Javier Vazquez: More on bunching and damage

For those following along, I dedicated last week to Braves SP Javier Vazquez. I first stated my belief that he’ll be a top five fantasy pitcher for the rest of 2009 and then briefly explored the claim that he bunches his hits and walks together. I found little evidence to suggest he did, but there were some things I didn’t get a chance to look at. Today, I’d like to go a couple steps further and look at some of these things.

First, while I found that Vazquez was merely league average-ish at bunching hits and walks, I didn’t check how similarly-skilled pitchers performed. I also didn’t check the quality of the hits, treating every hit and walk equally. Unfortunately, I ran out of time today and didn’t get to look at a few other things I would have liked to, so I’m sure you’ll all be happy to hear that there should be yet another follow-up in the coming days.

Also, please note that, because we’re digging into somewhat complicated matters, this may get a little technical for some readers’ likings, and the charts certainly aren’t as straight-forward as many of you would like. Please don’t feel overwhelmed. I’ll do my best to summarize, in simple terms, what’s going on at the end of each section.

Note: All data presented in this article was arrived at using the stupendous Retrosheet for the years 2004 to 2008.

Comparison to his peers

As I noted in my previous article (and as a few commenters also made note of), it would be best to compare Vazquez not only to league average, but also to similar skilled pitchers (henceforth known as ‘peers’).

To define “peers,” I selected all starting pitchers who were within 0.25 LIPS ERA points (to assure that they were exhibiting similar skills to Vazquez and not getting lucky) and within 0.05 WHIP points (to assure that there weren’t differences in the overall number of hits and walks allowed that would skew the study) for each year. Arbitrary, yes, but that’s kind of the nature of the beast. This gives us a sample of nearly 11,000 inning appearances from 2004-2008. The results are shown below:

image

The format of this chart is a little different than last time. Each column shows the percentage of time that this exact number of hits and walks were allowed in an inning (as opposed to the percentage of time that at least this many hits and walks were allowed, as was displayed last time. This was done to make for easier comparisons to the next couple charts).

The important thing to take away from this is that Vazquez’s peers don’t perform much differently than league average and that Vazquez doesn’t perform much differently than them. In fact, we see almost the same exact net result: the bunching of 1.6 fewer hits and walks than his peers per 216 inning appearances (his average number pitched since 2004). If we remove the innings with two hits and walks, it drops to the same 5.4 deficit we saw last time as well.

Let’s dig a little deeper…

Damage done by hits and walks

In my first study, I didn’t include the actual damage done by the hits and walks, but simply looked at the raw totals. It was suggested that perhaps Vazquez’s problem isn’t how many hits and walks he bunches together, but the types of hits (i.e maybe more doubles and homers than singles and walks). Using Linear Weights, we can check this pretty easily.

If you’re the kind of person who’s interested in the specifics, you can click here to see the average damage done per single inning of a particular type. Here, it appears that more damage is done to Vazquez than both league average and his peers in innings with two, three, or four hits and walks, but he has been able to make up for it a bit by bettering (or tying) both the league and his peers in innings with 5 through 11 hits and walks.

After finding this, I combined the frequency with which Vazquez allows each type of inning with the cumulative damage done by the walks and hits in that type of inning (scaled to 216 inning appearances). You can see the breakdown by inning type here. Vazquez seems to take the biggest (relative) beating in innings with four hits and walks, and these innings happen frequently enough to wreak a little havoc.

For those who would rather not be bored with the specifics (the majority of you, I’m wagering), below is a chart with (hopefully) an easily understandable version of the final effects. This takes into account both the frequency with which Vazquez allows each type of inning and the cumulative damage done by the walks and hits. It has been scaled to show the net linear weighted effect per 216 inning appearances (Vazquez’s average since 2004). I’ve also broken these effects up by types of innings: those with at least 2 hits and walks, 3 hits and walks, and 4 hits and walks.

image

To put it into simple terms, what we’re seeing is that Vazquez’s peers are a bit better than league average, but Vazquez himself is a bit worse than both. At best, he’s about 2.2 runs worse if we only focus on 2+ H/BB innings (about 0.08 points of ERA). At worst, he’s about 4.2 runs worse if we only focus on 3+ H/BB innings (about 0.17 points of ERA). While some of this may be noise, it still looks like it might be justifiable to dock Vazquez’s value a little bit… just don’t go crazy. I still believe that an ERA below 3.30 is a very real possibility for Javy.

