Setting the bar for defensive switches

Today the always-interesting Derrick Goold got me thinking some with this story:

As Cardinals manager Mike Matheny has canvassed coaches and managers in the Cardinals system and friends who are with other clubs, he has tried to gather advice and opinions on one of the trickier calls he and Don Mattingly have both made in this series.

“Talk to me about improving your defense late in the game,” Matheny says.

Late in Game 4, Matheny replaced third baseman David Freese and then switched left fielder Matt Holliday out of the game with nine outs to get and a two-run lead. The risk to improve the defense was being without two righthanded bats, including No. 3 hitter Holliday, if the Dodgers rallied to tie. The pitcher in the third spot would also neutralize No. 2 hitter Carlos Beltran, who would be intentionally walked to face the pitcher or pinch-hitter...

--snip--

“There are times when I want the middle of my order,” Matheny said. “Especially when you’re talking about replacing the 3-4 hitters in your lineup. That’s a big move. You’re taking that risk of the offensive spot coming back around and you not having your key guys.”

Hey, let’s give Matheny some credit for asking for some help on this on. He seems like a genuinely curious fellow. And as Joe Sheehan was kind enough to explain to me, Matheny slashed the Cardinals’ non-pitcher sacrifice bunting by roughly 50 percent this season, his second as a manager. It’s not that Matheny doesn’t have more to learn -- as Sheehan also points out, Matheny’s got a maddening habit of sticking with his starting pitchers too long in close postseason games -- but he doesn’t seem impervious to the cold logic of numbers.

He’s not going to find any numbers by canvassing coaches and managers and friends with other clubs, though. Not many numbers, anyway. Have you ever seen the numbers on this? I can’t find anything. Here are a few thoughts that come to mind, though ...

If you’re ahead and it’s highly unlikely that your sluggardly slugger is going to bat again, then why not make the move? And I don’t see that it matters if you’re ahead by one run or four. If you’re ahead by just one run, there’s obviously a better chance the game will wind up being tied ... but doesn’t that also mean there’s a much better chance of the improved defense making a difference?

At the same time, the spot coming around must be highly unlikely ... because isn’t it manifestly true that one plate appearance is more important than the prospect of some key fielding play? Or plays? We know that hitting is roughly 50 percent of the game, and Bill James has suggested that defense -- not including pitching -- is around 15 percent. Assuming those figures are at least approximately accurate, doesn’t it seem crazy to trade one plate appearance by a poor-fielding good hitter for two innings of a good-fielding poor hitter?

And if the game goes extra innings, it might be more than one plate appearance.

Obviously, the equation is never so simple. It depends on the pitchers, and especially on the relative abilities of the sluggardly slugger and the defensive specialist. But it seems to me that while improving your defense in the late innings is obviously a worthy goal, the bar should be pretty high. Because if it’s close, I don’t want my key guys losing at-bats. Ever.