A humble suggestion for the Orioles’ rotation

onlWell, we’re a month into the new year, so I think it’s safe to say that all resolutions have been broken, probably for a few weeks now. Forming a new habit generally involves breaking an old one, which can be agonizing, and all too often we decide that the sacrifice is not worth the benefit.

In a scientific survey I conducted of one person, the most prevalent of all resolutions is getting your pre-parenthood body back. Now, there are two tried and true ways to lose weight: incorporating a better diet or exercising more often. Some people have combined the two, but I’m fairly certain those people ended up either in a Mr. Olympia competition or possessed by cravings, scrounging every crevasse of their home and car for enough change for a hit of Krispy Kreme.

However, getting in shape would be a lot easier if the only adjustment you had to make to your lifestyle was to go to the gym five percent more often. That’s essentially like if you’re going a little less than two times per week, and then you go five more times all year.

All of this is to say: Five percent is the figure I’d like to present to the Baltimore Orioles’ new pitching coach, Roger McDowell. It’s a goal — a resolution, if you will — to urge his starters to increase their first-pitch strike rate by, as it would bump them up to a 60 percent first-pitch strike rate. Last year, the Orioles starting rotation threw first-pitch strikes 57.1 percent of the time. That was the nadir among all starting rotations; however, it was not a one-year blip.

Since 2012, Baltimore’s starters have thrown the fewest first-pitch strikes of any rotation in baseball. The MLB average over that span has been 60.8 percent, making the O’s 58.6 percent clip almost two standard deviations below the mean (1.97). And it’s not that the Orioles rotation can’t throw strikes; although they’re only 23rd in overall strike rate, I’ve highlighted their plot point to show what an outlier they are in terms of how overall strike rate and first-pitch strike rate correlate.

In fact, they’re 15th overall, and actually above average at throwing strikes, if we remove all first pitches.

But in regards to first-pitch situations, they’re even slightly worse than the Rockies, who play in a ballpark that greatly reduces the effectiveness of breaking balls, and where hits fly over the fences like kernels of corn popping out of a lidless pot.

Colorado appearing near the bottom of the list seemed like a beckoning of logic, and I thought that, since Oriole Park at Camden Yards boasted the sixth-highest home run factor from 2012-16, there might be some correlation between home park home factor and first-pitch strike rate. If you squint, it’s there, but it’s not strong enough to explain the Orioles’ first-strike woes.

Another factor that felt intuitively relevant to test was the correlation between pitch framing and first-pitch strike rate, since first pitches are much more likely to be called strikes. Baltimore’s catchers have graded poorly in this faculty, ranking 19th in the majors from 2012 through 2016, but I also found that the Brewers’ catchers — who have been the best pitch framers — haven’t had a significant impact on their pitching staff, who have thrown the 23rd-smallest clip of first pitch strikes.

I then moved to pitch types, which didn’t lead to any solid conclusions, though they could ultimately be pieces of the puzzle. In 2016, the O’s rotation threw the fifth-fewest first-pitch fastballs, but this regression showed a meager 0.034 correlation.

Even the fact that they threw the second-most first-pitch sliders didn’t lead anywhere. The Yankees threw the most sliders in 0-0 counts last year, but their overall first-pitch strike rate was third-best.

And in the case of a first-pitch slider, it’s easy to justify the Orioles’ thinking, since that can actually be an effective pitch to open a plate appearance with. For starters, they produce a lower exit velocity than fastballs, 87.6 mph compared to 89.8. They also induce nearly as high a swing rate while garnering more swings at pitches outside the zone.

The difference in approach between the Yankees and the Orioles in regards to first pitch sliders was the Yankees threw theirs in the zone 45.8 percent of the time, whereas the Orioles were in the zone at only a 39.4 clip.

This is where we get to the really interesting stuff. In fact, the Orioles rotation’s zone rate on all their pitches to start a plate appearance was just 40.2 percent. This also is the bottom of the barrel among starting rotations, and it’s not lost among opposing hitters: They only swung at 24.6 of the Orioles’ first-pitch offerings, which is more than a percentage point lower than the second lowest mark (Houston at 25.7 percent) and 2.97 standard deviations below the mean (average is 27.8 percent). Considering the league’s proclivity to be more patient in 0-0 counts is surpassed only by 3-0 counts, I think it really hurts the O’s that they are so cautious in first-pitch scenarios.

A count-by-count breakdown of how MLB exerted itself in 2016.

You’ll notice I’ve included zone 5 swing percent, and for a crucial reason. Some of you might already know where I’m going with this, but we can all agree that avoiding middle-middle is not a controversial idea.

Now, unfortunately I cannot separate starting pitchers and relievers at Baseball Savant, but I can set the search parameters to include just the first six innings. Last year, starters averaged 5.65 innings per start, so this isn’t perfect data, but it gets the job done. What we find is that the Orioles, most likely because they had to work from behind in the count so often, possessed the only starting rotation in all of baseball last year to have a higher zone 5 rate in all non 0-0 counts compared to their 0-0 pitches.

