Sunday, May 31, 2015

The Ghosts of Astros Past: Starting Pitching Edition.

I have written about the ins-and-outs of the Astros' starting rotation a few times recently, highlighting the difficulties they have around depth and possible future ineffectiveness.  I doubt many people think a Keuchel-McHugh-Feldman-Hernández rotation will match up well against the Tigers, Nationals or Dodgers, and A.J. Hinch recently added to the idea that the depth of starting pitching is being tested with injuries and ineffectiveness at the moment.

There isn't much going on with the starters in Fresno: only Tommy Shirley (and now Dan Straily) have ERA's below 4 out of pitchers that have started four or more games in AAA this year, and Jake Buchanan is already in the majors, riding the pine.  The real strength of the Astros' system in terms of starting pitching is at AA and below, with Vincent Velasquez, Mark Appel, Josh Hader, Chris Devenski, Mike Hauschild and Aaron West all possessing solid stat-lines in Corpus, and a handful of lesser-heralded prospects doing well in the graveyard that is Lancaster.  Daniel Mengden and his major-league moustache was the notable guy in Quad Cities, but now resides at Lancaster.

With McCullers and Feliz both currently in the Majors, I think that the Astros are looking at what they currently have to determine whether they need to make a trade or not.  McHugh and Kuechel are no sure thing (although Keuchel certainly looked like it tonight) and Feldman and Hernández could turn into pumpkins at any time, so the Astros may need anything between zero and four starting pitchers before the summer is finished if they are to push toward the playoffs.

The lack of depth is highlighted by the fact that the Astros have traded four young starting pitchers over the last 18 months.  Jordan Lyles (and Brandon Barnes) were traded to Colorado after the 2013 season, bringing back Dexter Fowler.  Jarred Cosart (with Kiké Hernandez and Austin Wates) went to the Marlins in late 2014, bringing back a 2015 draft pick, Handsome Jake Marisnick, Francis Martes and the recently-healed Colin Moran.  Nick Tropeano (and catcher Carlos Perez) were packaged up for Hank Conger, and Mike Foltynewicz was sent with Rio Ruiz and Andrew Thurman to bring back Evan Gattis.  The exit of those players was no coincidence - the Astros were clearly prepared to trade their upper-level pitching prospects and unestablished major league starters for established players with upside.  I think that the Astros had a plan to clear the way for their fabulous depth lower in the minors.  Allowing someone like Folty to get established in the Major Leagues may work to damage their stock if they were to prove ineffective.  They were deliberately trading potential while doubling down on the pitchers they had further down in the minor leagues.

The question that this article poses is relatively simple and purely hypothetical: would the Astros want any of these guys back to fill a rotation hole for a late-season drive toward the postseason??  Let's look at the four traded starters in order of trades, and see how they have progressed.

Jordan Lyles:
Lyles is the most senior of the four pitchers in terms of experience and service time.  He was drafted by the Astros late in the first round of the 2008 Draft out of Hartsville HS, South Carolina.  He first appeared in the Major Leagues in 2011 with the Astros, starting 15 games and relieving in five more.  In 2012 and 2013, he started 25 games each year, and had 22 games with the Rockies in 2014.  He has started 9 games so far this year.

Lyles' FIP is trending in the right direction: 4.53; 4.53; 4.57; 4.22; 3.87 (from 2011 to 2015).  His recent improvement is mostly on the basis of a halving in HR/9 between 2014 (0.9) and 2015 (0.4).  He has been striking out 5.3 per nine so far in 2015 (the lowest of his career) and walking 3.6 per nine (the highest of his career).  His WHIP currently sits at 1.490 - the second highest of his career.

Acknowledging that Colorado is an awful place to pitch, it is very possible that Lyles has not improved at all from when he was an Astro.  He has never managed more than 142 innings pitched in a single season, and has never posted an ERA better than league average.  Lyles would not be the answer to the Astros' possible rotation issues.  Additionally, the Astros would not be keen to undo the Fowler trade, especially since they later dealt Fowler from a position of depth to get interesting pieces in Luis Valbuena and Dan Straily.

Jarred Cosart:
Cosart is an interesting guy, and not just as a pitcher.  He has been involved in his share of controversy.  Cosart has one nasty pitch - a fastball with a natural cutting action - and he pumps gas in the mid-90's.  If he learns to spot that fastball, and develop a couple of other complimentary pitches, then he could become a dominant starting pitcher, capable of carrying a club.

In 2013, Cosart started 10 games for the Astros.  Prior to being traded in 2014, he started 20 games.  Over those 30 games for the Astros, Cosart pitched to a 3.57 ERA, 4.13 FIP, with a 5.5 K/9 and a 4.4 BB/9.  Control was always the issue with Cosart - in 2013, he famously walked more (35) than he struck out (33).

Cosart finished 2014 very strongly with the Marlins.  He lowered his ERA and FIP significantly (2.39 and 3.32 respectively), striking out 5.6 per nine, while walking 3.1 per nine.  2015 has not been as kind: a 4.08 ERA and 4.67 FIP, with very similar strikeout (5.4/9) and walk (3.2/9) rates to 2014.  The difference in both ERA and FIP between 2014 and 2015 has been nearly entirely due to his rate of home runs allowed: 0.3 HR/9 in 2014, 1.1 HR/9 in 2015.  Small sample size acknowledged.

