If they keep it up, the Marlins bullpen could cost Josh Johnson millions. Johnson is presently 4-1 with a 2.66 ERA, but he should have more wins than the four he has to his credit. That’s because Johnson has made quality starts in nine of 11 outings. His record doesn’t reflect just how well he has pitched, in part because of middling run support (4.3 runs per game), but in larger part due to his bullpen.

On April 24, Johnson started against the Phillies and pitched seven shutout innings, leaving the game with a 3-0 lead. When closer Matt Lindstrom came in for the ninth, all hell broke loose-the Phillies plated seven runs, climaxing with back-to-back jacks by Shane Victorino and Chase Utley. Johnson kissed a sure win good-bye. Almost a month later on May 19, Johnson left the bullpen in a tough spot, taking a 2-0 lead into the top of the seventh inning against the Diamondbacks at Florida, but he allowed a home run to Mark Reynolds to make the score 2-1. An error followed by a single and a sacrifice bunt placed runners on second and third with one out; Marlins manager Fredi Gonzalez figured it was time for Renyel Pinto to save the day, but Pinto hit the first batter he faced, then allowed consecutive singles, and by the D’backs were done, they had batted around and led 5-2. Johnson took the eventual loss. In Johnson’s next start, an interleague matchup against the Rays, he pitched seven innings and was leading 4-3 when he yielded to Leo Nuñez, who promptly allowed a game-tying home run to Jason Bartlett. Johnson’s win disappeared, and the Fish went on to lose the game in extra innings. Without these unearned no-decisions, Johnson would presently be 7-1.

Johnson’s one of the unluckiest of unlucky starters, joining Paul Maholm of the Pirates and Roy Oswalt of the Astros as the only pitchers in the majors this season to have three potential wins squandered by their bullpens. Twenty-six other starting pitchers have twice exited games with the chance to be credited with a win had the bullpen held on, only to receive a no-decision as the leads were lost. Overall, there have been 70 such games in the National League, with the Astros and the Marlins tying for the lead with seven such games each. There have been 57 games in the American League, with the Mariners and Indians leading with seven times each. In contrast, starters for the Reds and Red Sox have only been victimized once.

Last year, there were 193 starters’ wins washed away by bullpens in the AL, 223 in the NL. Leading the way in the AL were the Mariners, who saw the bullpen snatch away 21 wins (with Jarrod Washburn losing seven potential wins). At the opposite end were the Yankees, whose pitchers kept what they left with in all but seven games, and no starting pitcher lost a win to his pen more than once. In the National League, Mets‘ starting pitchers were pulled with a lead only to be burned by the pen 24 times, while the Pirates and the Phillies had the fewest such wins stolen, with ten apiece.

A few of these vanishing decisions would have changed a few careers and altered awards balloting. If the Mets bullpen had just held fast last year, Johan Santana might have won 23 games, not 16; if his career stalls out at 293 wins, remember this fact. Beyond Santana, 30 pitchers left games in a position to win four or more times, but their relievers couldn’t hold on, including Jon Danks, who with six lost wins restored would have gone 18-9 instead of 12-9, Randy Johnson, who with his four missing wins from last season would currently be looking for win 304, not 300, and NL Cy Young winner Tim Lincecum, who might have had a for-the-ages 24-5 season instead of a merely very good 18-5 season. Turn this into a historical question, and along those lines, note that Tommy John (288 wins) lost wins to the bullpen 45 times, Jim Kaat (283) lost 35 potential wins to his bullpen, Bert Blyleven (287) lost 43, and Mike Mussina (270 wins) lost 32.

Ironically, blown wins for the starters doesn’t necessarily indicate a terrible bullpen. The Nationals have a pen that might someday be listed among the worst in history, has taken away only five wins from its starters, two fewer than the pens of the Astros and Marlins. While the Astros’ bullpen, as measured by expected wins added over replacement level, as well as their rate of conversion of save situations into saves and holds and their percentage of inherited runners allowed to score, has been exceedingly poor, the Marlins’ pen is guilty of little worse than mediocrity. Meanwhile, the Brewers‘ pen, which rates fifth in the majors in expected wins added, has lost as many wins for its starters, five, as have the Nationals. That said, where there’s smoke there is often fire, and the many wins lost to starters by the Mets last year was certainly a symptom of a very bad pen, just as the Yankees having just seven of those games was indicative of a very good one.

Individual pitcher wins don’t matter much, of course-it’s team-level wins that count, and if a team wins despite a starter or two feeling jobbed it matters little. However, as much of the baseball community still values individual won-lost records, these affirmative decisions lost to the bullpen are important for the way they frame the discussion of leading pitchers. With his three wins back, Josh Johnson would not only have an ERA that ranks in the NL’s top 10, he would be tied with Santana for the league lead in wins, the combination identifying him as a star instead of just another semi-anonymous Marlins hurler.

Then again, if Danny Haren had received anything like real run support from the Diamondbacks instead of the paltry three runs per game he’s had to work with so far, he might be 10-1 instead of 4-4. As with Johnson, his record reflects the shortcomings of others. That is the real lesson of the no-decision kings: the win is an inadequate description of a pitcher’s performance.

A version of this story originally appeared on ESPN Insider Insider.