The unprojectable Clay Buchholz

Morgan's Magic Snowplow · Dec 27, 2014

Wow. A lot to ponder here but this thread is awesome. I love this way of visualizing the data.

Murby · Dec 27, 2014

iayork said:
Clay's been a regular starter since 2010 and in that time he's had three reasonably healthy seasons, a couple where he was very effective as long as he was pitching, a couple where he was decent but not great (ERA+ of 124 and 92), and one season (2014) where he was more or less healthy but just awful: ERA+ of 72, worst ERA in the league for qualifying pitchers, a little better but still pretty mediocre in terms of xFIP. What went wrong? What can we expect for 2015?

Short answer is that I have no idea, but maybe someone will have better ideas than me.

I looked at his pitch mix and effectiveness in 2014 vs 2013. Overall, his pitch mix was roughly similar, with a 4-seam and a cutter dominating. He used his curve relatively more in 2013, swapping his 2-seam for it in 2014, probably because his curve pretty much sucked in 2014. (I'm relying entirely on PitchFX for pitch identification here.)

In these plots, the size of each bubble represents the proportion of pitches in each zone, and the color represents its effectiveness, based on total bases as a percent of pitches in each zone. Green is good, red is bad, grey is league average (overall, not for each sub-zone). These are based on pitches that are not balls, so it's showing the effectiveness of the pitch in either drawing bad contact, or fooling the batter into a called strike.

2013:

2014:

(All his years since 2010 are shown here.)

His 4-seam didn't change all that much in effectiveness, and though his 2-seam isn't all that good it's not drastically worse in 2014 than in 2013. The biggest differences were in his curve and, especially, his cutter, which were both very effective pitches in 2013 and pretty bad in 2014.

How did they look in each year? Here are animations of all his curves/cutters for each year.
Cutter, 2013:

Cutter, 2014:

Curve, 2013:

Curve, 2014:

The most striking thing about these comparisons is how similar they are. Velocity is very similar (same with the other pitches I don't show here), and movement seems pretty similar too. To me, that's more encouraging than not: His pitches don't seem to have dramatically changed their quality from a time when they were highly effective, so there's no obvious reason why they can't be effective again.

There's only one thing that I do see as being rather different year to year, and that's the fact that in 2014 his pitches cluster a lot more tightly. Could he have simply been more predictable in 2014? I mean, I'd love to blame Pierzynski, and it's worth noting that before AJ was DFA'd Clay's xFIP was 5.11, while after August (when he and Vazquez should have been on the same page) his xFIP was 3.66. But I recognize I'm grasping at straws here.

Otherwise, nothing jumps out at me. Clay had an awful year in 2014, but I don't see an obvious reason why he should have another bad one in 2015. Nor, unfortunately, do I see a reason why he shouldn't have another bad one.

The similarity struck me as well. Interesting concept that he was too fine on his pitches in 2014, thus hitters knew better where they would be, thus they'd be easier to hit? Would be interesting to read some other people look at the data and provide analysis.

Fireball Fred · Dec 27, 2014

I look at Buchholz's record differently. To me he's had 1.5 very good seasons in seven years, and much of the rest of the time, when he wasn't hurt, he's been pretty bad. That's tantalizing but at his current age no longer promising; he's unreliable in two respects, and cannot be counted on. When he is pitching a good game, it seems to me, he has too often been left in an inning too long - he'd be more effective, perhaps, if his limitations were accepted.

PrometheusWakefield · Dec 28, 2014

iayork said:
Clay's been a regular starter since 2010 and in that time he's had three reasonably healthy seasons, a couple where he was very effective as long as he was pitching, a couple where he was decent but not great (ERA+ of 124 and 92), and one season (2014) where he was more or less healthy but just awful: ERA+ of 72, worst ERA in the league for qualifying pitchers, a little better but still pretty mediocre in terms of xFIP. What went wrong? What can we expect for 2015?...

The most striking thing about these comparisons is how similar they are. Velocity is very similar (same with the other pitches I don't show here), and movement seems pretty similar too. To me, that's more encouraging than not: His pitches don't seem to have dramatically changed their quality from a time when they were highly effective, so there's no obvious reason why they can't be effective again.

The comparisons are similar because Clay Buchholz was mostly the same pitcher in 2014 that he has been for his whole career. Here's his xFIP by year since 2007: 3.70, 4.24, 4.04, 4.07, 4.28, 4.43, 3.41, 4.04. Ignore the ERA and he's a standard, fairly predictable low 4s FIP pitcher. In 2013, he had a remarkably lucky season with a 83.7% strand rate and a 4.5% HR/FB rate, combined with a significant jump from career norms in strikeouts. In 2014, his luck reversed, his strand rate fell to an atrocious 62.1%, he stopped getting lucky on home runs and his strikeouts were more in line with his career averages. That's it. The strand rate, in particular, is the central thing you have to understand if you want to understand Buchholz' 2014 and it has no real relationship to actual performance or ability.

