Why are the Red Sox starting Joe Kelly? Blame Ian Kennedy

There is a very good chance you are not excited about Joe Kelly getting another chance to start for the Red Sox. He had a rough 2015, and while his second half was a bit better, it wasn't so good that it's easy to put much stock into it as his new normal. Kelly is going to be Boston's fifth starter, though, and really, he does deserve another shot at sticking in the rotation. Some of that has to do with what he's done in his career and could still do, but much of it also has to do with the alternatives. Especially those on the free agent market.

Mike Leake is a generally dependable, league-average starter who has spent his entire career in the National League not striking hitters out, and the Cardinals gave him five years and $80 million. They can get away with the associated risks because of their league and their pitcher-friendly park. The same goes for the Giants, who handed Jeff Samardzija $90 million and five years so he can try to make it four good seasons in nine tries in 2016. It'll probably work for them, for the same reasons Leake should work for the Giants. Neither was a good bet at that money for the Red Sox and their environment, though.

The most egregious example of questionable starters getting significant money, though, is Ian Kennedy. Kennedy last produced an average-or-better ERA+ in 2012, and that was just at 101. Since 2013, he's pitched 550 innings -- an average of 184 per year -- but managed an ERA+ of 84. Despite pitching his home games in Petco Park, a notoriously pitcher-friendly space, Kennedy allowed 31 homers in just 168 innings in 2015 -- that's 1.7 homers per nine innings, despite every environmental advantage a pitcher could ask for.

Ian Kennedy signed for five years and $70 million guaranteed.

Like with Leake and Shark, there is a chance this works out because of where Kennedy signed. The Royals play in Kauffman Stadium, which is notoriously pitcher-friendly, especially for the long ball. They also seem as if they have better coaching than the Padres have had on the pitching side, so maybe they see something in Kennedy they can fix to get him back to the good old days. Maybe not, and it'll be a disaster! That "maybe" would have been a "definitely" were Kennedy to sign that kind of deal with Boston, though.

Jake Roth-USA TODAY Sports

So, how does this all relate back to Kelly? Well, for one, Kelly had at least as good of a 2015 as Kennedy: his 89 ERA+ bested Kennedy's 85 mark, and he only threw 30 fewer innings by virtue of getting five fewer starts on the season. Neither was great, not by any means, but Kelly will earn $2.6 million in his first year of arbitration, while Kennedy has $70 million coming his way even if he never improves one bit on his most recent campaign. Kelly might never improve, either, but he comes with far less of a commitment on the chance he will than Kennedy or Samardzija do.

This is the market now. It's not a bad thing that pitchers like these are getting paid -- it's better the money goes to the players than the owners, who already receive much more of a share of revenue than the players as is. It is important to recognize that this is how mid-level (or worse, in Kennedy's case) pitchers are going to be paid these days, though, and that while we wish the Red Sox would spend every dollar of revenue they pull in, it's just not going to happen. They have a budget, as does every other team -- yes, even the Dodgers have one now -- and signing David Price for $217 million filled most of it up for the next few years. And that's okay!

It's especially okay because the Sox can turn to the inexpensive Kelly to fill that fifth slot, either for the next few years as he finally cashes in on his potential to be a big-league starter, or until June when the Sox realize it's never going to work and that it's time for Henry Owens to get a shot. There is a risk here that none of Kelly, Owens, Brian Johnson, Steven Wright, or Roenis Elias should be Boston's fifth starter, a risk that they could help drag down a pitching staff that might be leaned upon heavily given the potential issues in the lineup. The revamped bullpen helps lessen those issues, at least, so if Kelly can only provide five solid innings each time out, they can work with it, unlike last year, when it was often the prelude to an eventual loss.

Photo by Denis Poroy/Getty Images

Throwing money at Leake, Shark, or Kennedy wouldn't have changed those risks or the potential need to lean on the pen, though. And we can probably say the same about Yovani Gallardo whenever he finally inks his eventual deal -- signing any of them would have just made the fifth spot a more expensive risk, and that's difficult to endure at a time when no one is quite sure what Hanley Ramirez, Pablo Sandoval, and Rick Porcello are going to produce for the combined $61 million they're making in 2016 and beyond.

It's worth pointing out that former Sox general manager Ben Cherington saw this change in the free agent market coming when Porcello signed a four-year, $82.5 million extension before the 2015 season. The first year didn't pay off, obviously, but there were extenuating circumstances -- such as a now-fired pitching coach deciding Porcello didn't need to rely on the sinker that earned him his extension in the first place -- and there is still plenty of time for the deal to look as prescient as it was intended to. It's also part of why the Sox went with Kelly again in 2015 when given the opportunity, as neither free agency nor the farm were guaranteed problem solvers for the fifth spot in the future.

So, the Sox will give Joe Kelly one last shot to make this work and solve that very problem for them. If he can't hold down the fifth spot in the rotation, then Owens, Wright, Elias, or Johnson will get their opportunities to do what Kelly could not. Maybe none of them turn out to be the answer, but they also aren't going to clog up Boston's budget, or make it more difficult to find in-season upgrades that will cost money while sparing the farm. It's not the perfect solution, but it is a solution, and given the alternatives, it's at least a preferable one for now.