More on batting orders

The Monday print column explored some of the questions and issues involved in drawing up a batting order for the Twins. Here's a tangent I left out:

Which inning do you suppose has the most runs scored? For most teams -- for the league as whole -- it's the first inning. American League teams last year scored 1,326 runs in the first inning. The next highest: the fifth inning (1,308). Third highest: the third inning. And the inning with the fewest runs (excluding the ninth, which the home team often doesn't hit) is the second (1,166).

This makes sense, because the first inning is the only inning in which the lineup is set up as the manager idealizes it. And if the best hitters are arrayed at the top of the order, that leaves the weaker ones at the bottom -- and they're going to hit in the second inning. One could boost the even-numbered innings by moving some good hitters lower, but that ultimately takes at-bats away from whoever is lowered (about 13 plate appearances a year per lineup slot).

Second question: You may expect that the No. 1 spot in the order leads off the most innings, because it automatically does to start the game, and you would be correct. Which lineup spot leads off the second most innings?

The answer is No. 5. Which is interesting because the stereotypical No. 5 hitter is pretty much the opposite of a leadoff hitter. The stereotypical No. 5 hitter is a slow right-handed hitter with power but lacking on on-base skills. It's a hitter better suited to finishing trouble than starting it -- but he's often tasked with starting it. (For the 2016 Twins, the revolving door in the fifth slot combined for the second lowest OBP, better than only the No. 9 slot.)

Oddly, the 2016 Twins didn't follow the usual pattern of runs scored. Their most productive inning was the third (106 runs), followed by the fifth (95). The first tied with the seventh for fourth place (78), although the Twins lost out on a seventh inning once, so on an average basis the seventh was ever so slightly more productive.

I suspect this is connected to using Brian Dozier as the leadoff hitter so much. Their best power hitter was guaranteed to have nobody on base for his first at-bat. They had to turn the lineup over to have any chance of having somebody on for Dozier,

Batting orders probably have just a marginal effect on runs scored, but managers feed off marginal advantages. I'm not convinced Paul Molitor's batting orders last year maximized the possibilities.

More on Mauer as a leadoff hitter

A tangent I left out of Monday's print column about the Twins batting order:

Last week, when Paul Molitor unveiled his first lineup with Joe Mauer hitting leadoff, Fox Sports North studio analyst Tim Laudner hastened to assure the public that this has nothing to do with sabermetrics.

Laudner says a lot of silly things, few of them sillier than that. Of course Mauer as a leadoff hitter is sabermetrically inspired. He gets on base and he is neither fast nor a base stealer. You can probably count on the fingers of one hand the managers in the pre-Bill James era who would put such a hitter leadoff.

It's a old-school staple of long-standing: You lead off with a speedster, and you follow him with a guy who is adept at bunting and the hit-and-run. That was sensible in the deadball era, but it hung on for decades after it stopped making sense.

In the days of Johnny Evers and Joe Tinker, those moves were practically automatic; if the first batter of an inning reached base, the next at-bat would feature a steal, a bunt or a hit-and-run. Nobody bunts in the first inning nowadays, and seldom in the first seven (not counting pitchers). You might see one true hit-and-run play in a month (not counting putting the runner in motion on a 3-2 count). So why design a lineup to maximize the opportunity to use plays you're not going to use? Because that's the way Frank Chance did it? Nonsense.

It remains remarkable how many broadcasters and writers seem threatened by this kind of change. It would be understandable if Dan Gladden was; he spent his career hitting leadoff, and I doubt that would be the case in today's game, because his on-base percentages were mediocre. He was the traditional speed merchant hitting leadoff. But Laudner ... man, this evolution in the game wouldn't have affected his playing status at all.

But it certainly isn't going away. Did you notice who Terry Francona is using as the leadoff hitter for Cleveland? Carlos Santana, who is probably slower than Mauer and nowhere near as adept a baserunner (speed and savvy being two different things). Getting on base creates runs; making outs does not. That's not just "sabermetric." It's logical.

On hitting Buxton ninth

It took Byron Buxton 10.7 seconds to go home-to-third
on his eighth-inning triple Monday night.

