On WAR and the duties of the analyst

One of the biggest ironies of the new stat age is that the development of sophisticated and nuanced analytical tools like WAR provide the reader with a shortcut and enable the same type of lazy, simplistic analysis the tools were created to avoid. One doesn’t need to travel very far to find instances of this sort: see the use of single-season fWAR to settle debates on All-Star selections/snubs or MVP ballots. For the uninitiated, this is “doing it wrong”. WAR is comprised of many components: baserunning, fielding, and offense, among others. When it comes to fielding, a large sample of data is required in order to ensure reliability. In fact, many say that 3 seasons of UZR data is a good sample size. But single-season fWAR considers only the UZR data in that given year. This doesn’t mean that single-season fWAR is useless, just that some caution and editorial discretion is required in its application.

This isn’t the fault of WAR’s framework, although one can be forgiven for wishing there was some sort of warning sign attached to it on the Fangraphs’ leaderboards with a blinking light and a flashing message: “BEWARE! Small sample sizes still apply! Especially with the defensive component!”. Rather, the misuse of the framework is more user error than anything else. Drivers are responsible for knowing how to properly operate a car; analysts are responsible for knowing how to use WAR. They’re also responsible for not intentionally misuing WAR, or any other stat, to serve a preexisting agenda.

This gets us to a simple point, which is this: it’s the duty of the analyst to use the tools and frameworks wisely, with humility and honesty, and to create a margin of space which allows for tolerance and ambiguity. This is decidedly antithetical to the approach found so often in many popular forums: assert a controversial opinion, get pageviews, profit. But it’s a better approach. It’s not easy, and it requires far more work than making a few clicks on Fangraphs and spouting off an opinion on “Who’s better this year”, but it’s the way to circumvent the dogmatism and unsophisticated analysis we find so distasteful when we see it anywhere else.

One of the most hair-raising parts of George Orwell’s book Animal Farm is when the animals look through the window and find that the faces of the pigs have become indistinguishable from the faces of the humans that they all worked together to overthrow. The symbolism is unmistakeable: once they achieved their goals they became what they hated. The message is of course a political one, but it has bearing in the world of baseball analysis. This movement – call it a SABR movement, a stat movement, a mouth-breathing basement-dwelling movement, whatever you like – is only gathering more and more steam. WAR is on Baseball Tonight. David Cone broadcasts the virtues of FIP to the entire YES Network audience. It’s only getting bigger and stronger.

As the movement expands it will become easier to develop a more rigid orthodoxy. This isn’t necessarily bad. In a religious sense orthodoxy maintains the purity of a belief system, prevents false doctrines from gaining root amongst believers, and roots out heretics. In the world of baseball analysis it is far less coherent, systematic or discursive. But there’s still orthodoxy. There’s still a set of rules, however loose, analysts are playing by. There’s nothing wrong with this per se, but the risk is that orthodoxy can turn into dogmatism, which will stifle the innovative and free-thinking spirit which animated the movement in the beginning. Then the movement will stop growing, and it will be dead and boring. Consider this a call to keep that spirit alive, to keep hustling and thinking outside the box, to not use single-season WAR in an irresponsible way and to be ready to set aside WAR and any other metric, state or framework as inferior when the next innovation comes along.

Kevin Goldstein’s Midseason Top 50 Prospects

Kevin Goldstein posted his midseason top 50 prospects list today (subs. req’d), and he has Jesus Montero ranked as the seventh best prospect in the game. “He has not yet put up big numbers this year,” said KG, “there is clearly a frustration factor as he has nowhere to go in New York. At some point, the Yankees just have to trade him and accept the fact that he’ll rake elsewhere.” Montero was third overall in his preseason list.

The Yankees placed three others on the list. Manny Banuelos ranked 14th (“remains a lefty with two excellent pitches in his fastball and changeup … poised for a big second half), Dellin Betances ranked 24th (“whispers about him possibly being better off as a late-inning reliever are becoming more common these days”), and Gary Sanchez ranked 39th (“shown impressive power for an 18-year-old … scouts [still] project him as an adequate defender”). The Rangers are the only other team with four top 40 prospects, the Royals the only other with four in the top 50.

Chavez resumes baseball activities

Via Chad Jennings, Eric Chavez has resumed baseball activities after suffering some setbacks in his rehab from a foot injury. “He’s taking ground balls, and he’s taking BP, so he’s done okay,” said Joe Girardi prior to this afternoon’s game. “Hopefully we can get him in a game pretty soon too. I can’t tell you when, but it would be nice to get him in a game soon.” I’m not going to hold my breath, but it would be really, really awesome to have Chavez back while Alex Rodriguez on the shelf.

Update (2:39pm): Via Marc Carig, Chavez is slated to DH in a minor league game as soon as Tuesday. Wow, that’s unexpectedly awesome.

Game 91: Can we get a win?

(Photo Credit: Flickr user Jay?? via Creative Commons license)

Losing sucks, no one likes to do it. Thankfully we’re Yankees fans and our team loses a whole lot less than everyone else’s team, but dude … two straight losses after the All-Star break? Weak. The Yankees have the right guy on the mound to put an end to that nonsense this afternoon. Here’s the lineup…

Derek Jeter, SS
Curtis Granderson, DH
Mark Teixeira, 1B
Robinson Cano, 2B
Nick Swisher, RF
Andruw Jones, LF
Brett Gardner, CF
Eduardo Nunez, 3B
Frankie Cervelli, C

CC Sabathia, SP

YES will carry the game as (almost) usual, and first pitch is set for just after 1pm ET. Enjoy.

