Shelby Miller is back, maybe

Shelby Miller did not have great results in his first two starts of the season. He struck out just seven hitters against six walks and gave up four home runs in Cardinals losses to Pittsburgh and Cincinnati. Given his infamous disappearing act in the playoffs last season and his struggles toward the end of the season that precipitated his exclusion from the World Series run, questions have been raised about what kind of pitcher Miller is right now. A positive start against Milwaukee on Tuesday does less to answer questions than it does to raise more questions. Is this a new Shelby? Back to old Shelby? Or was Milwaukee an aberration in his continued struggles?

Prior to the start in Milwaukee, Dave Cameron at Fangraphs thought Miller looked broken.

Right now, Shelby Miller is not a good pitcher. He hasn't been for a while, and the Cardinals should absolutely be concerned. Perhaps his problems are fixable, and this is a thing will work itself out while he's on the mound; they should hope so, because an off-season of rest didn't seem to do any good.

The start against Milwaukee does not make Cameron wrong. Based on the end of last season and his first two starts of this season, Miller has not been a good pitcher. Cameron goes into his reasoning in the post, emphasizing the lack of a secondary pitch that generates swings and misses as well as declining effectiveness of the fastball that produced great results in the first half of 2013. Ben previously addressed Miller's swinging strike rate in the offseason, and Joe discussed just how effective Miller's fastball was in 2013.

Before getting to Miller's start against Milwaukee, examining Miller's results in 2013 should provide a solid backdrop to his performance thus far in 2014. According to Fangraphs, in the first half of last season, he was striking out 9.6 hitters per nine innings (26.4% K-rate) and walking just 2.49 hitters per nine innings (6.8% BB-rate). The walk rate was somewhat of a surprise given that in his last two years in the minors, Miller's BB/9 exceeded three. With a 3.07 FIP and a 3.34 xFIP the Cardinals received a first half comparable to Mat Latos or Madison Bumgarner. In the second half, Miller's numbers worsened to 7.47 K/9 (19.1%), 3.67 BB/9 (9.4%) with a FIP of 4.59 and an xFIP of 4.33.

Pinpointing when Miller's season started to unravel is difficult. In his post, Cameron posited an August 25th date as the falloff date. On August 7th, Miller faced just one batter after being hit on the forearm, and thereafter lost some velocity and changed his pitch selection. Bernie Miklasz noticed problems as early as June 12th. Bernie noticed a six-start trend of high ERAs and lower K/BB ratios and theorizing hitters were reading scouting reports and not chasing the fastball like in Miller's initial tour around the big leagues.

By focusing on pitch selection rather than choosing among a mix of good and bad starts, the turning point appears to be August 7th, when Miller left the game due to injury. Using August 25th is a fine date to use, but it is a results based date as Miller had just gone seven innings against the Braves, striking out six and issuing no walks. It was also the first game where he threw his cutter with great frequency providing more similarities to later in the season than his stellar first half.

From August 25, 2013 through the end of the season, Miller threw 65.6% fastballs, 13% curves, 12.8% cutters, and 8.6% changes. His fastball velocity was 93.8 (Unless otherwise specified, the rest of the data comes from BrooksBasebal.net). The 25th might not be the best cutoff because it matched his pitch selection during his previous three starts. After getting hit by a batted ball on August 7th, but prior to August 25th, Miller threw 65.8% fastballs (94.3 mph), 13.1% curves, 11.2% cutters, and 9.9% changeups. He pitched decently during this time, but began showing some signs of problems. He struck out 13, but walked six in the first two games before putting up the previously mentioned six-strikeout, no-walk game on August 24th.

Bernie did note some signs of struggle for Miller in late June, and the period included two bad starts. However, it also included four good starts where he struck out 26 batters and walked just five. Given that the period was followed by three more starts where Miller struck out 20 and walked six prior to August 7, those two bad starts look more like blips than a pattern.

