Introducing the 2019 Astros: The Rotation

The greatest strength of the 103-win, 2018 Houston Astros was the starting rotation. It led both leagues in ERA at 3.16, and most other statistical categories. It accounted for 21.7 Wins Above Replacement according to Fangraphs, which surprisingly to me, was second behind the Indians...barely. The Indians’ ERA- was also slightly better than the Astros. Still, it’s hard to beat a rotation of five All-Stars, three of whom appeared in last year’s summer classic.

Unfortunately, three of those five stellar performers will not reappear in an Astros uniform in 2019. Charlie Morton signed a free agent contract with the Tampa Bay Rays. Dallas Keuchel is still an unsigned free agent, and Lance McCullers underwent Tommy John surgery.

That leaves 8.5 unaccounted wins for the 2019 Astros, but on the good side, the best pitching combo in the American League remains the anchor of the staff. Returning for 2019 is Justin Verlander, runner-up in Cy Young voting in 2019, and Gerrit Cole, not far behind Verlander in the balloting. They were the best 1-2 combo in the American league, accounting for 12.7 fWAR together, more than any other pairing.

Can the Astros’ rotation of 2019 equal the 2018 production?

Who will fill the shoes of the departed All-Stars?

Let’s introduce the starting rotation of the 2019 Houston Astros.

The Rotation

Justin Verlander:

Career Stats: ERA 3.39, K% 23.9, BB% 7.1, FIP 3.42 WHIP 1.16, fWAR 65.6 (13 seasons)

2018 Stats: ERA 2.52, K% 34.8, BB% 4.4%, FIP 2.78, WHIP 0.90, fWAR 6.7

Steamer Projections: ERA 3.37, K% 30.4, BB% 6.1, FIP 3.28, WHIP 1.07 fWAR 5.1

All Astros fans know the story of how Verlander came to the Astros at the last second of the 2017 waiver trade deadline, and then became the play-off force that won the Astros their first World Series.

Last year was one of the best of his career at age 35. He had his second lowest ERA, lowest FIP and xFIP, most strikeouts, highest K%, lowest BB%, lowest WHIP etc. etc.

No wonder that projections anticipate some regression, almost a run more per nine innings, and about one and a half less wins. To paraphrase the late George Bush, I don’t think it would be prudent to dispute the idea that Verlander is due some regression, but if Steamer is correct, I’ll take that regression any time. Pencil Verlander in for another All-Star appearance, his 8th, (barring injury, the universal disclaimer)

And, as every Astros fan knows by now, Verlander is signed with the Astros through 2021 for $33 million a year.

Gerrit Cole:

Career Stats: ERA 3.37, K% 25.1, BB% 6.6, FIP 3.15, WHIP 1.18, fWAR 21.2 (6 seasons)

2018 Stats: ERA 2.88, K% 34.5, BB% 8.0, FIP 2.70, WHIP 1.03, fWAR 6.0

Steamer: ERA 3.53, K% 28.8, BB% 7.5, FIP 3.41, WHIP 1.15, fWAR 4.3

Much like Verlander, and the departed Charlie Morton, Cole had his best season last year. Do we see a pattern here? Is there a common denominator for Cole’s and many other Astros’ pitchers career renaissances? (hint, Brent Strom)

So in this case I dispute the Streamer projection, which was the most pessimistic of all the services on Fangraphs. Regression? yes. To a point worse than career averages? (and Cole had some rather mediocre seasons with the Pirates) I won’t go that far. At age 28? In a contract year? In line with the others projections let’s give Cole almost 5 wins and a 3.30 ERA.

Collin McHugh:

Career Stats: ERA 3.87, K% 22.6, BB% 6.6, FIP 3.66, WHIP 1.25, fWAR 11.5 ( 5 seasons)

2018 Stats (relief): ERA 1.99, K% 33.2, BB% 7.4, FIP 2.72, WHIP 0.91, fWAR 1.4

Steamer: ERA 4.35, K% 21.6, BB% 7.7, FIP 4.34, WHIP 1.31, fWAR 1.5

McHugh is the hardest pitcher to project because with him there are so many moving parts. He was a near All Star quality starter in 2014-2015, declined in 2016, was injured most of 2017, and became one of the league’s most effective relievers in 2018. At the same time he incorporated a slider, which Fangraphs says was the 3rd most effective slider in MLB for pitchers with more than 70 innings.

Was his success in 2018 due to being a reliever? If so, then can you project that success into a role as a starter? Maybe not. Or was his success due to the new pitch? If so, then maybe his success last year does project into the rotation. On the other hand, as his slider has become more dominant, his once elite curve has become more pedestrian, per Fangraphs, but still good.

To complicate the matter, McHugh has had a terrible Spring Training. He started slow with a nagging injury, and has gotten bombed every time he has pitched since then. Yes, he is historically a poor Spring Training performer, but nothing like this: 15.88 ERA, 2.47 WHIP in only 5.2 innings. His fastball was not getting close to 90 as batters teed off. He missed a Spring Training start due to rain but threw 70 practice pitches last Wednesday. There’s just a bit of question whether he is fully healthy.

If his slow start is just Spring Training background noise then I think Steamer again has underestimated another Brent Strom protege. This projection would equal McHugh’s worst season statistically. And even then he had 2.8 WAR. I think I prefer this evaluation in Fangraph’s profile:

“The Quick Opinion: Not only did Collin McHugh excel in relief, he added a slider to his repertoire. McHugh won 43 games over three seasons as a starter with Houston before transitioning to the bullpen. He will try to carry over the gains of his slider to starting adding it to his fastball, curve and cutter to keep hitters off balance. By using more curves and sliders as a starter, McHugh could be a sneaky late round flier this season.”