Cosart has been a player that has consistently outperformed his FIP in terms of runs allowed.  However, he strikes out fewer than what he should do, and walks more than his share as well.  Until now, this has probably been due to a lack control of the strike zone.  It is possible that his recent problems with home runs are more due to command difficulties inside the strike zone (i.e. missing in happy places for the batters), and that is not a good thing to happen when you already walk 3-ish per nine, and strike out only 5-ish per nine.

Cosart threw 180-odd major-league innings in 2014, so he has carried a decent workload in the major leagues before.  He probably has a little more at this stage of his career to offer than Jordan Lyles, but Lyles has been pitching in front of a worse defence and in a much harder park to throw in.  However, the Astros would certainly not rescind the trade that netted them Francis Marte, Colin Moran and Handsome Jake, especially not with Jake's .379/.422/.621 April already in the bank.  The draft pick is the icing on the cake.

Nick Tropeano:
Nick Tropeano was the next starter to get traded, nearly immediately after the 2014 season.  Tropeano  has only five major league appearances under his belt, despite being the same age as Jordan Lyles.  he has one start this year with the Angels, where he went 6 shutout innings, giving up 5 hits and walking one, while striking out five.  He has an ERA north of 5 in AAA Salt Lake, having thrown 44 innings, striking out 43 and walking 15, while allowing 44 hits, including 4 home runs.

At this point, it is worth quickly acknowledging that the Angels starters and the Astros starters rank fairly evenly via Fangraphs: 12 and 14 respectively in the Majors in starting pitching, as defined by WAR.  Matt Shoemaker continues to get starts, despite a 5.44 ERA.  His peripheral stats are strong, however - an 8.42 K/9 and a 1.75 BB/9 indicate that positive regression is likely.  Jered Weaver, Garrett Richards, Hector Santiago and C.J. Wilson have all been solid and worth more than 0.6 WAR each.  The Astros are in a similar situation, with Keuchel, McHugh, Feldman and McCullers all worth 0.5 WAR or more, and with Roberto Hernández playing the role of Matt Shoemaker, but without the equivalent upside.

So Tropeano appears to not be getting starts with the Angels because of Matt Shoemaker's 2014, where he (Shoemaker) had an ERA of just over 3 in 125-odd innings.  I am not sure how much longer that will last, but Tropeano - much like Shoemaker - has never wowed scouts, and Tropeano's work in a tough AAA park for pitchers is solid, but unspectacular.  Tropaneo is doing little to force the Angels' hand in terms of getting more playing time.

Tropeano does not have a major-league track record that allows for reasonable analysis, but his AAA stats look a lot like a bunch of guys that the Astros already have.  So it is difficult to say whether he would be the guy to fill a rotation slot for a stretch drive.  The scouts would most likely say no, and there is nothing in the stats that would indicate than a significant improvement has occurred since he was traded.  The fact that he is not getting major-league innings for the Angels is significant.

Ironically, I would not mind betting that this trade is one the Astros would like to have back.  They traded Tropeano (a potential useful piece) and Carlos Perez (a defence-first catcher currently hitting .273/.283/.386 in 46 PA's) for Hank Conger, who has not really impressed behind or at the plate with a .156/.283/.311 line in 54 plate appearances.  It was a curious trade when it was made, with many of us thinking that Conger would replace Castro as the starter, but no further trades occurred.  I would love to pick the brains of the Front Office to see what they think of that trade now.

Mike Foltynewicz:
Folynewicz is the youngest and hardest throwing of the four pitches traded away by the Astros.  With Houston, he made 16 relief appearances in 2014, and appeared pretty hittable.  In 18 innings pitched, he struck out 14 while walking 7, and was good for a 5.30 ERA.  He allowed 3 home runs - again due to command problems within the strike zone, a straight fastball, and the lack of an effective offspeed pitch combining to highlight his struggles.

With Atlanta in 2015, he has been used exclusively as a starter.  His stats have been much better: a 3.96 ERA, with a 3.69 FIP in just over 36 innings pitched.  He has struck out 9.4 per nine (up from 6.8 in the major leagues last year) and walked 3.2 per nine (very close to 2014).  He still allows 1.0 HR/9 (down from 1.4 last year) and while small sample-size caveats apply for all of those figures, he has always been prone to the long-ball.

Folty's minor-league stats may shed more light on the situation.  In Oklahoma City last year, he pitched to a 5.08 ERA in 102.2 innings pitched, with an 8.9 K/9 and 4.6 BB/9.  He allowed 0.9 HR/9.  With Gwinnett this year, he has thrown only 21.2 innings, but with much better results: 2.08 ERA, 12.5 K/9 and 4.2 BB/9.  Those are the sorts of improvements that one may expect with an improved offspeed offering, so perhaps Folty has developed a feel for spinning the baseball.

But with Foltynewicz, the scouts have always been more bullish than the stat guys.  After all, he throws in triple-figures, has a great build for pitching, and has a smooth action.  The stats are only going to say so much, and the sample sizes remain small at this point.  The Gattis trade was eye-watering in terms of what the Astros gave up, and an effective Mike Foltynewicz is someone who would look mighty good in the Astros rotation right now, and into the summer.

Key word: effective.  An effective Mike Foltynewicz  No guarantee that that will continue.  Foltynewicz was always all about his ceiling - but right now he looks a little closer to that than what he was.

I will check back later in the season, and update you all.  The issue may become more acute, and Vince Velasquez may be the next guy to make the jump from AA to the Major Leagues.  The Astros won't get the opportunity to do over any of these trades, but it is still interesting to look at what has happened, and comment on the overall tactics recently employed.

Thanks for reading, and please feel free to discuss in the comments.