Friday, May 16, 2014

Has the Tandem Rotation in the Minors Disappeared??

I think that most readers with an interest in the Astros know about the Tandem Starting Rotation in the minor leagues.  As this excellent summary article says, it is often used in the lower minor leagues in a lot of systems, but the fact that the Astros use it throughout their affiliates at all levels raises some eyebrows.  And, it claimed an early victim / got shoved under a white-hot spotlight, because when your 1-1 has to head back to extended spring training with an ERA of 6.23, that is never going to generate headlines that are supportive of this idea.

In late April, the Astros formally (and possibly temporarily) disbanded the tandem starting rotation at OKC.  And in the last few PreStros Recaps, there seems to have been some guys who have been stretched out a little more than normal, at least in terms of innings pitched.  This article is to look at whether this represents the ditching of the tandem rotation.

We already know that OKC do a six-man non-tandem rotation (although a six-man tandem rotation would be awesome!), so we won't bother to analyse them.  Lets look at the other affiliates...

Corpus Christi Hooks

First two pitchers of the last 10 starts:
May 6: Heidenreich (5) followed by Rogers (4)
May 7: Cruz (4) followed by Jankowski (3)
May 8: Smith (3.2) followed by Sogard (1.2)*
May 10: Shirley (4, but 91 pitches) followed by Bellow (2)
May 11: Rollins (2) followed by Rodgers (5)
May 12: Hauschild (5) followed by Heidenreich (2.1)*
May 13: Cruz (0.1) followed by Robinson (3.2)*
May 14: Cain (3) followed by Jankowski (4)
May 15: Smith (5) followed by Shirley (4)

* - one of the two pitchers removed presumably due to ineffectiveness

Lancaster JetHawks

First two pitchers of the last 10 starts:
May 5: McCullers (1.2) followed by Minaya (4.1)*
May 6: Hader (5) followed by Devenski (4)
May 7: Holmes (4) followed by Osborne (2)
May 8: Hauschild (5) followed by Minor (2)
May 9: Velasquez (3) followed by Westwood (5)
May 10: Emmanuel (5) followed by McCullers (4)
May 11: Devenski (4.2) followed by Rodriguez (0.1)*
May 13: Hader (6) followed by Osborne (2)
May 14: Holmes (4.1) followed by Rodriguez (1.2)*
May 15: Westwood (7.2) followed by Lambson (1.1)

* - one of two pitchers removed presumably due to ineffectiveness

Quad Cities River Bandits

May 5: Thurman (5) followed by Lee (3)
May 6: Gustave (3.1) followed by Cotton (2.2)*
May 7: Grills (5) followed by Frias (3)
May 8: Feliz (2) followed by Christensen (2)
May 9: Houser (6) followed by Minnis (0.2)
May 10: Thurman (4) followed by Lee (4)
May 11: Feliz (3.1) followed by Christensen (1.2) - Gustave (4.0) finished
May 13: Frias (4) followed by Morton (2)
May 14: Grills (7) followed by Cotton (1)
May 15: Houser (5) followed by Comer (3)

So, analysis time...

Firstly, this kind of analysis does not lend itself to much, unless a very clear confirmatory patterns is seen.  Which would be Pitcher A and Pitcher B going every 4 days, in a random, possibly alternating pattern.  Those clear, confirmatory patterns are not really seen in the above 30 games.  That said, we are working using rotations - it may be better to work using pitch counts.  But too late now!

At Corpus, there is no obvious pitcher-pairing.  Neither at Lancaster.  In Quad Cities, Thurman and Lee go, then go again 5 days later.  Feliz and Christensen go, then go again 3 days later, but the length of those starts seems to lend itself to the idea of a 'pen start.  Hardly what was spoken of at the beginning of the season.

But there are four potential sources of noise here that I can easily see.  Firstly, pitchers are getting promoted - Kyle Smith went from Hi-A to AA, Kent Emmanuel went from Lo-A to Hi-A to replace him.  Secondly, a pitcher that gets lit-up and has a short outing could potentially come back earlier than a pitcher that doesn't, and therefore the pairs can easily get muddied.  Thirdly, especially in the lower levels, guys may be having trouble with endurance and the rigours of new professional routines, and therefore not be ready to go on time, and get knocked back a day or two.

And fourthly (and I think most importantly in terms of this analysis), is a clue from the article linked above:
"A tweak this year is that their pitchers will get a five or six day break worked into various points of their season, the first possibly a month or so in."
I wonder whether much of the noise that muddies this analysis is a result of the first month of the season being finished, and the extra breaks are starting to be worked in.  Such as, Evan Grills, who went 7 on May 14, last pitched 7 days prior to that on May 7 (resulting in 6 rest days).  And Kyle Westwood, who went 7.2 on May 15 had 5 rest days, after throwing 5 innings on May 9.  So, a couple of those guys that had extra breaks came back and threw a greater-than-expected number of innings.  And really, when I remember what gave me the notion that the idea of the tandem rotation may have died, it was those two 7+ inning starts that got me mulling the idea.

That said, this may represent the very beginnings of a shift in the Astros' minor league thinking, but is more likely to be just noise, I would think.  Bears watching.  If anything interesting pops up, I will look again next month.

Please feel encouraged to leave comments below.  Thank you for reading, especially if you got this far down in the article.