The obvious reason this isn’t good:

Maybe Baltimore’s starting pitchers’ command is so poor that they really do struggle with throwing first-pitch strikes as much as they struggle with avoiding middle-middle. I suppose this represents two sides of the same coin. However, given the count-based swing rates I’ve shown above, it feels to me like the Orioles are doing this backward; they should have LESS fear of zone 5 to begin a plate appearance, and then nibble in counts afterward.

Now, one thing we know about the Orioles is that they have an excellent bullpen. Over the time period I’ve outlined, they rank second in both fWAR and RA9-WAR, as well as first in WPA. So it stands to reason that the Orioles haven’t been stressing efficiency from their starting pitchers. Evidence of this can be seen in the pitchers that have risen from the Baltimore farm system.

Only Dylan Bundy checks in above the big-league average, and he’s faced a mere 480 batters in his young career. Miguel Gonzalez is absent from this list because he came over from the Angels organization, as is Wei-Yin Chen who came over from Japan; it could be purely coincidental that both of these players have career first-pitch strike rates that are above average, but perhaps not.

Another table I want to share with you shows how pitchers that have started their careers with the Orioles, but have since moved on, fared with regards to first-pitch strike rate.

*Only pitcher in 2016 to have a top-20 fWAR and a below-average first-pitch strike rate.

**Has faced only 18 batters as a non-Oriole.

So Baltimore doesn’t develop zone-pounding starters, but it doesn’t target them in trades or in free agency, either. Here is a list of acquisitions that have started games for the O’s during the 2012-16 seasons.

*Data taken from only three years prior to joining Orioles.

This data shows that these are flawed pitchers — in this regard, at least — and because of this, the Orioles have never had to pay an exorbitant amount, either monetarily or with prospects, to acquire their services. So while that could be seen as exploiting a market inefficiency, the performances of the pitchers they’ve developed in their own farm system points to this being an off-kilter philosophy of sorts.

There have been murmurs throughout baseball from former Orioles that Baltimore’s one-size-fits-all approach to developing pitchers is not appreciated. Nor has it been effective: Dating back to 2012, Baltimore’s starting rotation ranks 24th in fWAR and 18th in RA9-WAR, and even that’s been propped up by a defense that’s ranked as MLB’s eighth-best during that time.

This is where things could get sour in 2017. Yes, the Orioles have Manny Machado and Jonathan Schoop, but the rest of their core is aging. In fact, the only other position player on the “right” side of 30 is Hyun-Soo Kim, and he has been viewed as a platoon-only, bat-first player by the organization. The reality is that the defense is not going to be what it once was — indeed, it dipped to 19th in the majors last year. Despite the fact that the offense will likely hit another quarter of a thousand home runs in 2017, the team could stand to improve a facet of their game to make up for their declining defense and to mitigate the detriment of possible changes in personnel.

So now we come back to my proposition for the Orioles starters to become more aggressive in 0-0 counts. The benefit isn’t astronomical — it might only equate to one or two wins over the course of a full season — but a win is a win. In fact, using the run values provided in this piece by Dan Meyer, and a little math, we can surmise (very roughly) how much a five percent increase in first-pitch strike rate would affect the Orioles rotation’s projected ERA for 2017.

Despite its strong year-to-year correlation, Steamer doesn’t project first-pitch strike rate, but we can use each player’s career number to figure that the O’s rotation projects to a 58.5 first-pitch strike rate. A 5 percent increase would bring that to 61.4, which would net them 118 more plate appearances that start with a strike. With the full benefit of a first pitch strike being -0.069 runs, this would knock ⁸⁄₁₀₀ off their ERA, taking it from a projected 4.53 to 4.45. I know that doesn’t seem like much, but if this tactic allows the starters to stay out of middle-middle over the remainder of a plate appearance, it could end up being more.

Baltimore’s pitching coaching staff has been gutted. Rick Peterson, Dave Wallace, and Dom Chiti are gone, and while new pitching coach Roger McDowell has a history with the Orioles, his work as Atlanta’s pitching coach over the last eleven years could serve as an indication that he’d be willing to take his staff in this new direction. Under his guise, the Braves rotation was efficient in first pitch situations, logging a 60.6 first-pitch strike rate rate. Of course, only so much of a team’s outcome can be attributed to the coaches, and McDowell had some quality arms to work with in Atlanta.

But the Orioles’ rotation finally features some real, top-of-the-rotation talent. Kevin Gausman is coming off a quietly strong season by whatever valuation method you prefer (4.0 RA9-WAR; 3.0 fWAR; 4.5 WARP), and Dylan Bundy finally seems to be healthy. Steamer projects those two to be worth a combined 4.7 WAR, but I’ll take the over all day, and especially if they’re throwing more 0-0 strikes. The landscape is shifting in Baltimore, and if there was ever a time to make this change, it’s now.

******

Mark Davidson is a contributing writer at Beyond the Box Score. You can follow him and send him bat flip gifs at @NtflxnRichHill