I saw this in the comments by a user on Metsblog, and thought it was very interesting and relevant so I'm posting it here:

"In regards to a post at MetsBlog earlier today, MetsBlog reader Jamie wrote...

    After reading an article about Ned Yost's bullpen usage, I wanted (naturally) to check out Willie's. The following info is the Leverage of each of the Mets bullpen appearances. Leverage is defined thusly:

    Leverage measures how important the situations a reliever has been used in are. A leverage of 1.00 is the same importance as the start of a game. Leverage values below one represent situations that are less important than the start of a game (such as mop-up innings in a blowout). Leverage values above one represent situations with more importance (such as a closer protecting a one-run lead with bases loaded in the 9th inning).

    Mathematically, leverage is based on the win expectancy work done by Keith Woolner in BP 2005, and is defined as the change in the probability of winning the game from scoring (or allowing) one additional run in the current game situation divided by the change in probability from scoring (or allowing) one run at the start of the game.

    Now, are you ready for this? Think you know how it'll shake out? Maybe you'll be surprised...

    Games/Innings pitched/Leverage (min. 20 IP)
    1. Billy Wagner 63/ 65.3 1.69
    2. Aaron Heilman 75 /79.0 1.41
    3. Jorge Sosa 38 109.3 /1.35
    4. Pedro Feliciano 72 59.7 1.18
    5. Joseph Smith 49 /41.7 1.11
    6. Scott Schoeneweis 65 54.7 /1.05
    7. Guillermo Mota 48 53.3 0.95
    8. Aaron Sele 32 /51.0 0.62

    In other words, these guys appear in the same order that most of us would rank the ability of each arm in the pen. The best reliever (Wags) has been used in the situations with the highest leverage, and the worst (Mota and Sele) have been used in situations with the lowest leverage."