There's no big mystery here. This is the normal variation of pitchers who are about the quality of Clay Buchholz. You look at other guys I'd put in his category like Ryan Dempster or Edwin Jackson or Ervin Santana. Some years the luck stats bounce their way and they look like solid #2s at least. Other years the luck stats go against them and people act like something must have happened to their talent. Buchholz isn't especially unusual, and unfortunately he isn't especially talented, just another decent pitcher who looks good when things break his way and bad when they don't.

canyoubelieveit · Dec 28, 2014

iayork said:
With all respect, that's not interesting. Clay has shown that he can pitch at an extremely high level for a significant part of a season. Saying that you think he's just not a reliable pitcher, or that he's just not a good pitcher, doesn't go anywhere. What is interesting is to ask whether the high-level pitching is something that can be defined or predicted.

If you believe he's been left in games too long, that's not interesting. If you have data showing that he's been left in games too long, that is interesting.

I actually found his post to be more interesting than yours (i.e. your post quoted above, not the interesting one that started this thread). His post made me realize that Buchholz had been effective at a quality level for a lower percentage of the time than I had previously realized, and at least he introduced a new idea (that perhaps his numbers are skewed by one bad inning at the end more often than other pitchers). Not everyone needs to write a thesis.

It's not like his post was some kind of poorly worded gamethread drivel. Some of us also enjoy posts that simply express relevant new ideas from a fan's perspective.

C4CRVT · Dec 28, 2014

PrometheusWakefield said:
The comparisons are similar because Clay Buchholz was mostly the same pitcher in 2014 that he has been for his whole career. Here's his xFIP by year since 2007: 3.70, 4.24, 4.04, 4.07, 4.28, 4.43, 3.41, 4.04. Ignore the ERA and he's a standard, fairly predictable low 4s FIP pitcher. In 2013, he had a remarkably lucky season with a 83.7% strand rate and a 4.5% HR/FB rate, combined with a significant jump from career norms in strikeouts. In 2014, his luck reversed, his strand rate fell to an atrocious 62.1%, he stopped getting lucky on home runs and his strikeouts were more in line with his career averages. That's it. The strand rate, in particular, is the central thing you have to understand if you want to understand Buchholz' 2014 and it has no real relationship to actual performance or ability.

There's no big mystery here. This is the normal variation of pitchers who are about the quality of Clay Buchholz. You look at other guys I'd put in his category like Ryan Dempster or Edwin Jackson or Ervin Santana. Some years the luck stats bounce their way and they look like solid #2s at least. Other years the luck stats go against them and people act like something must have happened to their talent. Buchholz isn't especially unusual, and unfortunately he isn't especially talented, just another decent pitcher who looks good when things break his way and bad when they don't.

I thought your post was underselling how good he was in 2013 so I went on fangraphs to look at the other stats to see what stood out about 2013 that might disprove the gist of your post. My guess looking at the numbers is that his superlative ERA in 2013 was partly (or even mostly) luck. But he did pitch better in 2013. Even if that only accounted for some of the reason behind the low ERA, his overall numbers were good.

His WHIP was 1.02 (career average 1.33)- obviously some of this is BABIP.
His K/9 was 7.98 (career average 6.88)
His BB/9 was 2.99 (career average 3.33)
FIP 2.78 (career average 4.06)
xFIP 3.41 (career average 4.08)

I went on the fangraphs pitching leaders for 2014 and sorted by xFIP. A 3.41 xFIP would have ranked #26 in all of baseball last year so your conclusion is more or less spot on. If he replicates his xFIP of 2013 (which we could guess is his ceiling), he's a solid #2.

Food for thought: I went on B-ref to look at his comps. #1 comp? Wade Miller (shudders).

The problem with Buchholz is that when he was just a wee lad, he dazzled us all with his potential. He was touted as having "ace stuff" and proved that billing spot on when he first came into the league. He's since more or less disappointed us with his results for years and finally seemed to be putting it together and became the star player that many of us had long hoped he'd become. So because he's flashed brilliance, I keep hoping that he'll eventually just be brilliant. But that hope seems to be just that; hope. He is no longer the 'future ace', he's just a frustratingly inconsistent guy who has (had?) great stuff who will never live up to his potential.

EllisTheRimMan · Dec 28, 2014

PrometheusWakefield said:
The comparisons are similar because Clay Buchholz was mostly the same pitcher in 2014 that he has been for his whole career. Here's his xFIP by year since 2007: 3.70, 4.24, 4.04, 4.07, 4.28, 4.43, 3.41, 4.04. Ignore the ERA and he's a standard, fairly predictable low 4s FIP pitcher. In 2013, he had a remarkably lucky season with a 83.7% strand rate and a 4.5% HR/FB rate, combined with a significant jump from career norms in strikeouts. In 2014, his luck reversed, his strand rate fell to an atrocious 62.1%, he stopped getting lucky on home runs and his strikeouts were more in line with his career averages. That's it. The strand rate, in particular, is the central thing you have to understand if you want to understand Buchholz' 2014 and it has no real relationship to actual performance or ability.

There's no big mystery here. This is the normal variation of pitchers who are about the quality of Clay Buchholz. You look at other guys I'd put in his category like Ryan Dempster or Edwin Jackson or Ervin Santana. Some years the luck stats bounce their way and they look like solid #2s at least. Other years the luck stats go against them and people act like something must have happened to their talent. Buchholz isn't especially unusual, and unfortunately he isn't especially talented, just another decent pitcher who looks good when things break his way and bad when they don't.