Concluding thoughts

As always, comments are welcome. As I mentioned earlier, I’ll probably be doing one final follow-up in the coming days.

A Hardball Times Update
Goodbye for now.

15 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Mark
14 years ago

Mike—I may be taking this in a direction you did not intend, but your comment reminded me of something I’ve been thinking about.

For those who understand how PitchFX works, what are the odds that we’ll see pitches categorized as “thrown from the windup” and “thrown out of the stretch” anytime soon? Obviously, you’d need someone keeping track of that information, but it would seem to be one of the next bits of information that may answer a lot of questions. I suppose a shortcut would be to have Pitchfx categorize pitches by “thrown with bases empty” and “thrown with runners on”, but that wouldn’t be quite as accurate unless you knew for certain that a given pitcher was 100% consistent about his windup and stretch usage.

To bring my question back around—I’d be interested to see how Javy’s control and movement suffer from the stretch compared to his windup.

John Ziccardi
14 years ago

The guy’s a sub 500 pitcher lifetime, 131-134 end of 2008 with a 4.29 lifetime ERA for 12 ML seasons. In spite of playing for contenders he’s had losing seasons three of the last four years. Ozzie Guillen dumped his fellow hispanic and went public, saying he’s just not a big game type pitcher after Vazquez bombed out late last year. It’s not about his stats or about PitchFX and other hyperstatistical over-analysis, it’s about something in him Ozzie realized. Great arm questionable makeup. Never come close to winning 20. Teams have overpaid and hoped his stuff would overcome the other stuff which in 12 years has not worked out. 131-134 and 4.29, at best mediocre career numbers.

Dave
14 years ago

John- Nobody disputes those things.  The goal of this analysis is to see whether that manifests itself in his statistics in some subtle way.  Ideally we’d like to be able to either pick out the Vazquez’s of the baseball world and devalue them or say definitively that he’s just been unlucky and we can expect his performance to improve.

Derek- Nice analysis once again.  I’d check out next how this peer group compares to Vazquez with respect to pitching with runners on base.  The stuff above is fine, but I think we should try and get to the cause of the problem which may be that Vazquez gets rattled in these bad innings you detect.  I’m guessing that on average pitchers are worse with runners on?  But maybe Vazquez’s decrease is worse than average by a statistically significant amount.  If so that could give us a hint to include such numbers in pitchers projections, which wouldn’t be too hard.

Mike Podhorzer
14 years ago

This is great and all, but I’m still curious as to WHY he is worse in that last chart. Is there any way to figure this out, or the furthest we could go is identifying that yes there is a problem, but we still don’t know the underlying cause?

Mark
14 years ago

Isn’t the essence of sabermetrics (at least according to Bill James) the process of re-examining “settled” issues? In this case, traditional baseball wisdom says that Javy’s just a headcase and isn’t a big-game pitcher. But while that has to be considered, it’s not necessarily conclusive, and even if it is, it doesn’t really explain much.

Firstly, saying that Ozzie is being candid about Javy and that all his teams felt the same way is simply an appeal to authority. That’s fine in an informal debate, but it’s not a particularly compelling or persuasive argument. To be quite blunt, I respect Ozzie Guillen’s opinion, but I also know that he keeps running Scott Podsednik out there, so my respect is tempered a bit.

Secondly, even if I grant that Ozzie’s right and Javy is just a headcase, what does that explain? I want to know what being a “headcase” actually means in baseball terms. Does a headcase give up big innings more frequently (what we’re looking at here)? Does a headcase give up extra-base hits more frequently? What is actually happening, pitch by pitch, that makes Javy a “headcase”? Moreover, Javy has had some excellent seasons—was he doing something different in those years that minimized his headcasery?

To me, that’s what I’m trying to figure out—what is actually happening out there, and how can it be used to predict future performance. It’s not a case of someone looking at Javy and saying, “Gosh, he looks great”, and then Javy not performing to expectations. It’s the fact that some of Javy’s statistics indicate very-good if not greatness, and he’s put together some very good years, but he’s also had some underwhelming years…what explains that?

John Ziccardi
14 years ago

I appreciate the academic and intellectual quality of trying to examine and predict based on statistical analysis. Remember, Javy Vazquez has a 12 year career record of mediocre performance despite his very good stuff. It’s not the comments that one draws conclusions about him from, but rather his high ERA and losing lifetime record. Pure numbers.

