Hector Carrasco helped the Nats play spoilers and shut the Marlins out tonight, 4-0. Carrasco went six innings and struck out six. This was Carrasco's fourth start and he hasn't given up a run in the last three outings and 17.2 innings.

That's not bad for a guy who never started more than a game in any of his previous nine seasons. As of two weeks ago, he had pitched 557 major-league games and started just two. He has gone four starts consisting in 22.2 innings and he owns a 0.79 ERA as a starter.

Carrasco is 35 and has had more address changes at the major-league level (8) than starts (6). He hasn't had a season with an ERA under 4.00 since 1996. But what if he turns into a legitimate starter?

How unprecedented would such a conversion be?

I looked it up.

I looked for all pitchers who were predominately relievers (no more than 9 starts in any season and no seasons with more starts than relief appearances), who became primarily starters after turning 35. (That is, at least 10 starts in any season or more starts than relief appearances in any given season.)

I found nine (in reverse chronological order):

Paige barely qualifies since his season as a starter consisted in one game as a promotional stunt for the Kansas City A's in 1965.

Wilhelm started 10 games in 1958 (out of 39 in total), 27 in 1959 (out of 32), and 11 in 1960 (out of 41). He had never started a game in his first six seasons—he was a late bloomer all around—and 361 games. After 1961, he started just four games and went into the Hall as a starter.

Lindell was actually an outfielder who pitched 55 games in his career. He pitched one season as a pitcher prior to turning 35. It was 1942, when he pitched 23 games, all but two in relief, for the wartime Yanks at age 25. He didn't pitch again until 1953, his last season in which he converted to pitcher predominately and in which he started 23 of 32 games for the Pirates and Phils and registered a 4.66 ERA.

Prim was a World War II replacement pitcher for the Cubs 1943-46. He had pitched for the Senators and Phils eight years earlier. His only year as a starter was 1945 (19 of 34 games were starts) and he amassed a 13-8 record with a 2.40 ERA. But then the regulars returned….

Logan was another wartime replacement. He had pitched four seasons previously with 23 career games, none of which were starts. In 1945, he went 7-11 in 25 starts and 9 relief appearances and had a 3.18 ERA. Then the regulars returned, and he never pitched at the major-league level again.

Barrett was, surprise, a wartime replacement. He had a nine-year break from the majors, he started 24 games (of 38 in total) in 1943, and then the regulars returned…

Niggeling was a late bloomer who went on to win 64 games mostly as a starter.

Heving had one stint primarily as a starter with the Red Sox in 1938 (11 starts in 16 games). He went 8-1 with a 3.73 ERA. He never started more than 7 again.

Betts had a seven-year break and became a starter with the Braves in 1932. He had a six-year run with the Phils previously and never started more than 9 in any season. In 1932, he was 13-11 with a 2.80 ERA. He pitched three more seasons with Boston, mostly as a starter.

So what Carrasco's doing is not unprecedented, but it's pretty darn rare. His success as a starter might be a fluke. It may end with another assignment to the pen. But he might just stuck around for a while as a starting pitcher.