the MLB, teams could be expected to exhibit rent-seeking behavior through their trades, rather than competitive equilibrium.