And it’s not just Ozzie Guillen and the White Sox who have soured on Vazquez, the contending D’backs and Yankees did also. Ozzie, whose nasty tongue I don’t condone, stated it publicly and bluntly. But Vazquez has been bounced off good teams like the home runs he allowed as a Yankee.  Many many pitchers with less stuff have fared far better. The baseball people know something here.

Baseball history is loaded with guys who had tons of talent but remained mediocre or inconsistent. Javy’s still active and thus a good candidate for study but again, it’s not like he’s Phil Hughes or another young phenom. This guy overall at 131-134 has been mediocre and inconsistent for 12 ML seasons with basically winning teams. Never come close to winning 20.

It looks like someone wants to be a genius and figure him out using metrics and then predict that he will come in with an ERA of under 3.30? That seems an ego-based projection in the 99.9th percentile of improbability, and a huge deviation from standard. Very statistically improbable.

I don’t know if he’s a headcase, but one suspects the problem is somewhere in his approach to pitching. You don’t need Sabermetrics to see that. I’m not convinced one can “fix” him or find his issue using Sabermetrics. But I’ll follow this one.

Derek Carty
14 years ago

Thanks, Dave and Mark.  You guys really hit on the purpose of this.  I’m not really interested in the opinions of Ozzie Guillen or even how fans perceive personnel moves, which aren’t 100% about the talent involved.  And even at that, while there are many smart people working for and running MLB teams, it is incorrect to assume that these people are infallible or that all people working for and running MLB teams are of the most intelligent breed.  There are a good number of writers on the internet (not to imply that I am a part of this group) who I’d wager would do a better job running a baseball team than a handful of GMs (of course given all the resources of a GM).

As far as his record goes, this matters very little.  Wins are a fickle thing, more a matter of the quality of teams you play on (Vaz played more than 1/2 of his career in Montreal) and even at that, are prone to much random variation.  Furthermore, his 12 bad seasons really aren’t 12 bad seasons.  I’d give you 5 if you really wanted to argue it (2000, 2004, 2005, 2006, 2008).  In the rest he either posted a good ERA or didn’t display the top notch peripherals he is displaying now.  Actually, my argument is that, this year, he’ll post the best peripherals of his career.  If you’ve read the CAPS article, you’ll see that a good portion of my argument is based on the fact that Vazquez has 1) faced a set of circumstances within the past few seasons that are completely and undeniably out of his control and 2) has entered a new, more favorable environment in 2009.  He *will* be good in 2009 – sub 4.00 ERA good (his peripherals should improve enough to withstand some bad luck), unless he receives ridiculously bad luck or unless he actually is doing something that isn’t captured in all the peripherals and underlying numbers, but what I’ve found so far doesn’t seem to indicate that.  What I’m doing is trying to figure out how good he’ll be as objectively as I can.

While it’s easy to simply look at a player’s surface record without considering context or the opinions of a manager with a mediocre reputation, at best, and make judgments, I’m much more interested in examining everything and seeing exactly what’s happening.

Hopefully that makes sense.

NadavT
14 years ago

Derek –

I wonder how the numbers would change if you only looked at 2006 and 2008, rather than all of 2004-2008.  I know that your stated purpose is to see what Vazquez’s overall pitching patterns look like, but if you want to focus on the specific question of exactly why he appears to be a consistently unlucky pitcher, it would make sense to focus just on those years when he appears to be unlucky.  Going just on FIP-ERA, you get:

2004: -0.08
2005: -0.46
2006: -0.90
2007: +0.18
2008: -0.86
2009: -0.91 (so far)

I could see a case for including 2005 as well, but when you have 2004 and 2007 in there, I think you might not have the best sample for exploring what kind of “bad luck” Vazquez has when he is, in fact, unlucky.

John Ziccardi
14 years ago

First of all no one said he has had 12 bad seasons. Don’t misquote or paraphrase or misstate what has been said to serve your own purposes. That’s intellectually dishonest. He has a 12 year cumulative record of 131-134 and a 4.29 ERA that adds up to a mediocre career. In the end players are judged by their career stats not by one or a few seasons.

As for your stated willingness to wager that there are some writers on the internet who could do a better job than some GM’s, the comment is preposterous. It can never happen in the real world, thus it comes off as a self-serving egotistical assertion. It seems designed to buttress your case with readers and ally yourself with them, without the possibility of ever being disproven via any means including statistical scrutiny. It’s not the comment of someone who says he wants an objective result and purportedly relies on science and mathematics.