Frankie Cervelli and the backup catcher plight

Earlier this week, I posted a very unscientific/unofficial poll on Twitter. The question was simple. “If I say ‘Francisco Cervelli,’ the first thought that comes to your mind is….” RAB’s very own Ben Kabak responded with the only (sort of) positive response, “fist pumps” (hurray for enthusiasm!).  The rest of the answers were either equal to, or synonymous with, “awful” which is pretty much what one would expect.

All of the candid responses, of course, got me thinking. How does Cervelli actually compare to other backup catchers from around the American League.  And, is our collective angst really justified?  Note: the stats compiled below do not include the games played since the All-Star break.

Offense:

Click to enlarge!

In terms of offense, clearly Cervelli doesn’t inspire a whole lot of confidence. His walk rate is lower and his strikeout rate higher than the average backup catcher in the AL.  Just let that thought resonate on your tongue for a second.  He doesn’t particularly hit for power or for contact either. Most importantly, our eyes — which are never biased in the least — tell us he is a supremely gifted rally-killer, right?

Well to be fair, Cervelli’s only had 77 plate appearances which is a pretty nominal sample size.  Obviously, his limited exposure at the plate is also by design. Now, I don’t want to come across as an apologist for Franky’s offensive contributions; they are what they are. However, my point here, is that most backup catchers are very mediocre at best (at least offensively), and Cervelli is only marginally below that mark in terms of individual production.

Roughly speaking, all Cervelli would really need is about five more hits than he currently has (20 hits instead of 15, out 70 at bats) and he’d have a .285 batting average with a wOBA hovering around .320 which for all intents and purposes, is right in line with the other backups of the AL East.

In other words, given the typically high octane offensive production being generated by the rest of the team, I think there’s a bit of wiggle room to be found here.  No one should expect Cervelli to match Victor’s production, and no one should be surprised when the results are less than stellar given his role on the team.

Defense:

I think one can absolutely make the case though, that a backup catcher should be at the very least, proficient on defense.  After all, purely defense-oriented catchers are theoretically a dime a dozen.  Much to my chagrin, this aspect of the game is substantially harder to quantify though (especially when it comes to catchers).

If we utilize Fangraph’s FLD metric — which is UZR, or TZR prior to 2002 — Cervelli is definitely trailing his peers. Of course, the only catchers of the group to garner really solid ratings are Tampa Bay’s Kelly Shoppach and Cleveland’s Lou Marson. The rest of the group is average at best. Even former Yankee Jose Molina, who is often raved about defensively, comes across as very average according to the numbers.

The other stat being displayed in the chart above is Caught Stealing Percentage. Now Frankie is obviously doing himself a rather large disservice every time he elects to throw the ball to Curtis Granderson or Brett Gardner when the runners attempt to steal.  At this juncture, he’s only successfully stopped two runners out of the 21 attempts.  Opposing teams will continue to smell blood, and will continue to test him accordingly until he proves otherwise.

How much of that CS% is related to outside factors; I don’t really know. Some pitchers have naturally slower deliveries. Some pitchers offer less control (I’m looking at you AJ) which can distort the pickoff movement. Hell, some divisions could simply have more quality base runners than others.  On the flip side, if someone has a reputation for having an excellent pickoff move, runners will become deterred; if only the best base runners are attempting steals, the percentages can become skewed from that (i.e. – three successful steals out of eight total attempts).   Anecdotally speaking, I think Cervelli’s 10% CS rate is unacceptably low though.

Overall WAR:

For quick reference, let’s take a look at the old WAR rankings. It’s a pretty pathetic picture, which I think, is the point. The position is marred by mediocrity and Cervelli is on the lower end of that mediocre spectrum.  The larger point though is that the contributions gained or lost with Cervelli are not overly substantial.  In my eyes, his job first and foremost, is to simply be available to spell Russell Martin.  Secondly, he should provide some decent defense (whether or not he’s doing this is questionable), and any offense is merely gravy.  Are these needs being satisfied?  Perhaps.

The only guys really raising the averages in the list above are Kelly Shoppach, Victor Martinez, and Mike Napoli. Now, Shoppach and Napoli are not solely backup catchers; they also do some first base work and are occasionally slotted as designated hitters. Frankly, Victor is about as much of a backup catcher as I am; his value obviously stems from his bat which is why he’s given so much exposure at the plate.

Additionally, finding quality starting catchers is seemingly impossible as it is. Finding a second quality one is really just not overly realistic.  Perhaps the reason why our attitude about Cervelli is particularly soured is because we have one of the most hyped prospects of recent memory waiting in the pipeline, who supposedly could fill that need.  The only problem is, unless, the Yankees are willing to give Montero those additional at bats that would make him valuable, he wouldn’t be overly impacting on the team (I can’t account for the benefit of big league exposure at a developmental level).  Granted, it’s absolutely debatable whether or not the Yankees would be able to get Montero sufficient exposure for the remainder of the year, but as of now, it seems they are not intent on doing so.