Of course managers count in the dugout. Someone has to be in charge, make the decisions, listen to the coaches, encourage players, etc. But it's only 25 players, and in most games only about half of them get to play, and in fact the lineup and the rotation work best when they are steady--the same 9 and the same 5--especially in the American League. For most of this year, however, Farrell has tended to platoon in LF and, lately, at 3B, but even those decisions are pretty straightforward. He used Ortiz less at 1B in NL parks than usual, but that made sense. He has varied the rotation only when injuries occurred to ERod and Wright, plus he used Pomeranz when he arrived. Now it's pretty steady except for #5, which goes back and forth between Buchholz and Wright. Indeed, as the season has gone on, the rotation has only improved with or without Wright.
The tricky part, as always, is with the bullpen--when and who to bring in. This is a lot easier with a good hitting ball club that gets good leads and less so in close games when the bullpen is unreliable. That's where you and I enter the fray. I think most reliever decisions by any manager are rational and defensible because they are a matter of stats, trends, experience, advice from two coaches, and time to make them in. But, when the relievers are consistently unreliable, every rational, defensible decision (my words) becomes, in your words, a clear indication of "clueless." You believe the bullpen fails because they were used too soon or too late or the wrong guy was used. I think they fail mostly because they are what they are.