I feel like there's a bit of a lack of rigor when it comes to analyzing manager decisions here, which is maybe appropriate for game threads, but it feels as though the main board should have something of a requirement that at least some presentation of the available alternatives be discussed. It's very easy to say "pull Porcello" or "let Barnes come back," but I think it's incumbent on us to at least try to understand what the manager was going for and what his alternatives were before we resort to assuming a decision is obvious or based on factors like loyalty instead of what the decision should be -- "what choice increases my championship win percentage most"?
So, anyway, I'll try to be a bit more methodical about how I view last night and the choices that Cora made. My ultimate conclusion is that he was managing his ass off and that all of his decisions had a purpose. Some of it was done by feel but by and large I think he was keeping the ultimate proper question in mind -- what is my risk reward here.
1. Cora made a choice that Hembree and Workman were not going to pitch in the game while it was winable. This gave him the following options behind Porcello: Kelly, Barnes, Brasier, Kimbrel, Edro, Price. In deciding among those pitchers he had to take into account two significant factors: (1) Brasier, Barnes, Kelly, and Edro had pitched the night before, and (2) his anticipated starter for game 5 was unavailable. Given all of that he managed Porcello correctly in my opinion. He did not have enough outs in the bullpen to pull Porcello after 2 or 3 unless and until it was clear that Porcello was pushing him towards a blowout loss or Vazquez came into the dugout and said Porcello had nothing. It's a shame Porcello did not perform better, but pulling Porcello early was not nearly as obvious as people in the gamethread seemed to be suggesting and it's a classic example of not considering the alternatives. I actually think the question whether to pull Porcello after 4 was a very tough one and I think Cora made the right choice to keep his pitch count low enough that he likely can be available in game 6 and/or 7.
2. With respect to Barnes, I feel like the point people are really missing is that Cora knew that unless we got another grand slam, Kimbrel was going to have pitch that game. The only thing he didn't know was how much. Ideal would have been 3 outs. But that certainly wasn't obvious. I know in the world of perfect hindsight where Matt Barnes is like Cy Young everyone imagines because he threw a nice curveball that he was just going to cruise through the top of the order in the 8th and get three outs with no damage and Kimbrel would only have had to pitch the ninth. But the truth is that Barnes had not been bridged over two innings very often this year, and his results were a bit mixed. Barnes walks guys too. About .6 an inning. There is a fallacy where fans seem to ignore the potential for things not going exactly to plan with respect to moves not made. A good manager, however, has to consider all possibilities. And one of the non-esoteric possibilities of going with Barnes was having to bring Kimbrel in with 4, 5 or even 6 outs still to get, and maybe with men on base. The choice was not between Barnes and Kimbrel not pitching. Kimbrel was going to have to pitch anyway. If he was going to suck, there was just as much chance he was going to suck in the 9th as there was that he was going to suck in the 8th. With 9/1/2 coming up Cora made a very tough decision and decided that he would worry about the 9th when it came and if there was anything to worry about. That people think this was obvious is astounding to me. Kimbrel was going to pitch sometime.
3. It's also useful to think about how the Sox had gotten to that point. Cora plainly had a plan with a lead. He wanted Brasier to be the guy he stretched. He wanted EdRo to get his lefty, and then he wanted Brasier to get 5 outs. If that had worked, he has a clean 8th and 9th for Kimbrel and Barnes. From his postgame comments about wanting Kimbrel for Kemp and the top of the order I think there actually is a possibility that, had things gone to plan, Cora might have been playing around with which order to use Barnes and Kimbrel in the 8th and 9th depending on the lineup that was due up. Anyway, EdRo did not get his man, and Brasier couldn't get that last out and gave up a double to put two men in scoring position in a two-run game. Cora had three choices to make: (1) Hembree, (2) use Barnes earlier than he wanted to, or (3) stick with Brasier and hope. It looks like he never gave serious consideration to Hembree because he wasn't warming and so Cora had already essentially elected against option 1 earlier in the inning by not even having Hembree take off his sweatshirt. That left options 2 and 3 and it was very hard choice. But once he went with Barnes that really dictated the very difficult end of the game scenario as it played out.
To me, the main question when judging managers is not how they perform when they are choosing between good options, but how they act when they are choosing among shitty options. I think if EdRo or Brasier gets that out we're never having this conversation. But they didn't and I think the risk-reward percentages of the choices Cora had to make were all way more narrow than people are thinking. Worse, as mentioned, I think there is this overwhelming desire to act like it's obvious that the choice not taken would have worked out, when the job of the manager is actually to try to weigh the odds that it might not.