Great post. If we as fans or the Sox FO have any notion that Clay can magically bounce back and lead this staff in 2015 then we're simply wish casting. Add to his solid but unimpressive performance since 2007 his injury history and his psychological makeup (chronically throwing to first, for example) and we better hope that some combination of a Scherzer or Shields signing, a Hammels trade, the emergence of Miley, Porcello or Kelly etc. and/or the offense is a clear powerhouse. The xFIP numbers and my impressions of watching him pitch for many seasons suggest that Clay is a nice complementary piece to a starting rotation, but it is doubtful that he'll turn into anything more than that.

This is a disappointing realizations for me, because I was hoping we'd get "dominant Clay" in 2015, That version does not seem to exist, at least for any sustained period such as a full season. I think the thread title is incorrect. Clay is very projectable, especially if you ignore the high chance of injury.

nvalvo · Dec 28, 2014

PrometheusWakefield said:
The comparisons are similar because Clay Buchholz was mostly the same pitcher in 2014 that he has been for his whole career. Here's his xFIP by year since 2007: 3.70, 4.24, 4.04, 4.07, 4.28, 4.43, 3.41, 4.04. Ignore the ERA and he's a standard, fairly predictable low 4s FIP pitcher. In 2013, he had a remarkably lucky season with a 83.7% strand rate and a 4.5% HR/FB rate, combined with a significant jump from career norms in strikeouts. In 2014, his luck reversed, his strand rate fell to an atrocious 62.1%, he stopped getting lucky on home runs and his strikeouts were more in line with his career averages. That's it. The strand rate, in particular, is the central thing you have to understand if you want to understand Buchholz' 2014 and it has no real relationship to actual performance or ability.

There's no big mystery here. This is the normal variation of pitchers who are about the quality of Clay Buchholz. You look at other guys I'd put in his category like Ryan Dempster or Edwin Jackson or Ervin Santana. Some years the luck stats bounce their way and they look like solid #2s at least. Other years the luck stats go against them and people act like something must have happened to their talent. Buchholz isn't especially unusual, and unfortunately he isn't especially talented, just another decent pitcher who looks good when things break his way and bad when they don't.

Sox Scout posted a Luck/Skill plot to twitter a few weeks ago that was basically FIP or something similar plotted against an aggregate of strand rate and BABIP. Buchholz (and Masterson!) were way over at the unlucky end.

I don't imagine he'd object to me sharing it here. Take a look: link.

czar · Dec 28, 2014

PrometheusWakefield said:
The comparisons are similar because Clay Buchholz was mostly the same pitcher in 2014 that he has been for his whole career. Here's his xFIP by year since 2007: 3.70, 4.24, 4.04, 4.07, 4.28, 4.43, 3.41, 4.04. Ignore the ERA and he's a standard, fairly predictable low 4s FIP pitcher. In 2013, he had a remarkably lucky season with a 83.7% strand rate and a 4.5% HR/FB rate, combined with a significant jump from career norms in strikeouts. In 2014, his luck reversed, his strand rate fell to an atrocious 62.1%, he stopped getting lucky on home runs and his strikeouts were more in line with his career averages. That's it. The strand rate, in particular, is the central thing you have to understand if you want to understand Buchholz' 2014 and it has no real relationship to actual performance or ability.

There's no big mystery here. This is the normal variation of pitchers who are about the quality of Clay Buchholz. You look at other guys I'd put in his category like Ryan Dempster or Edwin Jackson or Ervin Santana. Some years the luck stats bounce their way and they look like solid #2s at least. Other years the luck stats go against them and people act like something must have happened to their talent. Buchholz isn't especially unusual, and unfortunately he isn't especially talented, just another decent pitcher who looks good when things break his way and bad when they don't.

I agree with the overall thrust of this point, however, feel compelled to point out that the strand rate and HR/FB are normalized in xFIP/SIERA which showed Buchholz with career-best marks in 2013.

What that tells me is that a Buchholz who can post a K% near 23% instead of 16-17% can be an extraordinary effective (top-10 AL) pitcher. Obviously, with his career punchout rate sitting at 18%, it's questionable as to how fluky that number was (especially given the number of called strikeouts he got) but a large reason he's a mediocre 4.00-type starter is the sheer number of balls that get put in play against him (coupled with his "eh" walk rate).

A little part of me has always wondered if the concerted push to pitch to weak contact (Buchholz has even admitted this) has hurt his bottom line. I've always thought that Buchholz achieved a large part of his success in AAA and his early career "working backwards" (I'd have to go chart his no-hitter, but he certainly threw many more curves/changes than he does in a typical game these days). Unfortunately, this also meant running up pitch counts getting him yanked in the 5th-6th innings.

His two- and four-seam fastball have well-below-average SwStr%. It'd be interesting to see if Clay Buchholz for 6 innings throwing all sorts of offspeed stuff would be better than Clay Buchholz for 7-8 innings relying on ball in play behavior to sustain solid starts. Alas...

Reverend · Dec 28, 2014

canyoubelieveit said:
I actually found his post to be more interesting than yours (i.e. your post quoted above, not the interesting one that started this thread). His post made me realize that Buchholz had been effective at a quality level for a lower percentage of the time than I had previously realized, and at least he introduced a new idea (that perhaps his numbers are skewed by one bad inning at the end more often than other pitchers). Not everyone needs to write a thesis.

