I’ve always felt that the typical way to compile organizational farm rankings wasn’t good enough. Judging any one prospect is tough enough using all of the available data, if you can get it all.
That said, we have trades, and the order in which players were drafted, as hard evidence of teams’ opinions to anchor around.
Judging an entire farm system is much tougher. There is no hard data about the opinions that teams have about the strength of a system as a whole. And subjectively deciding that this group of 30 players is better than that group of 24 players is invariably full of bias; our brains can’t consider all 54 players independently from those two teams, much less 30 teams’ worth of prospects.
When I worked with an MLB club, then later at FanGraphs, and now at ESPN, I’ve used the Future Value system. It’s scaled to WAR (an all-inclusive number of how good any player is, measured in wins) and adjusts for proximity to the majors and risk. On the team-by-team top 10s, I spelled out the FV grades for more than 300 players, listed about a dozen more prospects for each team, and I had about another dozen or so for each team whom I could have listed: the average number of players…