Over the past few years, Nathan has consistently posted WARP3 scores around 6.0 (meaning he was six wins better than the average guy in Triple-A or on the waiver wire), and there’s no reason to believe that this year would have been any different. Those six wins last year would have completely rewritten the story of the AL Central, from a neck-and-neck race to the end to the Tigers clinching in the last week and resting up before meeting the Yankees in the ALDS. Nathan makes that much of a difference.

If you use a 1.75 ERA, 70 IP, and a leverage of 2 (all aggressive), you need a replacement level ERA of 5.60 to get 60 RAR using RAR = (repERA – playerERA) * IP/9 * LI.

If you chain the Twins bullpen using the default settings for the rest, you get Nathan worth 3.8 wins.

Simply keeping everyone’s innings the same in the After setup as the Before setup and giving all 66 of Franklin’s innings to Joe Scrub, I get that losing Franklin costs the team 8 runs.

Now, 438 IP is a really low number for a bullpen. I realize the Cardinals bullpen threw the least number of innings in the majors last year, but still, I don’t think you can project that few.

This is a really interesting situation, as your projections don’t show Franklin as a very dominant closer, and the rest of the bullpen is relatively strong with lots of small downgrades as you move down the line. This type of bullpen will handle the loss of the closer very well.

Finally, in general, everyone should keep in mind that this type of analysis does not answer the general questions of “how valuable is this reliever?” It answers the question, “how valuable is this reliever to this team?”