What is a top prospect?

I’ve been putting together my top-30 prospect list for this fall since the minor league season ended. Its not an easy task to take 40-50 prospects inside the Yankee organization, pick thirty, and say which one is better than the others. How do you do it?

We know basically what makes a good baseball player. Those players will, when you add up their pitching, fielding, and hitting contributions, help their team win more baseball games than a readily-available alternative. A good prospect is a player that is most likely to become a player who will contribute.

However, a layer of probabilistic variation makes it more difficult. Some players are more likely to contribute a little, some players are more likely contribute a lot. There are different axis which we have to consider how to value players. Would you rather have a player with a 100% probability to become Boone Logan or a player with a 20% probability to become Clayton Kershaw? I take these things very mathematically, like a hand of poker: I’ll take the highest probability of producing WAR. A player with a 20% chance of producing 6 WAR should be roughly the same bet as a player with a 100% chance of producing 1.2 WAR.

Of course, its not that simple. There are 25 spots on a roster, and 14 of those spots get significantly more playing time than the others. If the WAR distribution is roughly shaped like a bell curve, then we can expect it to be much easier to find five 1 WAR players than one 5 WAR player. Given the choice between twenty-five 1 WAR players and five 5 WAR players, I’ll take the 5 WAR guys every single time, because I can go out and fill my roster with replacement players and find value.

There’s one more important value to prospects: time. A player in the minor leagues isn’t doing much good for the Yankees. Like a curve showing how much interest someone will pay over a period of time, the same benefits far in the future are worth less to us than benefits right now. All else being equal, we’ll take the prospect who will produce 20 WAR over 2013-2019 over the guy who will produce 20 WAR over 2016-2022.

All of this means three things:

• We should value higher-ceiling prospects above and beyond lower-ceiling guys.
• We should value prospects who are close to the majors over guys who are farther away, in terms of ETA.
• We should value players who are more likely to reach their potential over guys who are not.