It's not like his post was some kind of poorly worded gamethread drivel. Some of us also enjoy posts that simply express relevant new ideas from a fan's perspective.

Eh, the problem with FF's post is that he is effectively begging the question: iayork is wondering if we can figure out if there is hope for Buchholz based on our data, and FF posited the answer as "No."

What assuming "No" simply because he's older doesn't consider any sort of causal mechanism as to why we sometimes have gotten Good Buchholz and other times have gotten Evil Buchholz. He has been injured at times, yes, but they have not been similar injuries, so to the extent that injuries cause poor play, there's still the possibility we could turn the Good Buchholz card next year. Also, in 2012, he forgot how to through his frickin' change-up--and there's reason to believe that happened again this past season as well.

The point being that we might be able to learn something by considering what caused different outcomes at different times than just saying that we think he's bad too often.

You can kinda see a not fully thought out idea about a player in things like the claim that we've only gotten 1.5 very good seasons out of him--if you throw out the first six starts in 2012 when Chuckles had apparently forgotten how to throw his change, his ERA comes down to 3.62 for the rest of the season which is pretty good for 23 starts.

maxotaur · Dec 28, 2014

canyoubelieveit said:
I actually found his post to be more interesting than yours (i.e. your post quoted above, not the interesting one that started this thread). His post made me realize that Buchholz had been effective at a quality level for a lower percentage of the time than I had previously realized, and at least he introduced a new idea (that perhaps his numbers are skewed by one bad inning at the end more often than other pitchers). Not everyone needs to write a thesis.

It's not like his post was some kind of poorly worded gamethread drivel. Some of us also enjoy posts that simply express relevant new ideas from a fan's perspective.

I don't think this post got nearly enough adoration.

While I've loved baseball stats as long as I've loved baseball, I too often find the insistence upon them by some can sap the joie de vivre right out of any subject. I think we sometimes forget we are discussing a game here. While to many, including myself, it is almost more of a religion, the need to employ esoteric computations in every single post can do as much to degrade a thread as an inane comment. The scientists at CERN are more lighthearted about their work than we are about a game.

Reverend · Dec 28, 2014

maxotaur said:
I don't think this post got nearly enough adoration.

While I've loved baseball stats as long as I've loved baseball, I too often find the insistence upon them by some can sap the joie de vivre right out of any subject. I think we sometimes forget we are discussing a game here. While to many, including myself, it is almost more of a religion, the need to employ esoteric computations in every single post can do as much to degrade a thread as an inane comment. The scientists at CERN are more lighthearted about their work than we are about a game.

Claiming that it doesn't make sense to aggregate non-consecutive seasons of poor performance that occurred for different reasons isn't exactly ground breaking methodology.

Anyone who thinks the Red Sox look at players without considering inflection points and the reasons for performance is nuts.

Edit: I would like to emphasize that I think the notion that non-uber SABR insights are somehow stifled on the board is mistaken; there is tons of such and it should be evident and nobody is opposed to them. I'm just trying to speak to what I see as a false dichotomy--even before the statistical revolution, people still tried to discern causation and a failure to do so in projecting performance would be considered sloppy thinking.

I mean, there wasn't that much SABR in that first post anyway unless you hate the mere mention of FIP or ERA+. It was just a way of visualizing his success by pitch and location. It was mostly a creative presentation of data that people have been using for generations.

PrometheusWakefield · Dec 29, 2014

iayork said:
I agree with your main point, but I don't think it's really true that this is same level of variance as we see with Dempster, Jackson, or Santana; all of those guys are actually pretty consistent in their ERA+, with nowhere near the wild season-to-season swings that Buchholz has shown. As regulars, Buchholz's ERA+ has swung from 72 to 237, while Santana has ranged between 74 and 127, Jackson from 60-132, and Dempster, over a much longer career, varied from 63 to 154. And when looking at year-to-year variance the difference is even more spectacular, with none of them showing anything close to the 165 points of ERA+ change that Buchholz did.

Yeeeeah, but lets keep in mind that 237 was only really a half season. But yes, his 2013 jump was about the confluence of greatly improved luck with a genuine bump in real performance. It was just a lot more the luck than the performance, and 2014 was really right in line with career norms but for the luck.

The thing that most jumps out at me looking over Buchholz' history is the HR/FB rate. That's the one factor that really swings in the years that Buchholz looks like an ace (2010, 2013) and the years when he looks like nothing special (the rest).

phenweigh · Dec 29, 2014

PrometheusWakefield said:
Yeeeeah, but lets keep in mind that 237 was only really a half season. But yes, his 2013 jump was about the confluence of greatly improved luck with a genuine bump in real performance. It was just a lot more the luck than the performance, and 2014 was really right in line with career norms but for the luck.

The thing that most jumps out at me looking over Buchholz' history is the HR/FB rate. That's the one factor that really swings in the years that Buchholz looks like an ace (2010, 2013) and the years when he looks like nothing special (the rest).

This may be a semantics complaint, but this isn't the first time I've seen the variance in HR/FB rate referred to as luck. The reason xFIP normalizes FB/HR rate is because one season of data isn't a large enough sample to measure a pitcher's "true talent level". IMO. that is different than luck.

There is no Rev said:
Eh, the problem with FF's post is that he is effectively begging the question: iayork is wondering if we can figure out if there is hope for Buchholz based on our data, and FF posited the answer as "No."