The truth is you don’t know how your wager would play out and neither will anyone else. Ever.

As for the prediction that he’ll be good in 2009, sub 4.00 ERA, I’m mindful that he’s off to a good start already in the new, more favorable environment of the National League with no DH. 

But even if he does have a good season the jury will be out on him for several years or until his career ends. It’s consistency over a long period at a high level that separates decent or good players from stars.

And while Vazquez’ career is far from over, 12 ML years is a pretty long track record. I’d like to see someone outside the game be credited with turning a career C player into a B+ or better for several years running. I’d find that statistically significant at the .01 level. Probably never been done before, even by ML personnel with all the resources at their disposal.

And if you could somehow convince a GM into giving his job to an internet writer of your choice, you could prove or disprove what I would call your fallacious pseudo-bet. That’d put you 5 SD’s above the mean.

Mark
14 years ago

Can’t say I understand the adversarial tone. Weren’t we talking about baseball? I have plenty of thoughts about Javy Vazquez, but I can’t say that I’m emotionally invested in what other people think of him as a major league pitcher or as a person.

Out of curiosity:

http://www.thebaseballcube.com/players/Z/John-Ziccardi.shtml

One and the same? If so, nice batting eye.

Derek Carty
14 years ago

John, sorry if I came across as confrontational.  I’m not trying to start a fight, and I certainly didn’t mean to misquote you.  This is the line I was referring to (“This guy overall at 131-134 has been mediocre and inconsistent for 12 ML seasons”), but it sounds like I interpreted it wrong.

This will be my last response as it’s clear we have different viewpoints, but I just wanted to clarify that I’m coming at this from a fantasy baseball perspective.  I want to know what Vazquez will do *now*, not what he’s done for his career.  Yes, at the end of it that’s what he’ll be evaluated by, but there’s no reason 2009 can’t be his best season in a while if all the data supports it (which to this point it has come pretty close).  And that’s what fantasy players care about.

As to the off-hand GM comment, I’m not sure why this has become a big deal, but I don’t think it’s entirely prudent to assume that the 30 GMs are the 30 most intelligent baseball people in the world.  As I said before, many are very, very intelligent guys, but I also believe that someone like Tom Tango would do a very good job running a team.  That’s all I’ll say on the matter because, as you noted, it’s really a moot point.  And the comment wasn’t meant to spark a debate on this particular subject; I simply wanted to convey that work published on the internet is capable of having merit.

John Ziccardi
14 years ago

With a career ERA of 4.29 and a 131-134 record him just being unlucky is not a factor. He’s been dumped by several contending teams in a 12 year career. Ozzie Guillen said it most candidly and said it best last year. The guy’s not a big game type pitcher and he’s had 12 years to prove it. It’s not like he’s at the stage of his career Phil Hughes is at.

I don’t think the goal of this analysis is what you allege in regard to Javier Vazquez. Someone wants to be able to tell those in the big leagues that he has discovered something that the ML clubs didn’t detect.

That doesn’t mean a statistical methodology isn’t or can’t be valid for many or most other players. But Vazquez? Major League GM’s and managers have a handle on the why of his sub-500 12 year career despite his excellent stuff. Regardless of the attempts to analyze using complex stats, I think in this case it’s clear: the baseball people know.

And FYI I have a strong background in – and believe in – inferential statistics. But performance over the long term at the major league level is what counts for a ballplayer.
And the correlation between Javy Vazquez’ stuff and his performance is low.

Dave
14 years ago

OK well I can’t speak for Derek, but my interest in this series is to see if he can find something in Vazquez’s stat line that explains why he has stunk and would be a good predictor of him stinking in the future.  I have no interest in scouring the papers for comments from Ozzie Guillen to update my player projections.

Derek Carty
14 years ago

Good to hear, John.  Thanks for reading and commenting.

John Ziccardi
14 years ago

That is the nicest and most sincere reply comment one could expect. I must say I don’t do anything at all with fantasy baseball, I’m just a long-time baseball (and Yankee) fan who appreciates the statistical and numerical performance data baseball generates. It is clear that your statistical approach is designed to help you draft or pick fantasy players.

If your stats are as good as your writing, there just might be hope for Javy Vazquez’ numbers to improve from here out. He does have the stuff.

Thanks again. I recently discovered hardball times and will look for your postings and work on the site.