Wednesday, March 19, 2014

Roark vs Jordan

As I saw it the 5th starter role was Detwilers to lose. Perhaps he didn't last super long in games* but he was moderately effective, especially considering the role. Expand out his 2012 stats to a full year and he'd be about the 40th-50th most valuable starter in the NL. Given that there are at least 75, there's no shame in that. The Nats had one of the better #4s in the role of a #5.

Yet 2012 also saw Detwiler perform really well in the pen. It's something I actually spoke up about a few times, but before finishing that horn toot, I'll note I brought it up to explain why he wasn't as good a starter as you'd think, not because I thought he should be moved to the pen. The Nats needed another dependable lefty in there and they could use another solid long relief option. Moving Detwiler to the pen solves both issues. As Matt Williams said
"We're a better team with him coming out of the bullpen"
Except that's a half-truth.  They are a better team with Detwiler coming out of the bullpen... if their other options at 5th starter can pitch as well as Detwiler did. Can they?  He went over this in a cursory manner a few weeks ago, let's dive in now.

Taylor Jordan is 25**  His minor league career doesn't exactly inspire confidence. He was very good in rookie ball in 2009, struggled in A ball in 2010, did better in A ball in 2011, had Tommy John, then struggled again in A ball during his comeback in 2012. Not an auspicious beginning, but in 2013 though he looked like a different pitcher, mastering his control and breezing through High A and AA.

That's the type of pitcher Jordan is, a fantastic control pitcher who keeps his walks down (2.1 BB/9 in his minor league career) and is nearly impossible to homer off of (14 homers in 339 minor league innings).  He's not a strikeout threat (ignore what you see this spring) putting up mild K numbers in the minors. This is exactly what we saw in his major league stint. No walks, no homers, but no Ks. He's not a guy that's going to dominate. You'll get your hits but they won't be homers and I'll get out of the inning before anything bad happens.

The worst thing you can say about him is he lacks experience and thus 2013 could just be a fluke. That might have been true about the domination in the minors. abnormally low BABIPs and high LOB% suggest ERAs that were well under what they should have been. But in the major leagues we saw a .322 BABIP and a 66.9 LOB%. Those aren't lucky breaks. They seem more than reasonable. I think 2013 MLB Jordan is a fine guess at what he'll be going forward.

Roark is 27. One of the whatevers dealt to the Nats for Cristian Guzman, he slowly moved up the minor league ladder by attrition. He was mostly mediocre but never flat out terrible, and when the Nats had need of organizational depth, they kicked him from AA to AAA. After another blah year in AA though it looked like he "got it" in 2013. Although his surface stats didn't look any better than Jordan's in the minors (more homers, every slightly more walks, same strikeouts) he didn't suffer the same drop in stats when he moved up to the majors. Everything remained the same, if not better and he put up that great 1.51 ERA line.

Roark has morphed into something a lot like Jordan, after being more of a wild strikeout type guy in his early years.  2008 and 2009 showed a guy striking out a batter an inning, but walking too many and giving up too many homers. The 2010-2012 Tanner gave up some K's for some better control (around 0.8 HR/9, 2.8 BB/9, 7.3 K/9). Until finally arriving at that 2013 that was a lot like Jordan's. I don't think the endpoint was fluky. I think this is the pitcher Roark is now.

If Roark is Jordan then why the big discrepency in major league results? Part of it is Roark adjusting better, but part of it is things going his way.  Roark had spent the minors with a BABIP in the .280-.320 range. Last year he had a .258 BABIP in the minors, a .243 in the majors. His LOB% was on the high side (though not crazy) in the majors as well at 79.8%. And while Roark doesn't give up many homers a 2.6% HR/FB rate is something that no one keeps up for an entire year.

To put it another way, in 2013 we saw this
Lucky Jordan Minors: 1.00 ERA
Jordan Majors: 3.66 ERA

Roark Minors: 3.15 ERA
Lucky Roark Majors: 1.51 ERA

Very similar results, very similar pitchers (at least in 2013).  There's a good sense of how either of these guys will pitch in the majors. An ERA aropund 3.50, maybe up to 4.00 as the league gets comfortable against them, feels about right. Basically Detwiler, so the "gamble" that either of these two will pitch like Ross is probably a fair one. Ross offers more security for reaching the above numbers, but injury history and age suggest his time has passed.

The question then becomes who to choose and to me the answer is obvious and is why I started the analysis of each player with a statement of ages. Jordan is 25. Roark is 27. Jordan is 2 years and 3 months younger than Roark. Jordan is more likely to continue pitching at this level for more years than Roark making him a more useful piece in the rotation, or a more useful piece in a deal. Jordan also holds out more hope for improvement.

At this point the choice is simple - Taylor Jordan should be the #5.

*just under 5 2/3 a game. For reference ZNN has been around 6.5 innings the past 2 years, Gio under 6 1/3 and Strasburg just under 6, though he was given a short leash in 2012 and in 2013 he was at the same level as Gio).

**By the way 6 months younger than Strasburg, so when you consider Jordan a "prospect" think about that.

9 comments:

  1. Chinatown Express8:53 AM

    Does it matter that Jordan is coming off TJ? Have the Nats said anything about an innings limit for him? He went under the knife in 2011, so I suppose his 2014 is equivalent to Stras's 2013: ie, the leash is off (or is very long).

