Two games I've watched live makes two losses for the A's. I feel like I'm not very good at this.

I actually missed the first three half-innings, so I didn't get to see Josh Reddick's first single, nor Nolan Reimold doubling to start things for the Orioles and then somehow being out at third when J.J. Hardy lined out to first? How does that work? (UPDATE: Answer: it doesn't. The play-by-play I was reading calls Reimold out at third, but Jon Bernhardt told me on Twitter that he was doubled off of second easily, as he was attempting to steal third on the pitch that resulted in the line-out. I don't know why the play-by-play says "third.")

Anyway, I came in for the long middle run, where each team got a few base-runners here and there, but were mostly quiet. Bartolo Colon was working the fastball like he does, changing eye-levels and going in and out on hitters. Tommy Hunter was probably pitching well, too. I dunno.

The main item of note was Kurt Suzuki hitting a deep fly to left, only to see Reimold reach over the fence and pull it back. I literally mean "reach" -- he didn't jump, he didn't get excited. He just camped under the ball and put his glove on the other side of the wall and made the catch. It was kind of unfair how casually he took a run away, especially from Suzuki, who needs all the help he can get on offense (.208 True Average coming into the game).

The A's then got single runs in the sixth and seventh with a Reddick-single, Cespedes-double in one frame and a Seth Smith homer that just carried out to left center (oppo taco?) in the other. The Cespedes double was what you might call "ringing," a line-drive shot over the head of the third baseman.

Through all this, Bartolo Colon kept on keepin' on, until we came to the middle of the ninth, with the A's up 2-0, Colon sitting on 95 pitches, and the following due up:

None of these guys are rookies. Chris Davis has had very few recent plate appearances, though, so his numbers should probably be taken with a very large grain of salt. Betemit doesn't have as few PA as Davis, but he's been a part-time player for his entire career, so his recent years don't have as much data in them. The first four hitters due up, though, have been regulars, so we can trust their numbers about as well as we can trust anyone in the league's.

Let's first note that my preferences just aren't a consideration here, because I'd always remove the starter if my only concern was how to win the baseball game. Whether it's because baseball managers have to manage egos and expectations or because they genuinely believe in the hot hand, though, managers will send the pitcher out for the ninth in this situation. Against the baseline of "what happens in MLB," then, sending Colon out for the ninth wasn't a bad decision. JJ Hardy isn't really a threat with the bat, and even if Nick Markakis or Adam Jones get on (via Jones's reverse split), it's only Wieters (who hasn't been anything special against righties) and Chris Davis (who hasn't been anything special period, though he's hit exceedingly well in 71 PA this year) up next.

So given the decision to have Colon start the ninth, you presumably want to have at least one pitcher ready in the bullpen. The question, then, is whether you warm up Brian Fuentes, the lefty, Grant Balfour, the righty, or both. If we again measure by the standards of major-league baseball, the answer is clear: you warm the closer and the closer only. If there's a tough situation, you hand him the ball regardless of matchups and let him pitch his way out.

Measured against what's right, part of the question is how much leash Bartolo Colon has. If he gives up a homer to Hardy to start the inning, are you going out to get him right then? If so, you should probably have a lefty ready to face Markakis and Jones before flipping to a righty to finish things out. If you assume, though, that you're going to have Colon face at least two batters (i.e. worst-case scenario where both of those guys score), then you're looking at Jones-Wieters-Davis-Betemit. This is tougher, in part because of Jones's reverse splits. Maybe your scouts think that those stats are just a weird fluke and that, between Balfour and Fuentes, you'd rather have Balfour.

In real life, of course, Colon gave up an infield single to JJ Hardy, struck out Markakis, and then gave up another infield "single" to Adam Jones on a weak dribbler that Colon himself threw away. Both of the "hits" would have been outs had Daric Barton managed to scoop tough throws at first.

So this brings us to Wieters-Davis-Betemit with runners on second and third and one out. For Wieters, you'd prefer to have a right-handed pitcher. For Davis, you're not sure given his statistics, so you revert to the general MLB population performance and figure a lefty gets him out best. You probably prefer a lefty for Betemit as well, but you're not necessarily confident in that from the numbers.

So at this point, do you go Balfour for just one batter, then Fuentes? What about this: Grant Balfour has shown little to no split in his career, so if you're looking ahead to Betemit and you're not sure which side he's better from, maybe you set up the match-up with both your better overall pitcher and your pitcher who doesn't have weakness against opposite-hand hitters.

In the end, I don't see a clear-cut mistake in Bob Melvin bringing in Grant Balfour to face the three hitters he did. Balfour didn't execute, failing to bury his 0-2 curve to Wieters in the dirt, giving him a pitch he could hammer for a game-tying double, and then leaving a fastball way out over the plate for Betemit, who duly knocked it out of the yard for the three-run homer and the Orioles win, but I think that when you're dealing with Wilson Betemit, you don't go monkeying around with your bullpen to accommodate whatever strengths and weaknesses you perceive. He's Wilson Betemit. The best pitcher in the bullpen should be able to deal with him.