Primary colors: Coincidence or evidence?

Kohl Stewart on Tuesday made his eighth appearance with the Twins. The first four were straight-up starts; the most recent four were as a "primary" pitcher, still charged with getting through the lineup at least twice but not pitching the first inning.

He allowed exactly three earned runs in each of the four starts but never finished the fifth inning: a total of 16.1 innings with an ERA of 6.61 and a dismal nine walks (with 11 strikeouts).

In the second four outings, including Tuesday, he worked 20.1 innings, allowing four runs, three earned. ERA: 1.33. The walk-strikeout ratio is better, but not good (nine walks, 13 strikeouts).

Coincidence, or evidence that the "opener" tactic works? Thirty-six innings and change doesn't really settle anything, and there are signs that a lot of luck is involved, but Stewart has been better while avoiding the top of the order while getting settled in. That's part of the theory behind the opener.

Paul Molitor pulled Stewart after six innings, 75 pitches and 21 batters Tuesday, and the bullpen -- specifically Trevor Hildenberger, who is suddenly struggling again -- promptly surrendered four runs. The decision to yank Stewart was second-guessed on social media, with the commentary focused on the 75 pitches.

More important in that decision, I think, is the 21 batters. That's twice through the batting order, plus three hitters. Opener Gabriel Moya faced the first five men, then Stewart began with the Tigers No. 6 hitter (James McCann). Molitor wanted to prevent Stewart from facing the top of the order a third time, and also wanted to give Hildenberger a clean start to the inning.

The pitch count was a secondary consideration. It clearly didn't work, but that's on Hildenberger.

Earl Weaver and the opener

The Twins are to deploy Gabriel Moya as the opener again today -- he had that duty on Monday as well -- with Stephen Gonsalves apparently slated for the primary pitcher role.

This new-school tactic clearly does not sit well with at least two prominent Twin Cities media voices, Bert Blyleven and Patrick Reusse. Their disdain for the opener strikes me as a feature, not a bug.

And it also seems to me misplaced, given that their stated concern is for the future of what we know as starting pitching. I think this approach fits well with the development of true starters.

Gonsalves has put up ugly numbers in his first five major league appearances -- 9.39 ERA in four starts and one "primary" outing with 17 walks in 15.1 innings.

His big league struggles come after a stellar season in Triple A, where he made Baseball America's Triple A full-season all-star team. The lefty has the ability to get major league hitters out; he just hasn't done it yet.

Using Gonsalves -- and, for that matter, Kohl Stewart -- as primaries rather than as true starters calls to mind one of Hall of Fame manager Earl Weaver's principles: The place for a rookie pitcher is long relief.

Weaver seldom -- if ever -- put a call-up straight into the Baltimore rotation. They had to earn their chance to start with some success in longer relief outings. But true long relief outings are passe today; two innings of work out of the bullpen is generally seen as a long stint. That's not what Weaver was doing with Mike Flanagan in 1976 or Dennis Martinez and Scott McGregor in 1977.

The opener, in a sense, formalizes Weaver's approach. Stewart and Gonsalves are getting their innings and their opportunity to learn how their stuff plays at this higher level. They're just not facing the first batter of the game.

Opening the opener

Well, that didn't go well.

Opening pitcher Gabriel Moya gave up a two-run homer in the first. I like Moya, and I expect him to be a quality relief pitcher, but he has now surrendered five homers in a bit more than 25 major league innings this year, and that's too high a rate. (He's not alone in this in the Minnesota bullpen.)

Primary pitcher Zack Littell didn't fare much better at the end of his stint. The key theory behind the "opener" tactic is that the nominal starter will begin his third trip through the batting order somewhere in the second, weaker, half of the lineup. On Sunday he entered to begin the second inning against the No. 6 hitter, and in four innings allowed one unearned run. His fifth inning began with that feared third trip through the order, and he gave up a double and a homer. And that was all for him.

And then Tyler Duffey was completely ineffective, and Matt Belisle got ejected, and the whole contest got out of hand.

The opener isn't a cure for bad pitching, and the Twins had too much of that on Sunday. The tactic is -- as is true of baseball tactics -- an attempt to achieve a marginal advantage. One game, one example, doesn't truly determine anything.

One thing I heard Dick Bremer say during the telecast that I wish to dispute: He said that Tampa Bay has used slider specialist Sergio Romo a lot as the opener. The Rays did use Romo five times in that role at the beginning of the experiment, but he hasn't opened since May 27.

What happened? Two things:

• The Rays traded then-closer Alex Colome to Seattle and shifted Romo back to the late innings;
• The Rays used Romo to open against two specific clubs (Angels and Orioles) that had no left-handed hitters at the top of their lineups.

The most-often used opener for Tampa Bay has been Ryne Stanek, with 22 "starts" in 46 appearances.

Notes from the Weekend

Well, that series in Kansas City should definitely put the Twins in seller mode. From Sunday morning's Twitter feed:

The Indians' chances to win the AL Central assessed at 99.6%: https://t.co/aW7oAjgFzc

— Buster Olney (@Buster_ESPN) July 22, 2018

All together now: So you're telling me there's a chance.

---

Let us not dwell on a weekend in which the Twins received few breaks and deserved fewer. Let us move on to a process item rather than results. The Twins are beginning to experiment with Tampa Bay's "opener" tactic in the minor leagues. On Sunday, for example, Trevor May -- remember him? -- pitched the first inning for Rochester, and Zach Littell followed with 5.6 innings.