    ReplyDelete
  2. While I agree with everything you wrote, I think the Nats will name Roark as the #5. If the Nats did select TJ for #5 what would you do with Roark? Back to Syracuse or would he be the final guy for the bullpen? I think I'd rather see a power arm (Garcia / Treinin) or a more experienced guy (Ayala, Gonzalez) than Roark back there. Plus, if whoever isn't the #5 becomes the #6, I'd think you'd want to keep them stretched out.

    Someone made the comment (maybe on Nats Insider?) that Roark doesn't really have anything else to learn in AAA, while an argument could be made that TJ needs a little more seasoning. For that reason, it's probably an easier choice to go with Roark, even if not necessarily the right choice.

    ReplyDelete
  3. They shut Jordan down last season for his TJ innings limit. That's why he wasn't in the rotation in September.

    I agree with the logic behind starting Taylor Jordan in the 5th slot now.

    I remember reading throughout the winter about the various Nationals who were staying in Washington and working out at the park. The names were the usual suspects except for Roark. And Roark was at the Matt Williams press conference sitting beside Werth and Desmond. His constant presence at Nats Park might be playing a roll here.

    ReplyDelete
  4. I agree with Donald, it makes sense to make Roark the #5 if they are equivalent except in age.
    I think he'll win the job with the new manager, and its because of his attitude, poise. And he says the right things -- whereas, Jordan said "my outing was hurt by the Ump's calls" Let him get some experience in the minors.
    The latest WaPost article on Roark made him sound invincible against the Minor league Tigers! Had to give them extra outs each inning.

    Also, whats with the rumors that Houston wants to trade for Tyler Moore?

    ReplyDelete
  5. I can't argue with TJ as #5, although I do feel for Roark a bit; it seems like he's earned it. If TJ tanks in his first ten starts, I think they put Tanner in though.

    I do however love moving Ross to the pen. Our pen gets stronger by the day, and last year it was one of our (many) weaknesses.

    This team is better top to bottom than last years "world series or bust" team. Pretty exciting.

    Now lets see those results.

    ReplyDelete
  6. I agree with the conclusion, and here are a few other tidbits by looking a little deeper at the numbers. Apologies for all the math.

    Despite a lower k/9 rate, Jordan misses significantly more bats. His contact rate (total percentage of contact made when batters swing at all pitches) is 78% to Roark's 86.5, and his Z-contact rate (total percentage of contact made when batters swing at pitches in the strike zone) is 86% to Roark's 93%. That is 8-10% more swings and misses by Jordan. Your BABIP point partly explains why their traditional numbers were different, and I think it is reasonable to expect regression from both (Roark up, Jordan down).

    But Roark had another thing going for him last year that influenced results. And I still can't get my arms around whether it is a fluke or repeatable. Roark's Z-swing % (percentage of pitches a batter swings at that are in the strike zone) was 54.8%, compared to Jordan's 63%. So basically, batters were swinging at slightly more than half of Roark's pitches in the zone. To give this perspective, of all pitchers in 2013 that threw at least 50 IPs, that ranked first. Read that again - no other pitcher in 2013 had batters swing at fewer strikes than Roark, kind of like the opposite of a walk. (btw, #2 was Fister). So while I think that this could be a skill (pitch deception), it mostly feels flukey high. So if hitters swing at more strikes, and Roark puts up a normal BABIP, I think that you'll definitely see a 4.00ish FIP. That isn't a bad #5 by any means, but I get the sense that some people, while maybe a little dubious, are hoping for a Greg Maddux like transformation. I just don't see it, but I do think that he'll have a career as a good back end guy.

    Whereas the peripherals for Jordan suggest he might improve. Just taking K/9, if you look at the pitchers near Jordan's overall contact % numbers, you get Shields (7.7 k/9), Kuroda (6.7), Latos (7.9) and Corbin (7.69). All of them quite a bit better than Jordan's 5.05. I am not saying that he'll jump that high, it is a lot more complicated, but I am saying that the peripherals suggest it wouldn't be a complete shock to see him tick up to 6.

    ReplyDelete
  7. CX - It may a little, even still one year out, but that's a nice problem to worry about if he's been good all year it's mid Sept and he's at 180 innings. Can't choose Roark based on that scenario playing out.

    Donald - Makes sense but assuming (like the Nats do) that Cole and Giolito need major league time in 2015. You have to evaluate Jordan now. May not be space to do it next year.

    Sirc - Is Roark a former Diamondback? No. Bald? No. I don't get it. Why's he always around?

    D28 - attitude and poise? Says the right things? This ain't no miss america pageant! Give me the young 'un.

    As for Moore, Astros desparate for 1B, and Porter loves Moore. Doubt GM does though.

    Strasburger - and stay healthy



    ReplyDelete
  8. Chinatown Express11:48 AM

    Anyone else read the Nationals Journal sectio nin the Post today? Under the headline "Jordan looks sharp, makes case to start," Kilgore has a writeup about . . . Roark. Whoops. JOARK 4 MVP!

    ReplyDelete
  9. Wally - I was hesitant to use some of the fancier stats since I could only see them for the limited MLB time, not the minors.

    Even if I'm a bit unsure about the predictive value, it's a fascinating study on why what happened last year happened. People were taking more strikes vs Roark than anyone. Why? Pitch selection, location, movement? Catcher framing?

    CX - can we give them the same facial hair and send them out there
    in identical uniforms? Switch 'em in and out during the game? "Roark into the 11th and he doesn't seem tired at all!"

    Note : i am too lazy to do two picture searches to find out if the remotely look the same.

    ReplyDelete