Blah blah blah getting ahead of yourself jinx curse blah. Get that out of the way, then consider this. The Cubs will be playing in the 2008 NLDS, they will probably have time to set their rotation and they will probably have home-field. Presuming we have time to set our rotation, the games would be on these dates: Oct. 1 or 2 at Wrigley Oct. 3 at Wrigley Oct. 5 on the road Oct. 6 on the road Oct. 8 at Wrigley Choosing the first would easily allow us to go with three starter (which I'd assume we would want to do, though maybe if it was Arizona we wouldn't want to give them the option, but I'll ignore that for now). I'm going to make a case that the Cubs should do something they probably won't do: Bump Zambrano to Game 3 and live with just one start from him. First, Rich Harden is the best pitcher, so you need to give yourself the best chance possible to give him two starts. That means he has to get the ball Game 1 and Game 4. Then I argue that you have to start Dempster at Wrigley and not on the road. ERA doesn't seem too much different this season home/road for him, but RA tells a different story. Wrigley RA: 2.86, Road RA: 3.96. This might be a statistical fluke, but there's no way to know for sure. Zambrano has no real split over his career, a much larger sample. Declaring Wrigley Dempster to be a separate pitcher from Road Dempster, we trot out the following RAs by game: 2.19 2.86 3.55 2.19 2.86 Compared with what we get if we do the more expected Zambrano/Harden/Dempster: 3.55 2.19 3.96 3.55 2.19 Or even worse, Zambrano/Harden/Dempster/Lilly: 3.55 2.19 3.96 4.50 3.55 Plus, with Zambrano only throwing one game, there's less pressure to save his arm.