The Limits of Projections: Franklin Gutierrez
Yesterday, we talked about Nelson Cruz and how his frankly bizarre late-career surge makes him especially hard to project. If the historical similarity scores bring a bunch of players with only slight resemblance to what Cruz’s career arc looks like, you’re not going to be all that confident in the result. This is a very different spin on the Ichiro problem – that PECOTA never knew what to DO with Ichiro, because his historical comps were all dead-ball era players. That’s just going to happen with every player at some tail end of the distribution.
Franklin Gutierrez, though… We’re not even looking at a distribution. Systems are trying to project a player based on sporadic playing time over several years, and violently conflicting production between bouts of near-crippling illness. If Nelson Cruz’s similarity scores bring back Jermain Dye, Hank Sauer and Ellis Burks, Guti’s top comps are catarrh, flux, and biliousness (most similar through age 32: King’s Evil). Projections generally don’t have to consider athletes who’ve missed this much time due to injury, because those players retire. Projections don’t generally have to consider players who hit .211/.270/.340 in the hitter’s haven of the Pacific Coast League, then sit out an entire year to rest their failing, almost mutinous, body and then hit .292/.354/.620 in a short big league stint.
To say that projection systems weren’t built to handle this is both obvious, and it gets you to pondering that Franklin Gutierrez himself may not have been built to handle this. Nelson Cruz is a couple of standard deviations from the mean aging curve, but he was a recognizable type of player whose yearly statistics just look funny when read sequentially. Franklin Gutierrez is part role-player, part symbol of hope and recovery, and a big part cautionary tale. Cruz’s case teaches us about the limited predictive value of edge cases. Gutierrez teaches us about our limited narrative abilities.
“Narrative” tends to be a pejorative in many segments of the game, or discussions of the game, but it’s a word that can mean a variety of things. To sabermetric fans, it’s often used regarding stories that jump from results to psychologically-tinged explanations of them. A declining batting average is the sign of a player wilting under the bright lights of New York, or fading in the doldrums of a go-nowhere Atlanta team. This closer’s blown saves prove he doesn’t have the intestinal fortitude to close, and this reliever needs a clearly-defined role to succeed. Even teams get in on the act, as this article about the Braves pushing back on the “narrative” that their past two offseasons constitute a rebuild.
Of course, everything about statistics needs some kind of frame, some story – this is what happened, and this is what I think it means. It’s been years since Fangraphs was just numbers and graphs; they’ve got tons of words on there now, and more of them every day. Journalists – traditionally the target of the pejorative sense of the word “narrative” – occasionally push back at this: you don’t like my explanation of what happened? Fine, say why – don’t just shut down discussion by scoffing “narrative” and walking away. Look, I hate pieces that substitute pop psychology (“they just didn’t have the will to win”) for meaningful analysis of a team’s successes and failings, but both are ways of explaining why something happened. We all actually WANT narrative, but we don’t want shitty, cliched narrative, and yes, sabermetric writers are capable of producing that sort just as much as any on-deadline beat writer. We don’t want familiar tropes about this or that specific 5-1 loss, or another bullpen failing. We all desperately want some kind of meaningful, illuminating story about what the heck happened to Mike Zunino or DJ Peterson last year.
Franklin Gutierrez is a human rebuke to dull, cliched, boilerplate stories. Everything about Franklin Gutierrez is insane, and teeters on the edge of credibility, from his defense in 2009 to the fact that HE SLUGGED .620 LAST YEAR, AND YES THIS NEEDS ALL-CAPS. Describing anything about his career simply *requires* exclamation points. Some of my favorite pieces on this website the past few years have been about Guti, a players who’s logged all of 500 plate appearances since *2011*. The thankful, the depressed, the philosophical and mournful – Guti *makes* us do this. You don’t “project” Gutierrez’s statistics, because you can’t know when you’re going to see any. You know you’re going to feel something when you do, though, and that’s much more interesting.
Comments
11 Responses to “The Limits of Projections: Franklin Gutierrez”
Leave a Reply
You must be logged in to post a comment.
Quantitative analysis of a qualitative subject. I understand the objective of trying to somehow make predictions based on “fact” rather than just some gut emotion.
It is a “fact” that Robby Cano had a bad year in 2015. How much of that was injury?
It is a fact that Nelson Cruz had a great year. Was it a “blind squirrel” or did he find a mechanical flaw that he corrected?
Guti blew the doors off of every possible prediction last year. Was it a hot streak or is he just that good and illness sidelined a HOF career?
You can analyze the crap out of it and even if your answer turns out to be “correct” did you really predict the result or did your “prediction” just happen to line up with the stars?
The KC Royals did not finish last in the AL Central the last two seasons and our beloved M’s are not the defending WS champs.
There are only five possibilities of where the M’s can finish in the regular season, so even Punxsutawney Phil has a 20% chance of being correct.
Nate Silver can go 50 out of 50 predicting statewide results in a presidential election but what is his “batting average” when predicting where a team will finish come the end of September.
Maybe that is why we love baseball more than politics. In baseball unlike politics everyone really does have a chance.
I love Franklin Gutierrez. He is ten kinds of awesome sauce.
.620? Lol i don’t know how I don’t remember seeing that, its all shades of awesome.
Theres not enough wood in my house to knock on for me to risk talking about Guti at any length, I’m going to abstain and hope for the best.
As a Blazer fan who’s lived through Greg Oden, Brandon Roy, Sam Bowie and everything in-between, I just cherish watching him healthy and performing.
And yeah, I’m positively giddy after Lillards absolutely insane video-gamish heat check night against Steph and the Warriors. Its been a fun evening, bartending and watching dozens of Blazer fans absolutely lose their minds for 3 solid hours.
Love Guti too much, can’t talk rationally about him.
If his health is equal to last year then I think his performance will be too.
If his health is equal to last year then I think his performance will be too.
Can you walk me through the rationale behind your conviction that someone with a career 10% HR/FB ratio has suddenly, at 32, established as his new normal the highest HR/FB in history–far, far higher than late-career Barry Bonds, or peak Albert Pujols, for example–and has become, at age 32, a world-historically great power hitter?
I really appreciate this explanation of narrative. The narrative is an attempt at an explanation in an accessible way, to me at least, and the cliche is the most accessible available device.
A guy on Twitter provides amusement by posting Alfred North Whitehead quotes. He tweeted this one recently: “There are no brute, self-contained matters of fact, capable of being understood apart from interpretation as an element in a system.”
Maybe that’s germane?
Franklin Gutierrez slugged .620 last year. He did not play everyday, but when he did play, he performed exceptionally well. I can only guess that he was extremely focused while in the batter’s box, and had the physical skills to match. I hope he continues, for whatever reason, to play at this video game level.
I’d like to talk about the other side of Guti, for a moment.
I think we can all agree that, defensively, Gutierrez is not the player he once was. He’s still decent as a corner outfielder, but he’s not making the absurd plays he would’ve in 2009.
So I would like to propose a new nickname for him – UTFT (Unfriendly towards flying things).
I second that notion, west. Let’s also not forget about Randy Johnson, the patron saint of unfriendliness toward flying things.
https://www.youtube.com/watch?v=FCNZg2xwl54
For those who want a quick puff piece on Johnson with a still and grainy vid on the seagull elimination.
Thanks Marc. Parsing Guti, regardless of how it’s done, is something akin to the dark arts. At the very least, I admire the gumption in trying.
The shame is that we got the return of Guti (the player we all love and got for essentially free this time) and his .620 SA in the wrong year. Imagine if we’d had him in 2014, when we missed the playoffs by a single game.
Or, in a single word, Mariners.