In his most recent Bytes, Bernie used the same August 7th checkpoint and noted the difficulties expected when modifying pitches midseason. From the beginning of 2013 through August 6th, Miller threw his fastball 73.3% of the time, averaging 94.5 miles per hour, with his curve at 21%, a changeup at five percent and a cutter at roughly half a percent. Miller suffered growing pains with his new pitch, but the long season may have been wearing on him as he also lost a small amount of velocity, losing close to a mile per hour on his fastball.

The results towards the end of the season, while including some encouraging signs, were overall not very good. The changes were made, not necessarily because the prior results were innefective, but because Miller lacked efficiency. Throwing so many fastballs leads to a lot of foul balls and greater pitch counts. Adding a cutter was supposed to help him go deeper into games. His initial tries did not help. Through August 2, 2013, he averaged 5.8 innings per start, 98 pitches per start, and 67% strikes thrown. After August 7th, he averaged 5.8 innings per start, 100 pitches per start, and threw 64% of his pitches strikes.

Miller continued the same pattern for the first two games of 2014 that he ended 2013. In his first two games, he averaged 5 ⅔ innings pitched, made just 95 pitches per game, and threw just 62% of his pitches for strikes. In those starts, he threw 64.6% fastballs, 20.6% curves, 9.5% cutters, and 4.8% changeups. His average fastball velocity was 93.4. The velocity is not alarming this early in the season. In Miller's first start of 2013, his fastball averaged 93.1 miles per hour. In his second start in 2013, Miller's fastball averaged 93.9 miles per hour. By his third game, the fastball was up to 94.5 miles per hour.

There are still mixed messages to be taken from the start at Milwaukee. In six innings he was fairly efficient, making 94 pitches, striking out seven, giving up only three hits and one run on a solo shot to Aramis Ramirez. He also walked three and his command was iffy. Milwaukee is an aggressive team, making Miller's fastball more enticing to swing than it may be for other teams. All three of the walks issued by Miller came in the first two innings. Yadier Molina came out to the mound after a 3-0 count to Khris Davis. After Davis walked, Miller struck out six of the remaining sixteen hitters with just two batters reaching three balls and no free passes issued.

At their very brief mound visit, Molina told Shelby he was falling off the mound. Command will always be an issue. He has a very deceptive fastball and getting hitters to swing near the zone is imperative for his success. After the game, Matheny said,

"I don't know what it is about his fastball," said Matheny. "But when it's right, guys just have a hard time finding the barrel with it."

BrooksBaseball.net did not have the Pitch f/x information until the end of the third inning. Of the 53 pitches after that point, 44 were fastballs, averaging 94.8 miles per hour. The velocity may be inflated somewhat compared to his complete game average if he picked up velocity as the game wore on, but the velocity and the strikeouts are a sign that he is closer to the pitcher that performed extremely well in the first half of last season as opposed to the pitcher who struggled down the stretch. I re-watched the first three innings of the start and counted 35 fastballs among his first 41 pitches although without radar readings, it is possible I counted some cutters as fastballs.

Throwing 84% fastballs may not be sustainable going forward, but if he can consistently pound the zone with his fastball and sparingly use his offspeed pitches, we may see more outings like the one in Milwaukee. Pitching up in the zone makes him susceptible to home runs and the occasional poor outing, but the Milwaukee start, especially the last four innings, showed what he can do when his fastball is working.

Attempts to expand Miller's arsenal, increase his efficiency going forward, and make him less predictable seem like smart issues to tackle given the small number of pitchers who have survived on a single pitch great pitch. For Miller though, he has achieved considerably more success the more he throws his fastball and using other pitches a quarter of the time. Pitchers can do well with just two pitches. Tony Cingrani has survived on one pitch thus far and is working to expand his game.

An improving cutter could make Miller a much better pitcher, but going from zero to ten percent at the expense of his fastball has yet to produce the desired results. Cutters get hitters to chase and induce weak contact. Unfortunately that duplicates his fast fastball in a less effective manner and does not solve the problem of staying ahead of hitters. His curve, cutter, and changeup will never be big swing and miss pitches, but when his fastball is on, it might not matter.