The struggles of Rubby De La Rosa against left handed batters have been a point of conversation for the better part of six months, at least. Just how bad he was, last year, has already been posted. And yes, his career numbers against LHB look terrible. But I do think there are reasons to suspect that his struggles last year had to do with strategy as much as performance, and that working with a new pitching coach might resolve these problems and take away the giant splits we saw last year. (NB: I do not agree with the decision to reward him with a spot right at the outset. I would have much rather seen a competition for two spots. I'm just writing about why we can look on the bright side of a bad decision, something I've wound up doing all winter and now into the spring, it seems.)
We need to remember that RDLR hasn't had much major league experience. Last year was his only full season in the big leagues, and accounted for more than half of his career innings. Last year, his performance against right handed bats improved; they posted an OPS of .611 against him, 77 points lower than his career number against right handed bats. On the other hand, left handed bats posted that .949 OPS, 74 points higher than his career numbers. In other words, last year he was substantially better against RHB, and substantially worse against LHB, than he had ever been in his career. To look at one stat only, in 174.1 career innings prior to last year, he gave up 12 home runs to RHB and 8 to LHB. Last year (in 188.2 innings) he gave up 12 to RHB, and 20 to LHB. Clearly, something made him much worse last year. In 2014 the difference in OPS between RHB and LHB was only 40 points, which is reasonable. In 2011 (the only other year where he saw a decent amount of playing time) the difference was 81 points; high, but nothing like last year's 338 points.
There's also the little matter of how he performed in the minor leagues. In 2013, in AAA, he was 41 points better against LHB than RHB. In fact, in each of his full seasons in the minors, he was either roughly identical or better against LHB than he was against RHB. So if this is a legitimate problem, it's one that only showed itself when he advanced to the major leagues, and, in fact, only when he came to the Diamondbacks. That is one possibility, of course. But there are other possibilities which I believe provide a better explanation.
First, there's the matter of Mike Harkey, the since-fired pitching coach. He hadn't been out the door long before complaints about his lack of tailoring the game plan to the pitchers surfaced. While this might have been the organization (once again) talking bad about someone on his way out the door, it could also explain why he was so good against RHB and so bad against LHB; something in the strategy wasn't working, and Harkey didn't help him fix it. If that is true, what might have been the strategic problems, and how could they be fixed?
RDLR saw dramatic increases of usage on two pitches in 2015: his slider and his sinker. He rarely threw the sinker prior to joining the D-backs; May of 2015 was the first time the sinker accounted for more than 10% of his pitches. As the year went on, he threw it as much as 28% of the time. While he had thrown his slider throughout his career, his usage of the slider peaked at 33% in August of 2015. Those are the overall numbers. Against right handed batters, he used primarily the fastball and the slider, with either the sinker or the changeup as a third pitch. That is somewhat significant, as his slider or his changeup are his swing-and-miss pitches, meaning that if he is being aggressive, he'll throw those pitches. Now early in the year, players were also swinging and missing at his sinker, probably because they weren't expecting it; he hadn't hardly thrown it before. Later in the year, they weren't missing it, but you don't really want batters to miss the sinker, anyway.
As far as those two "out" pitches are concerned, left handed batters missed the changeup about 35% of the time they swung at it, while they rarely missed the slider. Yet, as the year wore on, RDLR threw the slider more and more and the changeup less, to LHB. He was also throwing far more sinkers to left handed batters than to right handed batters. Basically, from the pitches he was throwing, it looks like he approached right handed batters looking to strike them out and be aggressive, and left handed batters looking to get a ground ball. Aside from any mental issues that might be conjectured with this, throwing fewer four seamers and more sinkers probably decreased the effectiveness of the changeup when he did throw it. Rubby, in previous years, was basically a fourseam-changeup pitcher to left handed hitters, and it worked. In 2015, he threw the sinker in almost 1/4 of his pitches to left handed hitters.
In fairness, when Rubby's changeup was hit by LHB, they have always hit it hard, to the tune of a slugging percentage in excess of .500 for his career. That explains going away from it, somewhat. But LHB slugged .684 against his sinker in 2015, so going for more groundballs wasn't helping him. (Right handed hitters, on the other hand, hit only .122 and slugged .220 against the sinker, yet he still wasn't throwing it nearly as frequently to them; only 12% of the time vs. 23% against LHB.)
There are two good things that can be taken out of this. First, that the problem pitch was something that was new to him last year. Second, that there is a problem pitch. If Butcher is less sinker oriented than Harkey, possibly the sinker can become a useful fourth or fifth pitch against LHB, instead of as one of his main pitches. Using his slider more against LHB might prove to be part of the answer; despite at .311 batting average against his slider, he had the lowest slugging percentage against his slider of any of his pitches. Simply throwing his slider something like 20% of the time (instead of 13%) and his sinker only about 10% of the time could make a big difference, unless batters learn to hit his slider with power like they did his sinker.
All that said, I think the most important thing management can do is to be flexible, and if RDLR can't figure it out, put someone in there that can. Depth, which has been a concern in the past, will be no longer. So while he's the fourth starter now, they need to make sure that if he doesn't show the capacity, they bring someone else into the rotation soon.