I think that the Surplus Value framework is more or less correct for valuing players, but I think that the specifics currently used by the larger sabermetric community are limited, and possibly incorrect. First, there is an assumption of unlimited roster spots, infinite playing time and perfect liquidity, and therefore players that are below average but above replacement level have significant surplus value when they are cheap. In a certain sense, it is definitely true that cheap, non-terrible players have value. But I'm not remotely convinced that it scales linearly with WAR, and that role-players and superstars can be plugged into the same formula. At the extremes this is very obvious, as everyone realizes that your favorite team can't package together 6 B prospect and expect the Angels to agree to give them Mike Trout, even if the Surplus Value calculations line up. But the Surplus Value framework doesn't build in any quality-vs-quantity tradeoff, so it is hard to know how to adjust the numbers for low-ceiling players. In addition, baseball players aren't super liquid assets, so there are significant transaction costs associated with converting surplus value into an asset that will actually help the team. In particular, a good team (90+ wins) basically needs its top (by playing time) 20 players to average 2 WAR per player. If one of those players is a 1 WAR talent (like, for example, Stewart), then someone else on the roster needs to be 3+ WAR to make up the difference. Second, if the Surplus Value implementation is correct, then one should be able to demonstrate that empirically. But in reality, the opposite is true. I've tracked roughly two dozen MLBer-for-prospect trades over the past three years. Using PECOTA, KATOH, The Point of Pittsburgh tables, and my own best guesses to fill in the projected values, over and over again the Surplus Value of the prospect side vastly exceeds that of the MLBer side. Generally speaking, the prospect Surplus Value is usually 150% to 200% of the MLBer Surplus Value. Examples of position players: Zobrist/Escobar to OAK - $40M vs $65M Zobrist to KC - $7M vs $30M Cespedes to NYM - $15M vs $30M Kendrick to LAD - $18M vs $30M Frazier to CWS - $31M vs $70M Upton to SDP - $17M vs $70M Myers to SDP - $81M vs $172M Lucroy to TEX - $50M vs $75M Gomez to HOU - $50M vs $100M Two significant exceptions: Donaldson to TOR - $85M vs $66M Simmons to ATL - $60M vs $15M // Maybe teams don't value defense? One even: Eaton to WAS - $100M vs $90M (using KATOH) or $110M (using TPOP table) Two hypotheses 1) Teams value prospects for their future WAA rather than their future WAR 2) The 8% discount rate on future WAR used by TPOP and others is grossly incorrect, and a more appropriate value is something like 40%. So even we agree that Dozier has roughly $50M in surplus value, I think past trades indicate that he is worth something like $75M to $100M in prospects. Someone who is super optimistic about De Leon and Stewart could agree that they get close to the $75M number. But if one isn't (and using your numbers, maybe we shouldn't be), then it seems completely reasonable to want another significant piece in the trade.