Here are Crawford's career #'s by batting order position (baseball-reference.com):
1. 1728 pa, .284/.319/.415/.734
2. 2705 pa, .303/.346/.460/.806
3. 863 pa, .291/.336/.448/.784
4. n/a
5. n/a
6. 198 pa, .282.308/.452/.760
7. 150 pa, .299/.351/.522/.874
8. 156 pa, .238/.277/.347/.624
9. 129 pa, .285/.305/.358/.662
From these numbers, it seems like the #2 or #7 spot is the best place for him, with the #7 spot being his best. Now, there's a LOT more data from which to make a judgment about the #2 spot, so maybe he's right that he just feels more comfortable there. So let's work with the assumption that he's correct and he should be in the 2 hole. Say you get the "really good Carl" if they move him back to the 2 hole, and he puts up a .303/.346/.460/.806 line, also taking advantage of his tremendous speed on the basepaths. I think we all would agree that this would constitute a pretty serious offensive weapon, especially given the rest of the lineup.
So that's how to maximize Carl. But what about the rest of the lineup?
According to (
http://deadcatsbounc...arances-by.html), here are the average number of plate appearances by spot in the order:
1st - 762
2nd - 743
3rd - 725
4th - 709
5th - 693
6th - 675
7th - 657
8th - 638
9th - 618
I'm sure it fluctuates from year to year, but let's work with this. If everyone stays healthy (a big "if", but work with me), and they go with essentially last year's lineup, here are the number of PA per player:
1st - Ellsbury - 762
2nd - Pedroia - 743
3rd - Gonzalez - 725
4th - Youkilis - 709
5th - Ortiz - 693
6th - Crawford - 675
7th - Ross/Sweeney - 657
8th - Aviles - 638
9th - Saltalamacchia/backup C - 618
If they move Crawford up to the 2 spot, where he says he's much more comfortable, here's what the lineup looks like:
1st - Ellsbury - 762
2nd - Crawford - 743
3rd - Pedroia - 725
4th - Gonzalez - 709
5th - Youkilis - 693
6th - Ortiz - 675
7th - Ross/Sweeney - 657
8th - Aviles - 638
9th - Saltalamacchia/backup C - 618
So everything else stays the same except for Pedroia, Gonzalez, Youkilis, and Ortiz. Here's the rough number of PA each guy would lose in this scenario:
Pedroia: 18
Gonzalez: 16
Youkilis: 16
Ortiz: 18
So (and I know this is a lot of speculation, because, really, who knows?) let's make the comparison....
Scenario 1 - Crawford in his "uncomfortable spot in the lineup" produces like he did last year overall: .255/.289/.405/.694, and Pedroia, Gonzo, Youks, and Ortiz do what they normally do (let's assume they aren't affected by batting order slotting, b/c they've not really complained about that ever, so far as I'm aware):
2nd - Pedroia - 743 - .305/.373/.463/.837
3rd - Gonzalez - 725 - .293/.375/.514/.889
4th - Youkilis - 709 - .289/.391/.492/.883
5th - Ortiz - 693 - .283/.378/.544/.922
6th - Crawford - 675 - .255/.289/.405/.694
Scenario 2 - Crawford moves to the 2-spot and produces his career norms in that role: .303/.346/.460/.806
2nd - Crawford - 743 - .303/.346/.460/.806
3rd - Pedroia - 743 - .305/.373/.463/.837
4th - Gonzalez - 725 - .293/.375/.514/.889
5th - Youkilis - 709 - .289/.391/.492/.883
6th - Ortiz - 693 - .283/.378/.544/.922
It would seem to me that the few PAs you'd lose for each of the four other guys would be worth it if you could get that kind of improvement from Crawford.
Of course, this little exercise assumes: (1) that Crawford actually will rebound pretty much back to his career norms by being in the more comfortable 2 spot, and (2) that the other guys that would move down wouldn't be negatively impacted by such a move.
Both of those assumptions, it should be clearly noted, are questionable. If we knew what would happen, then this decision would be easy. :-)