I've always been in the "trade him before he leaves" camp when it comes to players like this, but Cameron's work indicates that the draft pick compensation may be even greater, or at least is comparable, in value. Rany Jazayerli's draft study certainly indicates that draft picks are far more reliable than I ever imagined them to be, but my impulse until now has always been to go after good prospects in other systems rather than letting someone walk for "nothing." My reasoning was that these players will be closer to The Show than a draft pick, and therefore more likely to actually contribute in the major league level. However, I think the issue I hadn't considered is that, usually, you're not likely to get the outstanding prospect via those deals that you might get in the draft.
Furthermore, it's apparent from his data that the players received in trades aren't particularly more reliable (on average) in their ability to reach and perform at the majors than players taken via a draft. It's just that drafted players take longer to reach the majors than the returns from trades.
If the Reds are dead-set against picking up Adam Dunn's option at the end of the year, which is what John Fay, at least, believes (ref: Fay's on-air conversations with Marty), then the Reds are in the same position with him as the Mariners are with Ichiro. And I'm not sure what the best solution would be, given this new information...perhaps letting Class-A free agents walk is the better alternative, unless a truly outstanding prospect is made available via trade.
Kudos to Cameron for a nice study--I love it when I have to reconsider my views in light of new information.
Interesting tidbit: the Reds are currently in possession of two of the best returns on mid-summer deals for prospects: Aaron Harang (for Jose Guillen) and Brandon Phillips (via Cleveland along with Grady Sizemore, for Bartolo Colon). So clearly those deals can work out too.
Photo by AP/Jeff Chiu