Bill James once observed that there are more than 269 quadrillion possible lineup combinations out of a 25-man roster -- 741 billion possible sets of nine players out of the 25, and 362,880 options for batting order for each of the 741 billion. And that's not getting into defensive alignments.

The numbers start shrinking once you start ruling out having Trevor May, Glen Perkins and Mike Pelfrey in the lineup together, of course. But 269 quadrillion is a humbling number. Writes James (in his Guide to Baseball Managers):

The first point to make ... is that we can never really be sure what the optimal lineup is. The number of options is so large that it overpowers even the largest and most sophisticated computers. The only way to approach the problem is by whittling down the theoretically possible selections into those which seem reasonable, and then evaluating what seem to be the prime alternatives.

On Monday night, Paul Molitor hit his pitcher (May) eighth and Byron Buxton ninth. Buxton, of course, is as highly regarded a prospect as they come; May appears to be about as good a major-league hitter as I would be. Yet Molitor had reasons for putting Buxton ninth, and I assume those reasons will prevail again today at the Twins wrap up this visit to a National League venue.

The day will come when we look back on Buxton's arrival and laugh at the idea of hitting him ninth, But right here, right now, Molitor has two missions: Win games, and break Buxton into the majors. He calculates that he serves both by playing Buxton but limiting his exposure to major league pitching. And until Buxton starts recognizing breaking pitches out of the zone and laying off them, Molitor is absolutely right. (Buxton's triple Monday came on a fast ball, and I bet he doesn't see a fast ball strike this afternoon.)

A passing of the baton

Oswaldo Arcia accepts the congratulations of
third base coach Joe Vavra as he rounds the bases with
his sixth-inning solo home run Tuesday.

The Twins lineup the past few days has had the air of a new era — even beyond the absence of the injured Joe Mauer and the traded Justin Morneau.

The top half of the order was the same Monday and Tuesday: Alex Presley, Josmil Pinto, Brian Dozier, Oswaldo Arcia.

Three of those four opened the season in the minor leagues. The other (Dozier) opened the year at the bottom of the batting order.

And now they're filling the prime lineup spots — not only against another also-ran, as on Monday, but against a legitimate contender as well.

Obviously, Mauer's return, whenever that comes, will shake this up a bit. Maybe Mauer will hit second; maybe he'll hit third. (I'd prefer he hit second.) Presley figures to remain the leadoff guy; Arcia should be locked into the cleanup spot. Pinto, by process of elimination, will probably drop in the order.

But he'll probably remain higher in the lineup than the likes of Ryan Doumit and Josh Willingham — and he should. The veterans who hit in the middle of the order all year didn't produce. Now the kids get their turn.

Best man hits second. It's coming. It's here.

During Monday's FSN telecast of the Cedar Rapids Kernels game -- aka the Byron Buxton Show -- the studio boys spent some time talking about batting order position in the context of Buxton, who has been used strictly in the leadoff slot this year. Is he a leadoff man in the future, or a No. 3 hitter?

Tom Kelly's response combined common sense (basically, we don't really know yet what form his talents are going to take) and the outdated ("You don't want to hit him second").

Well. I suppose TK can't really be expected to be keeping up with always-evolving managerial strategies, but I'm comfortable with making this assertion: In five years, it will be as much the norm for teams to slot their best hitter second as it is now for teams to reserve their best reliever for save situations. And it will make a lot more sense.

There's resistance to this change, but it's already taking place, with the Twins and elsewhere. (Jim Margalus of the South Side Sox blog did a nice job this week dissecting the Hawk Harrelson-Steve Stone ridicule of the Blue Jays for hitting Jose Bautista second.) Ron Gardenhire went away from hitting Joe Mauer second for a couple of series, then switched back, presumably because he discovered that, yep, he was losing games while Brian Dozier got more at-bats than Mauer.

I don't know today who will be the Twins' best hitter five years from now. It might be Buxton. It might be Miguel Sano. Possibly it will be Oswaldo Arcia; it might even be a 35-year-old Mauer, although that seems unlikely given expected decline with age and the sheer talent coming up.

But I guarantee you: The expectation by then will be that the best hitter hits second.