Thad Levine, the Twins' No. 2 in baseball ops, said on the radio pregame show that the front office has been following the Tampa Bay experiment and sees the advantage of having a "reliever" deal with the top of the opposing order in the first inning, allowing the pitcher who is expected to work the bulk of the game to get more outs before seeing the opposition's best bats a third time. (Levine also said Littell profiles as the type of pitcher the opener would probably benefit.)

I now take the risk of parsing Levine's comments too closely. Levine said the tactic has been discussed with manager Paul Molitor and pitching coach Gavin Alston. I infer that Molitor and Alston are cool to the idea of employing the opener in the majors.

Today's game would really be a good opportunity for it. Aldaberto Mejia, today's scheduled starter, is not an innings eater. (He appears to eat pretty much everything else, but that's another matter.) The hefty lefty made 21 starts last year for the Twins and got more than 15 outs (five innings) six times. He can use the help in getting deep into games.

Plus Toronto -- at least off a spot check of recent box scores -- load the top of the their lineup against lefty starters with righties and switch hitters.

It would make sense to have someone like Alan Busenitz pitch the first inning, then give way to Mejia for the middle and bottom of the order. Say Busenitz faces four batters, than Mejia faces the fifth man to start the second. That would also be the first guy to face him a third time, and he's got a better chance to work another inning than if he were facing the top of the order for the third time.

Of course, if Molitor and Alston were interested in trying the opener today, it wouldn't be Busenitz; he got sent back to Triple A after his inning Sunday to make room for Mejia. I really do not understand the organizational fascination with keeping Matt Belisle around to be a mopup man, but that's another problem.

---

Ex-Twin watch: The Twins drafted Dereck Rodriguez, the son of Hall of Fame catcher Ivan Rodriguez, in the sixth round of the 2011 draft. He was an outfielder, but he didn't hit, and the Twins converted him to pitching in 2014.

He had his ups and downs on the mound in the Twins organization, and never made it on the 40-man roster. He left as a minor-league free agent after last season and signed with the San Francisco Giants.

And he's now in their major league rotation with a 5-1, 2.72 ERA after 10 appearances, eight starts.

San Francisco and the National League are a different environment than Minnesota and the American League, but I'd sure take that.

The Tampa Bay experiments

An intriguing statistical tidbit from the weekend (no longer up to the minute):

Your up-to-the-minute ERA leaders since the Rays started the Jonny Bullpen games on 5/19:

Rays: 2.95
Yankees: 3.03
Dodgers: 3.19
Red Sox: 3.28
Cubs: 3.37

Rays have played vs BAL, OAK, SEA, TOR, NYY, HOU, WAS in this run (collective .536 winning %). 24 last 27 vs teams > .500

— Jason Collette (@jasoncollette) June 26, 2018

Since the Rays shut out the Nationals on Tuesday, the larger point still stands. This is a team running with two traditional starters, and almost five weeks into it, with a bunch of games against good lineups, it's working.

I caught the end of the Rays' 1-0 win Tuesday, which featured some really out-of-the-box managerial strategy from Kevin Cash.

Top of the ninth, the Nats have up Bryce Harper (lefty), followed by Anthony Rendon (righty), Juan Soto (lefty) and Daniel Murphy (lefty). Four high quality hitters. Cash brings in Jose Alvarado, his top lefty reliever. Alvarado promptly walks Harper, who showed bunt on practically every pitch.

Now ... Cash doesn't want Alvarado pitching to Rendon with the tying run on, but he wants him to face Soto and Murphy. What to do?

His answer: He pulled right fielder Johnny Fields; moved Mallex Smith from left to right, shifted first baseman Jake Bauers to left field and had Alvarado play first base (with Bauers' glove), with Chaz Roe coming in to pitch. (As a sidelight to all this, the Rays lose the DH). Alvarado was laughing as he took his new position.

This kind of thing -- moving the pitcher to a defensive position for a batter -- isn't unprecedented. Paul Richards did it a few times in the 1950s, and Whitey Herzog did it in the 1980s. But they generally were putting the transplant someplace where they didn't expect the ball to go. First base may carry the lightest physical demands, but the first baseman is quite likely to touch the ball.

So the Nats announcers are speculating on whether Roe dares try a pickoff throw and if Rendon should bunt toward first and force Alvarado to make a play. There are a lot of ways for the Nationals to try to take advantage of this novice playing first base.

And they did none of them. Harper doesn't take a monster lead and steal second; Rendon doesn't bunt toward first. Rendon strikes out.

And Cash undoes all his earlier moves. Bauers reclaims his glove and returns to first, Smith goes back to left, Carlos Gomez comes in to play right, and Alvarado returns to the mound (and gets his warmup pitches). The Rays have lost their DH for the remainder of the game, but they have their favored lefty to face Soto and Murphy, so they're where they want to be.

Soto singles to center, Harper stops at second. Murphy singles to left; Harper stops at third. All this maneuvering to get Alvarado to face the lefties, and he retires none of 'em.

Now the Nats have right-handed hitters up, Trea Turner and Michael A. Taylor. Cash brings in Sergio Romo, who throws a few sliders and pops up Turner and fans Taylor, end of game.

And all the postgame attention went not to Cash's brainstorm but to the postgame yelling by Romo at Taylor, who apparently drew Romo's ire in a previous game by stealing a base. Eyeroll.

That, in my mind, is just stupid jock stuff. It's nowhere near as interesting as a pitcher playing first for a batter.