What assuming "No" simply because he's older doesn't consider any sort of causal mechanism as to why we sometimes have gotten Good Buchholz and other times have gotten Evil Buchholz. He has been injured at times, yes, but they have not been similar injuries, so to the extent that injuries cause poor play, there's still the possibility we could turn the Good Buchholz card next year. Also, in 2012, he forgot how to through his frickin' change-up--and there's reason to believe that happened again this past season as well.

The point being that we might be able to learn something by considering what caused different outcomes at different times than just saying that we think he's bad too often.

You can kinda see a not fully thought out idea about a player in things like the claim that we've only gotten 1.5 very good seasons out of him--if you throw out the first six starts in 2012 when Chuckles had apparently forgotten how to throw his change, his ERA comes down to 3.62 for the rest of the season which is pretty good for 23 starts.

I'm not trying to be snarky, but isn't the parsing of larger samples into smaller samples and ascribing a reason for throwing out the bad part of the sample why EV was regularly pilloried? Or was it more how he did it?

Regardless, I think Clay's career ERA+ of 109 is a fine indicator of ability when he takes the mound. Combine that with his ~21 starts per season (I"m throwing out 2007 as a few late season starts doesn't represent typical durability) and that's who he is. Buchholz is an above average pitcher who shouldn't be expected to take the ~~more~~ mound more than 2/3 of the time.

P'tucket rhymes with... · Dec 29, 2014

phenweigh said:
This may be a semantics complaint, but this isn't the first time I've seen the variance in HR/FB rate referred to as luck. The reason xFIP normalizes FB/HR rate is because one season of data isn't a large enough sample to measure a pitcher's "true talent level". IMO. that is different than luck.

I'm not trying to be snarky, but isn't the parsing of larger samples into smaller samples and ascribing a reason for throwing out the bad part of the sample why EV was regularly pilloried? Or was it more how he did it?

Regardless, I think Clay's career ERA+ of 109 is a fine indicator of ability when he takes the mound. Combine that with his ~21 starts per season (I"m throwing out 2007 as a few late season starts doesn't represent typical durability) and that's who he is. Buchholz is an above average pitcher who shouldn't be expected to take the ~~more~~ mound more than 2/3 of the time.

Clay strikes me as a walking example of the pitfalls in using the Mean as a measure of central tendency; his season-to-season performances have been black or white, and to call him "gray" serves to hide rather than illuminate an essential element of his career.

Sprowl · Dec 29, 2014

One further element in Buchholz' unpredictability -- he isn't just Good Buchholz and Bad Buchholz. He changes his repertoire often, both season to season and game to game, and he changes action on pitches within his repertoire. The slider became a cutter in 2010, and went from 10% of his pitches to 24%. The curve comes and goes (it was almost abandoned in 2010). He changed his arm slot in 2008. His fastball lost velocity and gained movement. The splitter showed up in 2012 in small quantities to LHB. The changeup went from 20-25% of pitches in 2007-2009 to 10-12% in 2013-2014. 2015 could bring a new assembly of old parts, a Frankenstein's Buchholz.

joe dokes · Dec 30, 2014

PrometheusWakefield said:
There's no big mystery here. This is the normal variation of pitchers who are about the quality of Clay Buchholz. You look at other guys I'd put in his category like Ryan Dempster or Edwin Jackson or Ervin Santana. Some years the luck stats bounce their way and they look like solid #2s at least. Other years the luck stats go against them and people act like something must have happened to their talent. Buchholz isn't especially unusual, and unfortunately he isn't especially talented, just another decent pitcher who looks good when things break his way and bad when they don't.

Dempster and Jackson have the added attraction of durability.

Reverend · Dec 30, 2014

phenweigh said:
I'm not trying to be snarky, but isn't the parsing of larger samples into smaller samples and ascribing a reason for throwing out the bad part of the sample why EV was regularly pilloried? Or was it more how he did it?

Thank you for not being snarky because there is a good answer to your question, not only for understanding Buchholz (if such a thing is possible) but for understanding how to analyze data, so if there are questions on the matter, it is good for them to be asked.

Building on iayork's explanation of how it depends upon what you are trying to do with the data, there are reasons to break up a sample and there are right and wrong ways of doing it. The problem of cherry picking is when you just pop around grabbing what supports your sample--sweet data--and leaving behind the rest of the sour data that does not support your theory.

Note, though, that I chose the first six games. That could be cherry picking, but we at least know there is a sense of "before and after"--it's not like I chose to drop all of April, half of July, a week in August, and three innings on August 17th or something. More specifically, we have reason to take a before and after snapshot at that point because we have reason to believe, from Buchholz himself, that he altered his pitching technique at that time. We had been tracking what the hell was wrong with him here, and had identified changes in his pitch performance between other seasons. (This speaks to Sprowl's note that Buchholz changes his composition a great deal from year to year.)

Imagine if you knew a pitcher had implemented a new pitch into his repertoire on, say, July 12th. It would make sense to look at his performance before then and compare it to his performance after then, yes? Sure, the sample sizes would be relatively small, but it's all you'd have so it's what you'd use--you would just adjust in your mind how confident you were that the sample was predictive, as iayork indicated.

