Originally Posted by dougdirt
I guess the problem with the historical value of the lists problems begin with different writers. Baseball America and John Sickels are the only places that have been forming a list for 8+ years with at least the main person making the call on the list. Even BA rotates some writers in and out as far as their Top 100 meetings go, and the team rankings have probably 10-15 different people making the lists, though I believe Callis does have final say there. With that said, Victor Wang has done some research on the value of those two lists rankings. You can see that at The Hardball Times
The problem with that is that the data is based on lists from 1990-1999 and surely things have changed somewhat since then in how evaluators are ranking prospects.
I would actually disagree with your premise, Doug. The problem at its core is the simple fact that there is no agreement about what is being assessed. We've seen in our own discussions definitions ranging from ceiling to floor to likely path to trade value...
Unless and until there is some agreement about what exactly is being projected, there is no way to compare the "accuracy" of the projections themselves, regardless of by whom or through what process they were created. In the article you reference, Victor is clearly basing his analysis on a likely production basis. Though he is doing it a bit backwards, using ranking to project WAR rather than the other way around.
When I look at prospect lists across the net, few of them seem to be clearly organized on such a clearly articulated principal, let alone using an actual quantitative assessment such projected WAR (or WAR-like) over say, the 6 pre-FA years for which the player will be under team control.