I said in an earlier article that McHugh is basically as good as the departed Charlie Morton. If he finds his groove I will stick with that. If. If not, the Astros have plenty of reinforcements who could be as good or better. (which will be covered in another article)

Wade Miley

Career Stats: ERA 4.26, K% 18.4, BB% 8.1, FIP 4.06, WHIP 1.39, fWAR 13.1 (7 seasons)

2018 Stats: ERA 2.57, K% 14.8, BB% 8.0, FIP 3.59, WHIP 1.21, fWAR 1.4 (80 innings)

Steamer: ERA 4.59 K% 17.4, BB% 8.7, FIP 4.63, WHIP 1.46, fWAR 0.8

Why does Steamer think that Miley will get almost half as many wins in 2019 as he did in 2018 even though they have him pitching 50% more innings? Look at some of the deeper peripherals in 2018. Above, we already see his FIP was a full run higher than his ERA, and his xFIP, 4.30, was almost two runs higher, right in line with his career ERA. His SIERA, Fangraph’s most advanced predictive metric, was even worse; 4.66, about what Steamer projects for Miley’s 2019 ERA. Despite all the talk that Miley was so good last year because of his new-found cutter, (it is highly rated in the pitch value scores) all this says is that in 80 innings he had some good luck last year. His BABIP was .269, well below his career average.

He did avoid hard contact better last year than in years past, but not remarkably so: 2018 barrel %; 3.5. Career: 5.5. His 2018 exit velocity; 85 MPH. Career; 87. Meanwhile his K% was in the lower 6% of MLB pitchers. His WOBA in 2018 was .283, well below the xWOBA of .303 that his batted ball profile would predict.

There’s only one reason to doubt the Steamer projection; the Strom effect. Ask Will Harris, Collin McHugh, Charlie Morton, Justin Verlander or Gerrit Cole if Brent Strom and the Astros’ Nerd Cave can help a pitcher. Then again, don’t ask Scott Kazmir, Pat Neshek, Mike Fiers, Ken Giles or Francisco Liriano that question if you want to assume that Miley will be better as an Astro.

I’ve called Brent Strom a wizard before, but I think it’s too easy to assume that every time the Astros get a mediocre pitcher, say Charlie Morton, that just by putting on an Astros uniform he will be transformed into an All Star. Yes it’s happened, but not every time. There’s a good chance that Miley’s peripherals improve as an Astro, and that he outperforms Steamer, but it is too optimistic to expect another Charlie Morton miracle.

I’ll give him 1.1 WAR. That is the most optimistic forecast. But maybe Strom has another miracle in store. I just can’t bank on it.

Brad Peacock

Career Stats: ERA 3.97, K%, 24.8, BB% 10.5, FIP 4.19, WHIP 1.32, fWAR 3.8 (5 seasons)

2018 Stats: ERA 3.46, K% 35.3, BB% 7.4% FIP 3.47, WHIP 1.17, fWAR 0.6

Steamer: ERA 4.23, K% 25.0, BB% 9.0, FIP 4.16, WHIP 1.31, fWAR 1.0

The Astros are hoping for a rebirth of the 2017 Brad Peacock who, pitching mostly as a starter, had a 3.00 ERA and 3.2 WAR. Peripherals were a little behind; xFIP 3.73, SIERA 3.76. His ERA as a starter was slightly worse, 3.22, while as a reliever in 2017 his ERA was 1.77, but with an xFIP of 4.12. Good home run luck for Pea in a relief role in 2017.

In 2018, pitching almost entirely in relief, Peacock had a 3.47 ERA, but his home run luck turned. His xFIP was 2.82 and SIERA 2.40, lowest of his career.

So as might be expected, Peacock, after outperforming peripherals in 2017, under-performed them in 2018. He was neither as good as he looked in 2017, nor as bad as he looked in 2018. On the other hand, the pitch values on both his fastball and slider were significantly worse in 2018 than in 2017.

Still, why is Steamer projecting ERA, FIP and xFIP a half run to almost a full run higher than what Peacock has done in either of the past two years, including when he was a starter? It is true that if he starts, he must be protected from pitching the third time through the order, where his career ERA is 8.48. But I’m sure A.J Hinch knows that, and won’t expose him promiscuously.

I’m going with the more optimistic Depth Charts or ZiPS projections, which credit Peacock with about 1.5 wins. These are based on statistics more in line with the average of his last two seasons.

Summary

Try as I might, I can’t say that the 2019 Opening Day rotation will be as awesome as last year’s. Using my conclusions above, Verlander’s WAR will be 1.6 less, (6.7 to 5.1) Cole’s 1 less. 6 to 5) Using the most optimistic projections for Miley, he will be 2.2 less than his nearest comp, Dallas Keuchel. (3.3 to 1.1) I said McHugh would equal Morton, but with the caveat that he gets healthy/finds the groove he hasn’t found this Spring. (2.9) If Peacock replaces McCullers he is projected to produce 1 less win. (2.5 to 1.5) In total that is 5.8 less wins from the rotation.

If that were subtracted from last year, the Astros would have ranked third in rotation WAR in the AL instead of second, just behind the Yankees.

But we just might want to see 3 through 5 falter in the starting rotation, because then we might get to see the future right now. Guys like Forrest Whitley, #1 pitching prospect in the solar system, Josh James, J.B. Bukauskas, who showed us that third pitch in Spring Training that was supposedly the missing link in his development. Armenteros, Bielak, Martin, Perez. By season’s end, I dare say, this starting five will be long gone.

Astrosfuture will fill us in later in the week.