Taking Nathan out of my projection system costs the Twins about 3-4 wins.  Rankings are after the jump.  (here's where I first posted about the way I do things: http://www.twinkietown.com/2010/2/18/1315802/power-rankings)

I updated the system to use the latest PECOTA data, which changed things pretty dramatically by itself on the pitching front.  E.g., the Red Sox and Cardinals got even better, the Yankees and Angels got much worse.  The Twins pitching also gets worse with the new data.  Neshek, previously projected as quite good, is now projected with an ERA around 6.00.  I just take the top 7 projected relievers under PECOTA to create my bullpens, so currently, Neshek doesn't even make the team.  Some of PECOTA's calls seem odd to me, but I'm just a slave to the system, so there it is. 

I also shifted from calculating runs scored using just team OPS to using team OBP, SLG, SB, and caught stealing all seperately.  So that should give a little truer read on run scoring, but it didn't make dramatic changes.  The Red Sox and Rays, e.g., edged up a tick since they figure to steal bases effectively. 

The Twins went from winning about 88-89 games with the old PECOTA data and run-scoring system to winning about 86-87 with the new--from the 5th best team in baseball to the 9th.  Still strong favorites in the division. 

The loss of Nathan now takes the Twins down to about 83 wins and to the 13th best team in baseball.  The next best team in the AL Central, now the Indians (who also improved with the new PECOTA data), project to about 79-80 wins.  The Tigers and White Sox are close behind.  So I think the Twins are still favorites, but it's become a pretty close race. 

I added strength-of-schedule data in the tables below.  One interesting thing for we Twins fans is that the Twins figure to have the easiest schedule in the AL.  As with last time, the first set of data is figured against a neutral team, and the divisional rankings are figured by actually running the teams through their 2010 schedules.