I had a lead all ready to go about how the Braves were one of the few teams that might be able to justify playing Ken Griffey Jr. in left field, how the .270/.360/.460 line they could reasonably expect from him in a strict platoon would be enough of an upgrade over their available options to allow them to suffer his defense. There was a paragraph about how a left-field platoon of Griffey and Matt Diaz, with Josh Anderson or Brandon Jones serving as the combination’s legs, would be comparable in value to what the rest of the division will get from left field.

With Griffey electing to play in Seattle-a match that makes little sense for either entity-the Braves are back to where they’ve been all winter: searching for production in the outfield that matches what they’ll get from the infield. This is an important issue for them, because the rest of the team is strong enough that what the Braves get from their outfield could be the difference between getting back to the playoffs or missing them for the fourth straight season.

They have no one to blame but themselves, of course. Adam Dunn and Bobby Abreu would have been fantastic and affordable solutions to the problem, and they were both sitting on the market like bruised tomatoes until just a week ago. The Braves, having signed Derek Lowe and Ken Kawakami and traded for Javier Vazquez, were reluctant to add more payroll, especially given the likelihood that the $13 million paid to Tim Hudson this year will largely be wasted. The Braves are looking at an Opening Day payroll of around $88-90 million, Hudson included, which would be in line with their 2004-07 payrolls. (The 2008 payroll was over $100 million in large part because of the Mark Teixeira contract.) Bumping that by 8-10 percent to bring in one of the impact corner outfielders may have seemed excessive, but the marginal value of that outfielder could be very high.

Remember the lessons that Nate Silver has offered: the value of wins is not linear, and where the Braves currently sit, projected to win 87 games, is the sweet spot where the marginal value of each win is astronomical, peaking at $4.5 million per win on the 90th (all data from Baseball Between the Numbers). If either Dunn or Abreu were to add two wins in that range above the current left fielders, they would pay for themselves.

Well, “if” seems like a silly word. Projected weighted-mean WARP for the players who might have been the Braves left fielders, and for the ones who will be:

Dunn  3.1
Abreu 3.2
Diaz  0.8
Jones 1.1

The combination of Diaz and Jones projects to 1.9 wins, but that’s deceptive, as it involves double-counting some playing time, giving the two players a combined 711 plate appearances, as opposed to the 598 and 574 PECOTA lists for the first two. Moreover, you’d rather have value concentrated in as few roster spots as possible. You would think that the difference between star-caliber players and fringe guys would be higher, but Dunn and Abreu are both poor defensive outfielders. We’re also seeing the effect of the higher replacement level in Clay Davenport‘s new system. Still, at minimum you would be talking about a 1.5-win gain, along with some certainty about performance and the gain of a roster spot. Not signing either outfielder may turn out to be a costly mistake.

The key is to not exacerbate that mistake by making a bad signing. Picking up Griffey would have been marginal-Griffey’s projected WARP is a mere 1.1, exactly that of Jones-justified by the lower cost and the possibility that he would have value in a strict platoon. With the Hall of Famer gone, talk has turned to Garret Anderson, who is something of a last man standing in this market. Anderson is durable and little else, with his power mostly gone and his OBP a perpetual problem. He projects to a .274/.319/.440 line, with slightly below-average defense and a value of about one win over replacement. He is no better than Jones at this point, and should not be signed to replace the younger player. The differences between Griffey and Anderson are OBP and the platoon split, neither of which favor Anderson.

I don’t mean to peg the Braves as penurious overall, just to point out that they stopped spending money at exactly the point where they might have gotten the most bang for their buck. Buying a left fielder like Dunn would have helped make the significant investments in the rotation-more than a third of the payroll is tied up in the three new starters-pay off. This is a better team because of the moves the Braves made this winter, acting to stabilize a rotation that didn’t provide innings a year ago and put pressure on an inadequate bullpen. Adding Vazquez and Lowe to replace the unready Jo-Jo Reyes and Charlie Morton is a massive upgrade, more than enough to offset some anticipated decline by last season’s surprises, Jorge Campillo and Jair Jurrjens. Kawakami won’t match Tim Hudson’s 23 starts, but Clay Davenport‘s projection was that the 33-year-old Japanese righty would be “better than an average National League pitcher.” The Braves’ rotation should be set to help the team lop as many as 40 runs off of last seasons’ 778 allowed. Sentiment aside, Tom Glavine is done, and should not be allowed anywhere near this rotation. Every start he gets will push the team further from the postseason.

The bullpen is once again high on upside and low on reliability, Mike Gonzalez and Rafael Soriano would be a devastating duo at the back end of games if they were ever both available at the same time. The most innings the two have ever combined to throw in any season, as teammates or otherwise, is 114. They tossed a combined 47 2/3 last year, after throwing 89 in 2007. The Braves need 140 innings from the two of them, as the drop-off from the pair is significant, especially on the left side. The idea is that Vazquez and Lowe can get more games to Manny Acosta, who hands them off to Gonzalez and Soriano, limiting the number of pitchers the Braves have to use in high-leverage situations.

What makes these discussions relevant is the Braves’ infield, which is clearly championship-caliber thanks to Chipper Jones‘ late-career surge, an underrated Kelly Johnson, and the best player no one knows about, Brian McCann. With Vazquez and Lowe added, the Braves have a core of players you can build a successful team around, with the main problem being that they go up against two of the best five-man cores in the game in the Mets and Phillies. That’s why they can’t afford to make mistakes in assembling their outfield.

In the absence of a power-hitting left fielder, the Braves would probably be well-served to try and make run prevention their main goal. Their top five-Johnson, Yunel Escobar, Chipper Jones, McCann, and Casey Kotchman-should be good enough to give them a league-average offense. Rather than put Brandon Jones in over his head as the big half of a platoon, what the Braves could do is play Josh Anderson in center field and Gregor Blanco in left, giving them the best defensive outfield in the NL. Anderson can’t hit, and Blanco can’t hit well enough to carry left field, but combined with Jeff Francoeur, the trio would cover broad swaths of ground, enough to make Javier Vazquez a very happy man.

There’s another reason that I think this might work. If you consider the teams that have surprised us recently, it’s all teams that did so by preventing runs with great defense. The 2005 White Sox, the 2006 Tigers, the 2008 Rays… these were all teams that had good pitching staffs backed by excellent glove work, and which rode a terrific performance on balls in play to the World Series. It seems like defense is more affordable than offense, and more likely to be the backbone of a surprise team. When was the last time a team surprised us by bashing its way to the postseason? The 1993 Phillies, perhaps? Maxing out the defense is the best use of the Braves’ current resources; their left fielder and center fielder aren’t going to hit no matter who plays, but they could combine to be outstanding defensively.

Whatever they do, this is going to be a good team and a real contender. The 2008 season was a blip caused by a terrible run of injuries and some awful fortune in close games. Unless both make a reappearance, the Braves will be playing meaningful games deep into September-and possibly October.