It all made sense on paper. The best pitchers should throw the most innings. There's no doubt that Daniel Bard was one of Boston's best pitchers in recent seasons. He wanted to be in the rotation, and the Red Sox had a slot to fill.
Except it's never that simple.
Bard's transition to starting pitcher has taken him from a relief ace to the minor leagues.
On Tuesday, the Red Sox demoted Bard to Pawtucket, where he will continue to work as a starter. For now.
Is the starter experiment a failure? Perhaps not quite yet. But that's only because Boston hasn't found an alternative.
Simply put, Bard is no longer one of the best pitchers on the team. He's lost his velocity, lost the strike zone and lost his spot in the majors.
Despite talk of release points and pitch repertoire, Bard's decreased velocity is the inescapable source of his woes. He simply isn't a good enough pitcher to dominate without his 98-mph fastball. At least, not right now.
Although he's tried to mix in a changeup, Bard is clearly a two-pitch pitcher. And his dominant fastball has disappeared in 2012, reducing the effectiveness of his slider.
A decrease in velocity was expected with a move to the rotation. Last season, only two starters -- Alexi Ogando and Justin Verlander -- had a fastball that averaged over 95 mph. By contrast, six relievers (including Bard) averaged over 97 mph.
However, Bard's velocity decline has been more drastic than imagined. That 97-mph fastball is down to 93.0, and he's reached 97 on only five pitches. The fastest pitch he's thrown in 2012 (98.2) is just slightly faster than his average fastball velocity in 2010 (97.8).
Manager Bobby Valentine attributed Bard's velocity issues to a "mechanical mystery." Pitching coach Bob McClure points to his arm slot.
"Coming out of spring training, he was a little higher but more accurate," he said Tuesday. "When he's lower, it's harder and not as accurate. That's why you saw the velocity down a little bit, too. He was kind of caught in between a little bit."
The result of that slower fastball? Fewer misses and fewer strikeouts. Opponents are hitting .272 against Bard with two strikes. Compare that with .125 over the first three seasons of his career.
That certainly helps explain the drop in strikeouts. After averaging 9.7 strikeouts per nine innings in his three seasons out of the bullpen, Bard has dropped to just 5.6 in 2012.
Of course, that velocity drop wouldn't be as worrisome if Bard was locating his fastball. However, only 47 percent of his fastballs have been thrown in the zone. That's 10th-lowest among starters and down from 53 percent last season. Bard has thrown the fourth-most fastballs above the strike zone, a pitch that certainly would benefit from his old velocity.
Only Ubaldo Jimenez has a higher walk rate than Bard, who has 13 more walks than he did in 2011 in 18 fewer innings.
Bard has gone to a 3-0 count 27 times this season. That's the most in the majors and one fewer than he had in the previous two seasons combined.
Bard has never thrown a high percentage of first-pitch strikes, and that has continued in 2012. The difference is that he's been unable to recover after falling behind early. Opponents have a .453 on-base percentage after getting ahead 1-0. The league average is .379.
With Bard's control issue, hitters can now be more selective. Opponents are swinging at 20 percent of pitches out of the zone, down from 27 percent last year.
Bard's woes are amplified by his previous struggles as a starter. In 2007, he posted a 7.08 ERA while walking more than a batter per inning in Class A. His career turned around with a move to the bullpen.
Bard wouldn't be the first flamethrower who couldn't cut it as a starter.
Robb Nen, once clocked at 102 mph, was an enigma coming up as a starting pitcher in the Texas Rangers organization. In 399 minor-league innings, he posted a 4.69 ERA and issued 266 walks. After running out of options, Nen was shipped to the Florida Marlins, where he moved to the bullpen.
Nen threw 697 innings as a reliever, posting a 2.94 ERA and 314 saves. That same pitcher who struggled with control as a minor-league starter had no such problems out of the bullpen. Despite throwing 298 more innings as a major-league reliever, he issued 22 fewer walks than he did in the minors.
Feliz maintained most of his velocity, losing just 1.6 mph on his fastball compared with 4.2 for Bard. The Rangers' acquisition of Roy Oswalt may ultimately push Feliz back to the bullpen, but the move won't be performance related.
Sale's fastball has had a more significant drop off than Feliz's, but his changeup is a legitimate third pitch (opponents are hitting .115 against it). With an American League best 2.30 ERA, Sale's transition represents the best-case scenario for a converted reliever.
So what happens now for Bard? Perhaps he'll fix a mechanical flaw and regain his control and velocity, and return to the rotation. A far more likely scenario appears to be a return to the bullpen.
For now, the Red Sox will likely turn to Daisuke Matsuzaka as fifth starter. It's hard to escape the irony of Matsuzaka being called upon to replace a pitcher who couldn't find the strike zone.