Dipping Their Toe In Knowledge: The Kansas City Royals Shake Up Their Lineup

In 2005, after 35 years as a Kansas City Royals fan, I renounced my affiliation and took up with the Washington Nationals. It had finally become evident that while I cared for the team, the owner, David Glass, did not. He operated the franchise like a WalMart, of which he had been CEO, which is to say his primary concern was earning a profit.

So Glass sucked the franchise dry and ran it on the cheap, relying on low-cost young talent to induce the flipping of turnstiles before earning their freedom.

In the ensuing eight years, KC is 506-685, a .425 winning percentage. This year, their 25-32 record is mostly attributable to their work in the batter's box: seven of their 11 most-used batters sport OPS+ of 80 or less. In other words, much of their starting lineup is at least 20% worse than average, including keystoner Chris Getz, who, at .207/.267/.289, is only the second-worst regular in the lineup. 

The entire team has struck 30 home runs, three fewer than the Orioles' first baseman and shortstop.

Their manager, lightly-regarded Ned Yost, has not wasted any time perusing the newest research on game strategy. So, for example, he's given sabermetric punching bag Jeff Francoeur 167 plate appearances despite a predictable .213/.246/.319 performance. (Actually, it's not the batting average or lack of power that are so lamentable, since both of those can increase over the course of the season. It's the 3% walk rate, which tends not to change much over time and which would keep his OBP below average for an outfielder even if he batted .300.)

When it comes to lineup construction, it's no surprise that Yost has clung to the old ways. Before there was research that could muddy the waters with annoying "facts," managers arranged lineups by folk wisdom, some of which proved wanting. They batted a punch and judy lightning bolt first, a speedy bat handler second, the best hitter third and the slugger fourth.

Hundreds of studies later, covering tens of thousands of games, there is a whole matrix of rules for optimal lineup construction, which can be summarized this way:
1. Bat your best hitters at the front of the order so they get to the plate more.
2. Table-setters in the first two positions need high on base percentages.
3. Alternate handedness as much as possible.
4. Beyond that, it doesn't matter much.

Managers still walking among the great lizards fail the test most often with respect to Rule 2 (and by extension Rule 1). For example, throwback strategist Dusty Baker bats shortstop Zach Cozart and his .270 OBP in the two spot, costing the Reds bits of runs every game.

That's exactly the error Ned Yost makes, albeit with fewer options than Baker has. Yost hits the Royals' best player, Alex Gordon (.315/.360/.468), at the top of the lineup and then follows him with speedy, punchless shortstop Alcides Escobar. The Gordon choice is defensible because he's likely to author the top OBP on the team. 

As for Escobar, remember him from his Milwaukee days? His Kansas City glory? I didn't think so. He might not be the worst hitter in the Royals' starting lineup, but if you made the case you wouldn't get an argument. That's the guy Yost sends to the plate the second most often. 

But! Earlier this week, Yost finally caved to the seamheads employed by the team (though one wonders why it bothers) and allowed them to re-arrange the order. Against a RHP, they dropped Escobar to the nine hole and moved first baseman Eric Hosmer, a .273/.332/.440 lifetime hitter against righties, up to second.

When asked whether he'd extend the experiment following a four-run effort the first night, Yost replied, "We'll see how it works tomorrow." (They scored seven the following night.) Yost is understandably wary, in as much as the new lineup represents a repudiation of everything he's believed for all his decades in cleats, but even a stubborn innumerate should be able to understand that a game or two is an insufficient sample size for testing a strategy that might boost scoring by a run every few games.

Obviously, optimal lineup construction is only optimal for the lineup in question. No batting order rife with dead weight, like KC's, can be transformed into the '27 Yankees simply by rearranging the uniforms. The improvement on offense that a more rational batting order conveys might bump the Royals to .500, but no further, and only if they improve elsewhere as well. But the decision to start listening to what the research says could transform the entire organization and produce occasional contenders.

The likes of Ned Yost probably won't be there to experience it. And I'll stick with my Nats all the same.

The Order of Things

Tom, a Met fan from Charleston, asks how much the Mets are hurting themselves by batting Jose Reyes first and Luis Castillo and Alex Cora second?

In one sense, the Mets are killing themselves by playing a middle infield that can't hit. Reyes (.222/.266/.284), Castillo (.250/.338/.292) and Cora (.239/.325/.313) are all offensive weak knees so far this year, their other skills not withstanding. Reyes has shown some life the last week, but living to 100 may not be enough to lift his performance to respectability. (Cora mostly spots Castillo at second -- both on the field and in the batting order -- so I treat them as one entry.)

Generally, batting order doesn't much matter.  Our baseball noggins have been wired to conceive of lead-off hitters as skeeters, #2 batters as bat control gurus and the middle of the order comprising the best hitter, the cleanup hitter and the secondary slugger, respectively. Research shows there's nothing sacrosanct about this, or any other, configuration.

There are, however, three concepts that do matter in a lineup:

1. Your best hitters should bat most, so write small numbers next to their names.

2. Alternate handedness to mitigate the value of pitching changes.

3. Lead-off batters matter because they're up with no one on base at least once every game.

It's here that Jerry Manuel has handcuffed his offense. Reyes may be a hummingbird on the basepaths, but he puts the Mets in a one-out hole nearly three-quarters of the time. Castillo/Cora comes to the plate more often than David Wright, Jason Bay and other players who can actually hit major league pitching. Over a season, batting second will add something like 50 plate appearances versus batting eighth. Castillo this year and Cora this lifetime have "batting eighth" written all over them.

Except, then where would Jeff Francoeur bat? For a guy with his athletic ability, Francoeur is cheap beer, and his expiration date has passed. Despite his speed, power and cannon arm, Francoeur's complete lack of plate discipline will soon end his major league career. He has now followed last year's .280/.310/.425 with .211/.273/.362 so far this year. If you are what your record says you are, Frenchy is rated Triple-A.

In short, the Mets have myriad woes. The order of their lineup isn't one of the big ones, but installing poor hitters at the top of the order is bad medicine, even if they're slap-hitting speed burners.

b