Manager Fredi Gonzalez of the Atlanta Braves makes notes in the dugout during the first inning of their game against the Washington Nationals at Nationals Park in Washington, DC. (Photo by Jonathan Ernst/Getty Images)
Nearly two full seasons into his tenure as Braves manager, Fredi González finally seems to have embraced some basic tenets of sabermetrics. So what took him so long?
When Fredi González was managing the Marlins, O! How we loved him!
In theory, anyway. González loved to talk about sabermetrics. There were actual books on his desk with numbers in them. Also on his desk: a Bill James bobblehead. Every so often, González would compulsively reach over, Queeg-like, and give li'l Bill's head a tap.
Okay, so I made that last part up. But Fredi González was supposed to be our guy ... except he wasn't, really. It's not that González ever seemed like a bad manager, really. But he seemed pretty conventional, except for using his best relievers slightly more, last season, than most managers would. But when that didn't lead to a positive outcome, he ratcheted back their workloads this season (as I recently wrote about).
Lately, though, as David O'Brien writes, González hasn't been conventional at all:
In the past week Fredi Gonzalez has made three moves that he had never made before, indications the Braves manager learned from last September’s collapse not to stay with the horse he rode in just because it still has a pulse.
The third such move came Sunday, when Gonzalez had closer Craig Kimbrel pitch the ninth inning of a tie game on the road against the Mets at Citi Field. Kimbrel struck out two in a scoreless ninth and Brian McCann’s sacrifice fly in the 10th brought in the go-ahead run for the Braves in a 3-2 win for a three-game sweep and five-game winning streak.
First, three unprecedented moves in one week is a lot. Via O'Brien, those moves in order:
- Installing, unofficially, Martin Prado at shortstop, because he's a much better than the incumbent, Paul Janish, even though Prado's considered (with cause) a weak fielder. Janish, of course, is one of the worst hitters in the major leagues, with a sub-600 OPS in 1,158 career plate appearances.
- Last Wednesday, González used Kimbrel for a four-out save. This was literally the first time in Kimbrel's two-season tenure as close that he'd been asked to do that, the first time this season that he'd recorded more than three outs in an appearance, and just the third time in his career he'd gotten more than three outs.
- And of course Sunday. According to O'Brien, last year González said he would never ask Kimbrel to pitch in the ninth inning of a tie game on the road. In (dare we say it?) a non-save situation.
And then he did. From O'Brien, again:
"It was one of those situations, [pitching coach Roger McDowell] and I talked it over," Gonzalez said of using Kimbrel in the ninth. "They’ve got Davis and Duda coming up, and really you feel comfortable – it’s first time I’ve done it, bringing closer in a tie game on the road – but you feel with the way our pitching matched up later in the game, that that was the game right there."
With the bottom of the ninth coming up, here were the Mets' next six batters, along with OPS:
Davis and Duda haven't exactly been world-beaters this season ... but a) they rank first and third on the Mets in home runs, b) Kimbrel's exceptionally stingy when it comes to home runs, and c) those first three hitters, while not MVP candidates, are significantly better than those next three.
Well, better anyway. Certainly more powerful. In the event, the pinch hitter was Jordany Valdespin, who's actually having a pretty good hitter. But there's no question, if you were playing a computer simulation you would want your best relief pitcher pitching in the ninth, just to keep the game tied.
Now, most of acknowledge that managing a baseball game with real players isn't the same as managing a baseball game with electronic or Strat-O-Matic simulacrums. But I suspect that most of us do believe that a manager should occasionally deploy his best relief pitchers because of match-ups rather than whether or not it's a save situation. I suspect that most of us do believe a manager should occasionally ask his best relief pitcher to get four outs rather than three. I suspect that most of us do believe a manager should occasionally trade good fielding and bad hitting for good hitting and bad fielding.
To this point, González has hardly done anything revolutionary. Martín Prado has started four games at shortstop in the last week, and rookie Andrelton Simmons -- who can, it seems, both hit and field with great skill -- is expected back in the lineup soon. González has used Kimbrel less than utterly orthodoxically exactly twice all season.
I doubt if González has suddenly got religion. I suspect that he's always known what he should do, but was simply afraid to flout convention. Most men are. Recently, though, his fear of the unconventional has been trumped by his fear of losing his job, which is probably what will happen if his team fails down the stretch in a second straight season.
Maybe you can't manage like this all season long. Maybe the players can't handle it. Maybe you can't handle all those questions from the pesky writers. But this is how you should manage when one or two runs might be the difference between playing into October, or not.