Starter or reliever?

Yes, pitchers who put up good AAA numbers also tend to be better major league pitchers (i.e., in your terminology, they have lower MLE ratios). If you're going to claim that good knuckleballers are particularly special in this regard (i.e., that knuckleballers with good AAA numbers project to be better in the majors than non-knuckleballers with good AAA numbers), well, I'm going to be skeptical until I see some more than a theoretical argument and one example.

I can also think of some notable counterexamples. Charlie Zink had a 2.84 ERA in 174.1 innings in Pawtucket in 2008, but that success did not carry over and his career ended soon thereafter (yes, that's cherrypicking Zink's best season somewhat, but I'll note that from 2006-2008, he had a combined 3.70 ERA). Charlie Haeger threw 170 IP of 3.07 ERA ball in AAA in 2006 and similarly never amounted to much. 

Let me clarify what I believe to be the case about acquiring a second starter. It may be that we're all more on the same page than we think we are.

There are two types of moves that teams contemplate in any offseason: holes to fill, and ways to improve. Although obviously they exist on a continuum, I still believe it's useful to think of them as different, because there are often clear-cut cases of each.

Backup catcher is a hole to fill. 3B, if you don't think Mookie Betts should be given a shot there, is a hole to fill. While Cecchini could likely step in and be an adequate MLB starter, he would likely be below average, and so he'd be better off at AAA working on hs defense and consolidating the offensive gains he seemed to make late last year. Holt could play there as well, but he'd also be below average and has much more value as a bench player. WMB does not now project to be starting-caliber. So you have a hole.

I don't think there's any argument that 5th starter is a hole to fill, once you've added one frontline starter. Brandon Workman is 7th on the depth chart to fill that job (by season's end), and he's already established himself as an adequate MLB 5th starter. Steven Wright, even if you're not high on him like I am, looks like he's certainly an adequate MLB 5th starter. Allen Webster is one if he continues to pitch like he did at the end of last year in MLB, and Matt Barnes is one if he continues to pitch like he did at the end of the PawSox season. And then you have Owens, Rodriguez, and Johnson in the mix by mid-season. (I'm assuming Ranaudo is dealt, since they'll need to clear room on the 40-man roster.)

So adding another top starter is one of the ways to improve the team. You are looking at this, by mid-season:

Potential acquisition, e.g., Samardzija, Latos, etc.
versus
The fourth best pitcher among Buchholz, Kelly, De la Rosa, Wright, Webster, Barnes, Owens, Rodriguez, and Johnson.

Now, no one is saying that's not likely to be an upgrade. (It might not, however, be a large one.) Furthermore, at the start of the season it gives you important extra depth; it means that if one of the 5 projected starters is hurt in ST, it's Wright (who has been bumped to long reliever / spot starter in this scenario) who is the 5th starter rather than Webster, who has a significantly higher chance of being inadequate.

What I take issue with is the characterization of adding a second starter as a "need." It's not a need, and thinking of it as a need runs the risk of getting you in trouble. Instead, it's one of the ways to make a good team better.

When you have a need, you have to fill that need, even if you're not thrilled with the best available solution, or by the price you have to pay for it (in money or prospects). But there are almost always multiple ways to improve the team, and choosing to do so is always an option.

In this case, if Jason Heyward is unavailable, or if the Braves want a prospect package for him, then I have no problem with trading Cespedes for a second frontline pitcher. He's a spare part, so you ask yourself how you can improve the team by dealing him, and that's one obvious answer.

But if they can trade Cespedes for Heyward (and I think they have a decent chance to do so), or for a 3B, or even for a shutdown late-inning reliever with a few years of team control left (the theory being that such a guy has extra value in the post-season), then it might not make sense to deal for a second pitcher. Another way to improve the team would be to deal for Stanton, and maybe the prospects it would take to get that pitcher are the prospects you'd need to eventually get Stanton. This may be especially true if they've already dealt prospects for Hamels rather than signed Lester.

The bottom line is that adding a second pitcher should simply not be thought of as a need. They need a frontline pitcher, they need a 3B, and they need a LH bat (getting two would be a way to improve). I would argue that they also need to leave enough chips in the farm system to trade for Stanton at the next deadline. If after you've done all that, if trading for a second frontline pitcher is doable, and is the best way to make the team better, by all means do so. But thinking of it as a move you need to make runs the risk of using finite resources incorrectly.

Eric-

Your claim that Wright is "sure fire" #1 or #2 pitcher is spurious and not supported by facts or evidence.

There are players on that list that had multiple strong minor league seasons. Dickey btw, wasn't one of them.

Knuckleballers are always highly risky because it is not uncommon for their statistics to suddenly fall apart. Even Wakefield experienced this and was released as you recall.

You can't declare Wright to be Phill Niekro because he has hit some sort of arbitrary performance bar in the minors.

The fact remains that virtuallly every knuckleball pitcher in the last 20 years has failed. Many, with track records of success as good as Wright's.

That said, I actually think I'm higher than most folks on Wright. I'd put him a small step behind or maybe even on par with Ranaudo on the SP depth chart (though, for what it's worth, I'm also pretty low on Ranaudo), and I think he's a major-league caliber pitcher. He certainly has mid-rotation upside, but unlike Eric, I think he's also one of their most high-variance pitchers, and I'm not comfortable with him being the fifth or sixth starter when camp breaks. As such, I think adding a second starter this offseason should be the plan, with the caveat that if FA or trade prices get too high, I could be talked into getting a reclamation project (Morrow, Anderson, etc.) to compete with Webster/Ranaudo/Barnes/Wright for that last rotation spot (I have Buccholz, Kelly, and RDLR penciled into the other three).

