As I write this, the Diamondbacks are using Jose Valverde to "protect" a three-run lead in the bottom of the 12th inning, after they threw out their second-line relievers for four innings of a tie game.
I, and many others, have written on the subject of reliever leverage in the past, but this is a very egregious example. This is a critical game between two contenders in a division race that could very well go down to the wire. To even compare the relative importance of the two situations--tie game vs. three-run lead--is a slap in the face of the tie, which towers over the three-run lead like Randy Johnson to David Eckstein. But David slew the Big Unit with his trusty sling, and the save rule dictated the Diamondbacks' usage of their relief ace. Again.
I get that relievers want to collect a bunch of saves. It helps them collect big contracts, and makes them feel like the king of the pitching staff. But aren't they just as excited to pick up a personal W, even if they're in the dugout rather than on the mound? As Bill James pointed out, ideal relief pitcher usage might prevent your closer from racking up 35 saves, but it might give him 20 wins. Even in the sabermetric age, teams are willing to pay for gaudy won-loss records. Kenny Williams traded Keith Foulke for Billy Koch, and it's hard to argue he didn't at least consider Koch's 11 wins the year before, many of which he earned by entering a tie game and shutting down the opposition long enough for the A's to plate a run.
The intelligence gap between the A's and the rest of MLB manifested itself again the next year when Foulke won 9 games as the A's "closer." Huston Street won two games in the season's first two weeks, and could end the season in double digits.