My point was that we have reasons both in the data (looking back over it, that thread was pretty awesome--too bad so many of the pics are gone) and in his public statements that something changed at that point, so I was pointing out that across his 23 starts in 2012 after the adjustment was made, Chuckles was a pretty good pitcher.

So this then all speaks to trying to identify the causal factors behind why he sometimes is successful and other times is not. If you think he's been bad because of freak injuries and correctable, um, memory loss, then you might think he can be very good. If you think he's bad because he's injury prone and dumber than a pile of rocks, then you might not, which is why I agreed with iayork that the poster who just said he's been note very good more often than not and now he's 30 was not terribly interesting with respect to figuring out what we think we know about Mr. Buchholz.

gryoung · Dec 30, 2014

He's always been an enigma - a guy with terrific stuff but not much upstairs who seems to get rattled when there's a runner on first (you know, the multiple throws over to the base). I'd be interested to see his history of pitching with first base occupied and with no runners on.

joe dokes · Dec 30, 2014

He's always been an enigma - a guy with terrific stuff but not much upstairs who seems to get rattled when there's a runner on first (you know, the multiple throws over to the base). I'd be interested to see his history of pitching with first base occupied and with no runners on.

www.icangooglethatforyou.com

http://www.baseball-reference.com/players/split.cgi?id=buchhcl01&year=Career&t=p#bases

Reverend · Dec 30, 2014

gryoung said:
He's always been an enigma - a guy with terrific stuff but not much upstairs who seems to get rattled when there's a runner on first (you know, the multiple throws over to the base). I'd be interested to see his history of pitching with first base occupied and with no runners on.

[tablegrid= Buchholz: Runner on 1st (only) ]Year ERA IP H R ER HR BB SO AVG WHIP 2010 0.98 36.2 30 5 4 2 8 19 .248 1.04 2011 1.06 17.0 18 2 2 1 3 11 .290 1.24 2012 1.82 34.2 37 8 7 2 10 20 .303 1.36 2013 0.87 31.0 9 3 3 1 9 20 .107 0.58 2014 2.70 36.2 36 11 11 3 8 26 .273 1.20 [/tablegrid]

phenweigh · Dec 30, 2014

iayork said:
Anyway, point being, small sample sizes are usually bad, but not hopeless.

There is no Rev said:
My point was that we have reasons both in the data (looking back over it, that thread was pretty awesome--too bad so many of the pics are gone) and in his public statements that something changed at that point, so I was pointing out that across his 23 starts in 2012 after the adjustment was made, Chuckles was a pretty good pitcher.

Sprowl said:
One further element in Buchholz' unpredictability -- he isn't just Good Buchholz and Bad Buchholz. He changes his repertoire often, both season to season and game to game, and he changes action on pitches within his repertoire. The slider became a cutter in 2010, and went from 10% of his pitches to 24%. The curve comes and goes (it was almost abandoned in 2010). He changed his arm slot in 2008. His fastball lost velocity and gained movement. The splitter showed up in 2012 in small quantities to LHB. The changeup went from 20-25% of pitches in 2007-2009 to 10-12% in 2013-2014. 2015 could bring a new assembly of old parts, a Frankenstein's Buchholz.

P'tucket said:
Clay strikes me as a walking example of the pitfalls in using the Mean as a measure of central tendency; his season-to-season performances have been black or white, and to call him "gray" serves to hide rather than illuminate an essential element of his career.

Thank you iayork and There is No Rev for responding directly to my question ... much appreciated.

I think Sprowl's post gets to the heart of my opinion and why I disagree with P'tucket. Based on watching and reading about MLB, knowledgeable commentators and writers consistent remark that baseball is a game of adjustments. Clay may be an extreme example, or maybe he's simply more open about it. Regardless, the constancy of his changes is why dropping a six-game sample doesn't seem like the right thing to do. It's not because there isn't a documented reason that it's a transition point, it's because it is one of many transition points. Clay often makes adjustments so it's reasonable to expect he'll do so again in 2015. Some may work well (for a time being), while others may be disastrous. In other words, IMO the best prediction for Clay's 2015 season are his career averages. Granted, the error bar on that prediction is probably larger than most pitchers (and maybe that's what P'tucket's point is) but I don't see any reason to shift the mean.

Reverend · Dec 30, 2014

phenweigh said:
Thank you iayork and There is No Rev for responding directly to my question ... much appreciated.

I think Sprowl's post gets to the heart of my opinion and why I disagree with P'tucket. Based on watching and reading about MLB, knowledgeable commentators and writers consistent remark that baseball is a game of adjustments. Clay may be an extreme example, or maybe he's simply more open about it. Regardless, the constancy of his changes is why dropping a six-game sample doesn't seem like the right thing to do. It's not because there isn't a documented reason that it's a transition point, it's because it is one of many transition points. Clay often makes adjustments so it's reasonable to expect he'll do so again in 2015. Some may work well (for a time being), while others may be disastrous. In other words, IMO the best prediction for Clay's 2015 season are his career averages. Granted, the error bar on that prediction is probably larger than most pitchers (and maybe that's what P'tucket's point is) but I don't see any reason to shift the mean.

