The WGOM

So the beat reporters at the Minneapolis Star Tribune were busy on Wednesday. First, Joe Christensen wrote about how the starters now have sufficient amounts of chest hair to pitch deeper into games. Or something like that.

As a generic position, I'm not really sure what to say about this. Pitchers ought to be removed from the game when it is to the team's benefit to do so. "To the team's benefit" is a tricky thing to make broad statements about, though, because the Twins have to worry about winning games today, tomorrow, Friday, next year, and five years from now. Those are obviously in decreasing order of importance, but they all figure into the decision on some level. I don't think Gardy's going to let these guys throw more pitches than he let Santana pitch, so I'm not really worried about him blowing out their arms. While 107 for Perkins on Tuesday is a little on the high side, it's nothing to worry about and he didn't have any especially stressful innings.* And it seems that in recent years Gardy's been more prone to warming up everyone in the bullpen and rifling through all his options in one inning, racking up appearance totals (which seems to be bad for relief pitcher health.)

*There's a certain school of thought that asserts high-exertion innings are a bigger problem than high pitch counts. I tend to fall in this camp. I think that the most dangerous thing is for a pitcher to be throwing when he is fatigued, because then his mechanics start to fall apart and he'll do things that his body isn't used to him doing. That's when I think that things tend to snap.

In Tuesday's specific case, I think that it would have been completely reasonable (and very likely the optimal strategy) to bring Joe Nathan in for a 4-out save. Unless we know something that makes Nathan more fragile than Crain, I don't understand why putting Nathan out there for 1 1/3 innings would have been so much worse than Crain going out for 2 innings. It's not like it's a slight to someone to tell them that they are coming out for Joe Nathan--he's one of the very best pitchers in baseball.

Then there's the matter that hitters gain an advantage each time through the order against a particular pitcher. This is not an insignificant advantage and it makes sense that as a pitcher tires and a hitter sees more of his pitches that the hitter will gain an edge. In Tuesday's specific instance, it's a matter of Perkins facing a hitter who has platoon advantage on him and was facing him for the fourth time that day. Perkins isn't very good at missing bats, and with a single all that's necessary to tie the game, I think that going to the pitcher who is the very best at missing bats (and the best overall) would have made a lot of sense.

Now, I've said a lot about Tuesday, but I don't want to make this sound like a more serious mistake than it was.* I think it was sub-optimal, but a lot of times in that situation, the manager can get away with keeping his pitcher in the game and holding on to the lead.

*Here's how I figure the breakdown goes, roughly speaking. Perkins has a lifetime .298 batting average against as a starting pitcher. Figure that he's improving and he'll allow about a .283 batting average against this year. Figure that Wells is about a .273 hitter to give before platoon advantage in a league with about a .270 average. Giving Perkins a bit of the benefit of the doubt, that puts us at .283 for Wells' average against Perkins before platoon advantage. Add in platoon advantage and that goes to about .296. Now add in that Perkins is facing Wells for the 4th time in the game and that goes to about .306.

Since everything here is fairly rough, say that Wells ties the game 30% of the time against Perkins and 20% of the time against Nathan. As much as that's a pretty big difference in the spectrum of major league pitching talent, making this decision costs the Twins the tie 1 out of every 10 times, so figure it costs a win once every 20 times this situation comes up. That's still a pretty huge deal if you make that decision every day, but this particular situation isn't going to present itself all that often. If we happened to get one of these once every ten games, which I would consider rather frequent, that's only 16 opportunities so we'd be looking at Gardy costing the team 0.8 win over the course of the season by making this decision. So that could still matter, but there are probably things that Gardy does that we can't directly measure which may help the team by 0.8 win over the course of the season, which makes it difficult to say how many games we would win if we replaced Gardy with a league average manager.

Anyway, let's get to the contradiction I see here:

“Last year, it was kind of like you were pitching against a shot clock,” Perkins said. “You knew you had X amount of pitches, but this year, they’re going to let all of us go.

“[Gardenhire] talked to me after the game, and he said, ‘I had no thoughts of taking you out of that game. You got us that far.’

“Wells got up there, and there was no reason to take me out right there, and I appreciate that. It’s just nice to know they’re going to stick with you. It gives you even more confidence to be able to pitch.”

Now let's go to LEN3's comments about the Duensing situation.

Manager Ron Gardenhire will now go with a bullpen that includes one lefty in Craig Breslow. It’s not the best situation to be in.

We have to remember that Breslow, in a sense, is still the new kid on the block but there was no way the front office was going to risk losing righthander Philip Humber on waivers. (I hope it’s obvious that there’s a difference of opinion going on with the Duensing demotion).

Reading between the lines, it seems to me like the front office doesn't want to lose talent (because they are thinking about next year and beyond) while Gardy wants to have another lefty in the 'pen (because he is thinking about today.) That they are thinking on different time scales is fine (and maybe even optimal), but it seems really contradictory to me that Gardenhire would think that:

A) He needs to switch his relievers out more or less at every chance he gets to gain platoon advantage (the only real reason you would strictly need two match-up lefties in the 'pen)

and

B) The starting pitchers should be kept in games longer because they will pitch better knowing that they get to finish the job

Because if B applies to starting pitchers then it ought to apply to relief pitchers. And in that case, they should be allowed to pitch more full frames without needing to worry about being lifted in the middle of the inning because they gave up a hit or two.

As a closing thought, I wanted to mention The Eddie Guardado Closer Experience. Eddie, if you'll recall, had a pretty big platoon split.

.225/.268/.371 -- Guardado vs. LHB, career
.264/.340/.464 -- Guardado vs. RHB, career

Had Guardado not been anointed "The Closer," we would never have had The Eddie Guardado Closer Experience. All of those innings where he put runners on base and then walked a tightrope out of the situation? They wouldn't have happened in middle relief because with runners on first and second and a right-handed batter striding to the plate, Guardado would have been lifted for a right-handed reliever. In an alternate universe, LaTroy Hawkins held on to the closer's spot and Guardado was a LOOGY. But in our universe, Guardado is living, breathing proof that pitchers can and do get out of tight spots even if they have to (gasp!) face opposite-handed hitters.

I'm not fully confident in the bullpen, but I'm also not worried that we only have one left-handed reliever. I'm worried that we don't have enough quality relievers.