Balki-B wrote:I understand the failed-starter route now, and agree with the strategy for team building as it relates to sustainability. Just not sure how the bullpen can improve without giving up Groshans or Orelvis when Boston will also be shopping for pen help. Is Kikuchi or Pearson the answer? Is a trade on the margins to acquire a pitcher who provides Cimber-like results enough?
My bad, initially misread which part you were saying didn't age well.
My feeling is the latter: identify a pitcher who gets strong results and/or has strong peripherals without a super-long track record of success, who doesn't look/feel like a high-leverage reliever (as with Cimber), or who can benefit from a change in pitch mix (something we've had success with). It's how the Rays perpetually seem to have five great relievers despite never actually spending on them (and seemingly trading several of them a year): they're great at identifying the traits that make for good relievers, and comfortable with taking the risk of guys whose numbers aren't very good but have underlying stuff/statistical components that they feel can be tuned for optimal performance.
Take Jason Adam. He was waiver wire fodder for a few years, a nothing guy for the Royals (and for us) who showed a bit of utility as a high-K pitcher with the Cubs, but still walked a tonne of guys and got hit fairly hard, because he mostly just threw fastballs. The Rays picked him up and overhauled his pitch mix: no more curveballs, fewer fastballs, and a lot of sliders and changeups. About 65% sliders and changeups actually, which is extremely high, especially for someone who threw nearly 60% fastballs the year previous.
Looking back, it seems really obvious: his fastball got hit pretty hard, but his outcomes on the small number of SL/CHs that he threw were quite good. The Rays gambled that those results would hold in large samples, and they were right.
We don't necessarily need to get that scrap heap-y, but you can find players of a similar ilk. Take Joe Jimenez. He sucked for years for the Tigers, and largely hung around because he was a power arm, and they were the Tigers. He's getting good results this year, both his raw numbers and his percentile rankings for expected outcomes, hard contact, etc, but relievers in small samples are all over the place, so the question becomes: is it sustainable? Are there things he's doing now that he wasn't doing in the past that explain his success, or is he just getting lucky?
And the answer is...maybe! One thing jumps out looking at his Baseball Savant page: the clustering on his slider is really good this season. He seems to be able to consistently locate it in the lower-right quadrant where RHP sliders belong. His slider gets good results generally over the past 4 seasons, and you might be able to eke out a bit of extra value by getting him to lean on it even harder. And then you can look at some of his other past numbers: last season, he had a completely out-of-character high BB rate, but otherwise his numbers were quite good. He's better about limiting hard contact than he was in earlier seasons, and it hasn't come at a cost to his K rate. If he doesn't forget how to throw strikes again, you could have a pretty cheap reliever (given his past struggles) with a solid chance to provide useful middle-inning relief, and a chance to be a high-K late-innings guy.
It's not an exact science (unless you're the Rays) because relievers are black magic, though. This time last year I was stumping for Paul Fry, who had sleeper relief ace components over a year and change with the Orioles. He then completely imploded and is currently languishing in the D-Backs AAA bullpen.
Certainly would have been a Dickian level blunder in the pantheon of bad Jay trades. 
