On Phil Hughes And Regression

All good points. To me, the idea of regression as it applies to Phil Hughes is because people look at his past track record and believe he will regress to that. Those who have paid closer attention to how Hughes has approached the 2014 season -- the repertoire change, the mechanics, etc -- should see that the peripheral stats that he put up -- the lowered walk rate, the improved K rate, the lowered fly ball rate, etc -- are a direct result of that. ERA and record will fluctuate but with the new approach the peripheral stats should remain consistent.

The question is, can he sustain these new mechanics and approach?  I heard the same kind of comments about Suzuki's approach to batting last year and he went right back to having the same kind of results he had been having. In fact, we get these same kind of comments any time a player suddenly significantly improves upon his career norm. Most of the time, players fall back into old habits.  

So, I don't think believing he'll regress is unwarranted or due to not paying close enough attention. We'll see if 2014 was an outlier or the start of a new trend for Hughes.  I'm guessing he regresses some if only for the fact his season was so awesome, it'd be hard for too many pitchers to repeat that.  That includes pitchers who have a longer track record of very good than him.

I've posted several times in the past that Hughes sharply regressed after and successful season in which he reached career highs in innings pitched.  Seeing the video evidence on his mechanics, the new season may be different.  The 1st video shows a delivery that's out of control.  The 2nd more controlled, but I'm concerned that it seems from this view that it's all arm.

I'm still hopeful for another successful year, but they really need to pay more attention to the number of innings pitched.

Hughes FIPs this decade:

4.25

4.58 (half season)

4.46

4.50

Then last year 2.65.  That's a huge difference. More than a run and a half lower than his BEST season this decade and almost 2 runs better than his past couple years. I think it's unlikely that we can expect to see such a huge difference in performance from the rest of his performance this decade.   We all want that to continue but that FIP ranked him like 6th in the majors.  He's not the 6th best pitcher in the majors. He can continue to be very good but what would be sad is if he went to somewhere in the middle of what he had done prior to this year and what he did last year and people got disappointed because they were convinced he'd maintain that level of awesome based on one season.

Yeah, that was in jest. The point of the entire post is that those things *won't* happen.

 

 

It's not a huge difference when you consider what FIP is comprised of. First, he greatly reduced the number of walks (lowest in MLB). Second, he greatly reduced the number of home runs (HR/9 ranked 25th). Last, he had an improved strikeout rate (31st). All three of those things are the basis of FIP. 

 

FIP rankings don't tell the entire story of a pitcher but if he's doing those three things well, there is a good chance he'll have strong numbers in the future.

I know all about FIP, and you seem convinced past the point of any chance of convincing you other-wise, so I hope you are right.

If you are trying to throw shade on Hughes' potential, by all means go ahead but if the only thing you are citing is his past track record then it is not a convincing argument for all the reasons listed in my article. 

Again, I hope you are right.

The Twins will have to hire Rick Anderson to be Hughes' "mental coach."

The Twins will have to hire Rick Anderson to be Hughes' "mental coach."

If that's what it takes to have Hughes have another season where his FIP is better than Kershaw's (the only current SP with 1000 IP or more and a career FIP under 3.00), King Felix's, and Pedro's career FIPs, then DO IT! :-)

I think the point he's trying to make is that some stats are predictive and some stats tell us what happened.  FIP isn't a very good predictive stat although you can argue that the ingredients in FIP are.  That's why the changes in those rates might better explain Hughes outside of Yankee stadium.  Most of the recent articles on fangraphs go in that general direction - "he's a good pitcher but real changes have made him better, etc, etc".  I don't think anyone rationally expects him to be a 6 WAR pitcher again but he shouldn't fall back to the 2 WAR Yankee pitcher either.  

Exactly, FIP isn't predictive.  We shouldn't be using that to predict next year. xFIP is better for that and it had last year at 3.18

Somewhere in between last year and what he's done before is where I was saying we should expect him to be. And that I hope people won't view that as a negative.  Look at it this way, he had the 6th best FIP and that resulted in like the 42nd or so best ERA due to our defense. Imagine what happens to his ERA if he even goes to a 3.50 FIP (which is still quite good), with the defense behind him, he's looking at a mid 4.00 ERA and some people will freak.

Steamer predicts Hughes to 3.89 FIP and a 2.5 WAR.

I'm shocked! No one seems to notice the biggest difference between Hughes from NY to MN... it is so obvious that it must go overlooked. The Yankees failed to notice that what Hughes needed the most was in fact a beard. That thing reeks of confidence.

Well, as a young padawan, you should know that having a beard was the reason Obi-Wan and Qui-Gon were so confident :-)

I don't think that sways many people - Steamer usually predicts low.  In fact, they predicted 6 pitchers to be above 4 WAR and only 3 to pitch 200 innings (unless they haven't issued all players/teams yet.  I confess to not caring).  Hughes' projection is the 29th best WAR in MLB.

ZIPS is the most accurate projection model for baseball out there.  I look forward to seeing theirs.  And, I need to note, I said Steamer 'predicts' in my previous post.  These models don't actually predict.  They project. It's a subtle difference.  I've seen projections for next year where no team is projected to win 90 or more games in the AL.  I really don't think that the people who do the projections are really predicting every AL team to fall short of 90 wins. So, yeah, grains of salt for sure.

His GCTWTG percentage last year was 75%.  The last projection I saw for this was the same so I am not looking for a huge regression.

This is much more accurate than FIP, Zips or Steamer.

One thing to keep in mind, and maybe I am peculiarly aware of this since I had Phil on my fantasy team back in his early days (thinking I was all smart in grabbing him early in drafts), is that Hughes was touted as being the kind of pitcher we saw in 2014 back 8 years ago. Things went astray and sometimes the talent just is really there but doesn't translate to the field for whatever reason (I maintain that he was never a guy to succeed in New York, in general). That changes all talk of "regression" somewhat.