Well, this is part of where it's frustrating to be a fan in that we have imperfect information. The teams, of course, should theoretically have more information on inflection points with regard to injuries, intentional changes in delivery and maybe even some unintentional changes in delivery if they pour over the data and film, changes in pitch calling and so forth. So it's likely that the teams parse up the data set more than we can reasonably do. This is one of the reasons, as Farrell has pointed out, that a team's own players are always at least a bit more valuable to that team because they have a more complete understanding of the player's health.

So the question becomes: If we do think we know about an inflection point, do we use it or do we discard it because we don't know all of them?

I can see arguments for both sides. If we're trying to figure out how a player will do going forward, I'm of the school of thought that it makes sense to look at what it tells us in terms of possible causal elements for his future success because then we can think about whether or not certain causal factors are likely to be replicated in the future or not.

phenweigh · Dec 31, 2014

There is no Rev said:
Well, this is part of where it's frustrating to be a fan in that we have imperfect information. The teams, of course, should theoretically have more information on inflection points with regard to injuries, intentional changes in delivery and maybe even some unintentional changes in delivery if they pour over the data and film, changes in pitch calling and so forth. So it's likely that the teams parse up the data set more than we can reasonably do. This is one of the reasons, as Farrell has pointed out, that a team's own players are always at least a bit more valuable to that team because they have a more complete understanding of the player's health.

So the question becomes: If we do think we know about an inflection point, do we use it or do we discard it because we don't know all of them?

I can see arguments for both sides. If we're trying to figure out how a player will do going forward, I'm of the school of thought that it makes sense to look at what it tells us in terms of possible causal elements for his future success because then we can think about whether or not certain causal factors are likely to be replicated in the future or not.

I agree with the bolded sentence. I just don't think splitting Clay's 2012 season was a good way of making an argument that Clay had more than 1.5 good seasons. In addition to his very good 2010 (ERA+ 187) and his spectacular 2013 half-season (ERA+ 237), he had two good half-seasons in 2009 and 2011 (ERA+ of 111 and 124). Maybe the issue with the poster who felt Clay only had 1.5 good seasons is an issue with setting the bar too high for what is good. To me, he's had 2.5 seasons of at least good.

Even taking 2012 in it's entirety, though below average, wasn't terrible (ERA+ 92). Adding in your 2012 inflection point, one could argue that he's had a total of 3 good seasons, but it's not really needed to show that only 1.5 seasons of good Clay doesn't fit reality.

Sampo Gida · Jan 19, 2015

To me the biggest question I have about Buchholz is his health. This is a guy with almost 1000 MLB innings under his belt who dropped 165 ERA+ points in a year, so it has to be more than not making adjustments IMO.

Since suffering that shoulder injury/neck strain, he has pitched as poorly as the 2011 Lackey, which was pretty awful (although his few starts in Sep 2013 were fine he struggled his final 2 starts in the post season). Lackey of course required TJ Surgery, and Buchholz is said to be hea/thy, but it would not surprise me to find out somewhere down the road he has had an issue physically. Sometimes injuries only manifest as poor command in the zone, and according to Brooks Baseball he established career highs with grooved fastballs in 2014

Another possibility is he dropped some weight coming into last season for reasons which escape me which may have affected his overall strength and conditioning. This would be the best case scenario since it would be something he could correct this offseason

Fireball Fred · Jan 19, 2015

I think I'm the poster who first said, explicitly, that Buchholz had had only 1.5 good seasons. Here's the reminder I'll offer: Buchholz has never quite pitched a full season. He did get to 189 IP once, and 173 (working from memory here) another, but a number of his calendar years are fractions of seasons. As I think I said, sometimes he's good, sometimes he's bad, sometimes he's just not there. He has an established track record of unreliability, although at his best he's certainly very good.

Dan to Theo to Ben · Mar 29, 2015

projected Opening Day starter
http://www.boston.com/sports/baseball/redsox/2015/03/29/red-sox-name-clay-buchholz-opening-day-starter/tCwjeCjhCgMYCLpLOgyQPP/story.html

HriniakPosterChild · Mar 31, 2015

Dan to Theo to Ben said:
projected Opening Day starter
http://www.boston.com/sports/baseball/redsox/2015/03/29/red-sox-name-clay-buchholz-opening-day-starter/tCwjeCjhCgMYCLpLOgyQPP/story.html

He's the Ace.

(Says so on the T shirt, folks.)

Rovin Romine · Apr 1, 2015

How often do the "aces" of each club line up across the season? Assuming the opening day starter matches up with the other teams' number one starter at a higher rate, this is a good choice. Buchholz at his best can hang in there with anyone. At his worst he's going to lose no matter who he's paired up against. If possible, I'd rather see his sterling efforts going against strong competition.

If there's no "matching" of starters that's significant, then it really does not matter.

mt8thsw9th · Apr 1, 2015

HriniakPosterChild said:
He's the Ace.

(Says so on the T shirt, folks.)

Ace? As in Ace Ventura? I'd say he looks more like Fire Marshal Bill.

Toe Nash · Apr 1, 2015

Rovin Romine said:
How often do the "aces" of each club line up across the season? Assuming the opening day starter matches up with the other teams' number one starter at a higher rate, this is a good choice. Buchholz at his best can hang in there with anyone. At his worst he's going to lose no matter who he's paired up against. If possible, I'd rather see his sterling efforts going against strong competition.

If there's no "matching" of starters that's significant, then it really does not matter.