He pitched the most innings he ever had by far last year. I think it's a very poor assumption that he will just always suck after 100 innings, and I would keep him as a starter.

Well, the age point you make is very deceptive, whether you intended it to be so or not (I'm assuming the latter). First off, let's speak correctly here - he turns 26 in March 2015. You make it sound like he was 26 this season.

However, the bigger point is that his progression to the majors was extremely abnormal. He signed at 18 (weighing 130 pounds and throwing 89-91, btw), spent two years in the DSL, and didn't come stateside until he was 20 for rookie ball (in which he may have gotten hurt, based on his only having pitched in 5 games, but I'm not sure). He didn't make it to full-season until 2010 at age 21, jumped from Low A straight to Double-A, throwing about 100 innings (which is normal for that age/level progression) and he was in the majors by June of the next season at age 22, which is bonkers.* His last start of that season is when he blew his arm out, and he basically missed all of his age 23 season rehabbing. His age 24 season was spent on a restrictive program to build him back up, as we know, and this year he was 25.

So there is a reason why he has yet to throw 150 innings without either getting hurt or wearing down. That reason is that in what are essentially three cracks at getting past that point, in one year he got hurt and needed Tommy John, in one year the club had him on a restrictive innings limit that he had no control over, and one year was this year and he did, in fact, predictably falter. You're basically looking at a two-season sample, so pointing to his age is kind of unfair.

I don't necessarily disagree with your assessment regarding his arm action, delivery, etc. My point is only that history, in his case, is borderline useless in judging his ability to pitch deep into a season. He spent less than a season and a half in full-season minor league baseball and skipped both High A and Triple-A entirely before making his MLB debut. He's not a guy who made a normal progression through the minors. It's entirely possible that he can develop into a horse who can give you 180-200 innings and that wouldn't be crazy.

* - The closest Red Sox equivalent I can find from this season would be if Mario Alcantara, who started the year in Greenville's rotation at age 21 in his first full-season assignment, suddenly started sitting mid-90s, got a midseason promotion to Portland this year, and then jumped from Portland to the majors next June. That's how fast this all happened.

I have no idea - I just don't have that kind of encyclopedic knowledge of the backgrounds of pitchers.

You'll just have to settle for me saying that I'm not saying that's the likely result. I'll give you that the fact that he tired makes it more likely that is what will happen next time than it would if he hadn't tired. That's kind of obvious. But I'm just saying that to only cite his age is misleading given how little experience he has pitching as a professional in the U.S. given that age. Unless you're faulting him for injuring his elbow, missing the next year because of it, starting his rehab late because he got traded (forgot about this in the above post, but remember, he didn't pitch rehab games in August/September 2012 because he had to be a PTBNL), and then being put on restrictions the year after that, he's had exactly two chances to pitch entire seasons in full-season leagues - his Low A/Double-A year and last year. If it happens again this year, I'm on board with it being a pattern, but I just don't think there's a pattern yet.

The point was that his progression appears to be unique. Good job thinking of the comp with Pedro in terms of his quick rise through the minors, but Pedro also had his first full-season year at age 19, whereas RDLR's was at 21, which is part of the issue. (Also, can you fault a guy for not being Pedro, really?)

I'd assume that most pitchers who shoot through the minors like that wind up being studs. But I'm not sure how many hit the injury hurdle that RDLR did at the time he did. If, to use a really rough example, Jose Fernandez comes back in 2015 with restrictions that keep his innings down, and then at the end of 2016 he tires when he pitches way more than he did in 2015, would it be fair to say, "he's going to be 24 and the only season in which he passed 150 innings he wound up needing Tommy John!" The only differences in that case would be a) that Fernandez was 2 years younger than De La Rosa, even though they had the same amount of full-season pitching experience basically when they got hurt, and b) that Fernandez blew his arm out later in the season than RDLR did.

At any rate, if you disagree, I'll agree to disagree with you. I just think it's too soon to draw conclusions. A lot of people assumed that De La Rosa was going to tire down the stretch - many of us were thinking he'd get shut down at the end of August/early September at one point.

He did? How do you figure? He had one full season in the minors.

Looking at players who've had TJ would be the way to go. Good call. I don't care enough to do the legwork, honestly, and don't have time, but if someone wants to take a gander, I wonder if there's any comparables.

Yeah and during that season he had 134 innings pitched 534 batters faced.  RDLR by comparison had one season before he reached the majors and before he got hurt with 110 innings and < 450 BF.  That's about a 20% difference.

I am curious to see if there is a comp as well but am not sure I have time.  There is an all-time list of TJ patients out there.  You would be looking at players who had at somewhere between age 22 and 24.

Please keep up the good work on the podcast.  I listen regularly and very much enjoy it.

I'm just saying that there isn't a huge sample size of pitchers to look at.  There are much stricter innings limits now and pitchers tear their elbow more frequently.  So even looking at someone like Kershaw, he pitched over 230 innings as a 19 year old in 2007.  That would never ever happen now so you can't even compare.  He'd probably get 100 innings now as a 19 year old if that.

You'd basically have to be looking at pitchers about as old as RDLR.