It’s amazing what a general manager can accomplish when desperation drives him to forget all the things about which he’s sure and to make a move that can’t possibly work. We often think of a GM’s mission as a quest for the best players at each position, as if he spends the hours between 9 a.m. and 5 p.m. deciding whether to trade for Manny Ramirez or Mark Teixeira. More often, though, the choice is not whether to get this or that star, but whether to stick with this Sean Casey or pursue that Scott Hatteberg — the far more challenging task of trying to distinguish the subtle differences between journeymen.

Generally, these kinds of decisions are only made when a team has bottomed out a position, having run through all the in-house veteran and prospect options (if the team has either), because even the most adventurous GM doesn’t want to stake his season on a player like Jack Cust. Billy Beane had to reach for the defensively challenged, strikeout-prone outfielder when whole swaths of his team hit the trainer’s office, never to reemerge. A 28-year-old, former first-round draft pick, Cust had been stuck in the minors since 1997, failing to gain traction in several, brief major league trials. When Mike Piazza went down, Cust took over as the primary designated hitter. In the roughly seven weeks since then, Cust has given the A’s an unexpected, perhaps season-saving lefty power source in the mold of a latter-day Jason Giambi. It’s worth noting that although Cust is streaky and positionless, his 11 home runs would be the second-highest total on the Yankees.

Similarly, the Braves now find themselves frequently listing Willie Harris as their starting left fielder. The speedy utilityman, 29, had long ago seemed to prove himself incapable of hitting in the majors, having batted .238 AVG/.306 OBA/.294 SLG in roughly 1,000 career plate appearances entering this season. He spent most of last season with Triple-A Pawtucket, batting .220. The Braves signed him last winter, tinkered with his swing, and saw him hit .362 at Richmond. When Ryan Langerhans opened the season in a 3–for–44 slump (.068), they decided to give Harris a shot. The odds were that he would fail nearly as badly as Langerhans had — Harris had just 58 at bats at Richmond, a sample small enough to suggest either that he was having a hot streak, or that an inordinate number of outfielders were tripping over their shoelaces while chasing his fly balls. Still, John Scheurholz and Bobby Cox took the chance, and Harris has rewarded them with .383/.440/.500 rates in 45 games.

Will Harris keep up the heat? Almost certainly not; even the Tony Gwynns of the world fail to hit .380 for very long. When Harris’s average starts to slide, the Braves will have to make another change, because with little power, if Harris isn’t hitting .330 he’s not helping. But that’s not important. What matters is that the Braves didn’t let their Langerhans problem fester. They attacked it, however unlikely their proposed solution. As Franklin Roosevelt said in justifying the countless number of federal programs he threw at the Great Depression, “It is common sense to take a method and try it. If it fails, admit it frankly and try another. But above all, try something.” Roosevelt died in 1945, but you’d swear he was trying to say, “It just doesn’t look like Carlos Delgado is going to pull it together. What else you got?”

Not all of these desperate patches work. In fact, more often than not, they don’t. General managers are further hindered by guaranteed contracts. Even if the Mets really felt it was time to give up on Delgado, or were able to swing a deal for a replacement, there’s a minimum of $20 million — next year’s guarantee, plus 2009’s buyout — that makes the case against it. Still, that’s not a coldly logical argument since the money is gone whether or not Delgado plays. It’s easy for an outsider to argue that the Mets should treat the dough as a “sunk” cost and do what it takes to win, but it’s harder for Omar Minaya to make that point to his owner.

Under more financially fluid conditions, though, players can make the jump from the discard pile. The Devil Rays have scored with 26-year-old shortstop Brendan Harris (.311/.369/.483), and 29-year-old Carlos Pena (.293/.393/.633). This is the same Carlos Pena the Yankees refused to try last year, in spite of their first base situation being nearly as desperate then as it is now.

Perhaps the biggest leap from afterthought to essential has been made by outfielder Reggie Willits of the Angels. Willits is not an Angels sort of player: He walks, something that few Angels do. Since 2003, no Angel has taken even 70 walks in a season, and as a result the Angels offense has rarely been above average in the Mike Scioscia era. Willits, who was a .300 hitter in the minors, had a tenuous grip on a bench job this spring, but when Garrett Anderson went down in late April, the 26-year-old was promoted to the starting lineup and literally ran with the job, batting .337/.432/.401 with 18 steals and seizing the leadoff spot from Garry Mathews Jr. He is second on the team in walks with 36; Vladimir Guerrero leads, but only because pitchers have intentionally walked him a ridiculous 20 times.

Anderson is the antithesis of Willits, impatient and slow. He hasn’t been an offensive asset in five years. Yet, when Anderson came off the disabled list earlier this month, Willits found himself back on the bench. Anderson rescued the Angels from their foolish loyalty by immediately injuring himself again. To paraphrase the Rolling Stones, you can’t always get what you want, but sometimes you get what you need — if only you’re willing to look for it.

Mr. Goldman writes the Pinstriped Bible for yesnetwork.com and is the author of “Forging Genius,” a biography of Casey Stengel.