It gets messed up pretty much immediately with off days, rainouts and of course, injuries.

He was named the opening day starter because he has been on the team the longest. If it motivates him, good, but otherwise it's pretty meaningless.

TheYaz67 · Apr 1, 2015

Toe Nash said:
It gets messed up pretty much immediately with off days, rainouts and of course, injuries.

Clay agrees with you...

"It's fun to be out there and be called a No. 1 or whatever but my point of view, No. 1 is the guy who starts the first game of the season because that's the hype, but after that, I've been No. 5, been No. 3, seems like I'm always facing the aces of other staffs after that first, second week of the season because of off-days."
Buchholz said he's healthy, feels good after throwing about four bullpen sessions here.

ToeKneeArmAss · Apr 1, 2015

That's an admirably self-aware quote. Speaking from experience, some of us mature late. Hopefully Buchholz keeps his head screwed on right this season and has results that match.

joe dokes · Apr 2, 2015

The only tangible difference is that the "#1" & "#2" starters, in theory, will get 1 more start than the remaining 3. (33 to 32). But that's in theory only, as rotations rarely hold together like that over a full season.

semsox · Apr 2, 2015

ToeKneeArmAss said:
That's an admirably self-aware quote. Speaking from experience, some of us mature late. Hopefully Buchholz keeps his head screwed on right this season and has results that match.

Speaking of self-aware quotes from Buchholz, Bradford did a nice piece on him yesterday that touched on a little bit of everything. Buchholz is quoted throughout and I felt comes across very well. Worth a read:

http://www.weei.com//sports/boston/baseball/red-sox/john-tomase/2015/03/31/making-clay-it-has-taken-eight-years-buchholz-

touchstone033 · Apr 2, 2015

What? This has been up 24 hours, and no one's posted it here??

Curt Schilling knows a No. 1 starter when he sees one, and the outspoken ex-Red Sox ace doesn't think Clay Buchholz has the makeup to be a No. 1 starter.

"Well, I don't think he wants to be one," Schilling said Wednesday in a conference call to promote ESPN's Sunday Night Baseball. "I think there's a level of commitment mentally and physically you have to have. You have to have a little bit of a dark side, I think, in the sense that losing has to hurt so bad that you do whatever you can do to make sure it never happens again. Clay is just kind of, 'Hey, I'm going to pitch today.'"

joe dokes · Apr 2, 2015

touchstone033 said:
What? This has been up 24 hours, and no one's posted it here??

"Well, I don't think he wants to be one," Schilling said Wednesday in a conference call to promote ESPN's Sunday Night Baseball. "I think there's a level of commitment mentally and physically you have to have. You have to have a little bit of a dark side, I think, in the sense that losing has to hurt so bad that you do whatever you can do to make sure it never happens again. Clay is just kind of, 'Hey, I'm going to pitch today.'"

Right. Because #1 starters "want it" more.

kieckeredinthehead · Apr 2, 2015

touchstone033 said:
What? This has been up 24 hours, and no one's posted it here??

"Well, I don't think he wants to be one," Schilling said Wednesday in a conference call to promote ESPN's Sunday Night Baseball. "I think there's a level of commitment mentally and physically you have to have. You have to have a little bit of a dark side, I think, in the sense that losing has to hurt so bad that you do whatever you can do to make sure it never happens again. Clay is just kind of, 'Hey, I'm going to pitch today.'"

This is exactly why Manny could never be a middle-of-the-lineup bat.

Leather · Apr 2, 2015

Tim Lincecum: never a TRUE #1 starter.

Not a big enough asshole.

Ditto for Barry Zito.

soxhop411 · Apr 6, 2015

Peter Gammons ‏@pgammo 9s9 seconds ago
Clay Buchholz combined spring training and opener:31 strikeouts, 5 walks, 29 IP

Soxfan in Fla · Apr 6, 2015

I'd like to have this version of Clay project throughout season. If this is the version it is a huge benefit for the Sox.

Fireball Fred · Apr 6, 2015

I've been a Buchholz skeptic, in terms of reliability and application, and I thought Schilling's recent comments seemed right. But maybe he's finally matured and is rising to his newly-defined responsibility. That, too, seems plausible. The Phillies lineup is admittedly weak, but Buchholz was awfully good today.

soxhop411 · Apr 6, 2015

Jnai · Apr 6, 2015

Careful with the FX data from that game, it's all screwy. Need to wait for them to reparse it. Happens sometimes on opening day.

soxhop411 · Apr 6, 2015

Jeremy Lundblad ‏@JLundbladESPN 7s7 seconds ago
Phillies went 0-12 on AB ending in Clay Buchholz's off-speed stuff (1st time he's done that since 2012)

The unprojectable Clay Buchholz

Member

Member

New Member

New Member

Member

Member

Member

Member

Member

Member

Member

fanboy

for king and country

Member

for king and country

Member

Member

Member

Member

mikey lowell of the sandbox

Member

Member

for king and country

Member

Member

for king and country

Member

for king and country

Member

Member

New Member

Banned

Member

Johnny Rico

anti-SoSHal

Member

Member

Paul Byrd's pitching coach

Member

Member

New Member

Member

Member

given himself a skunk spot

news aggravator

Member

New Member

news aggravator

is not worried about sex with goats

news aggravator