An investigation into a major hole in Google’s knowledge
I discovered something strange yesterday. It began when, in an effort to make a stupid joke on Twitter, I tried to figure out the names of the Hobbit movies. What baffled me was that this quickly turned out to be a harder task than anticipated when Google refused to tell me the name of the 2nd Hobbit movie.
Now, to clarify, I’ve never seen a single Hobbit. While I’m a big Game of Thrones fan and I understand that George R. R. Martin’s work would not exist without J. R. R. Tolkien’s, I also think that Peter Jackson doesn’t make movies I enjoy and that three Hobbit movies seemed like nothing more than a money grab.
These are the things I knew—or, at least, thought I knew, before I started using Google to get some answers:
- There is one Hobbit book (which I’ve read and mostly liked) and three Hobbit movies.
- Peter Jackson made the three Hobbit movies. By all accounts, he probably should’ve just made one. But he did, certainly, make three.
- The first is called “The Big Journey” or something. The second is called “The Hobbit Versus the Dragon Smaug.” And the third is The War of the Five Battles or something.
- Yes, I knew these weren’t exactly what they were called, which is why I had to google this.
My awareness of The Hobbit has remained heightened over the last several years, largely due to the ongoing YouTube series On Cinema, in which Gregg Turkington repeatedly mentions The Hobbit movies as some of his favorite movies.
Turkington’s obsession with The Hobbit—and his self-proclaimed status as a #hobbithead—is one of the most compelling running themes in On Cinema, including his running belief that a write-in vote will help The Hobbit sweep the Oscars.
As mentioned before, this went from a simple search of “hobbit movies” to going down a bizarre search engine rabbit hole, in which I realized that the Google Knowledge Graph appears to have no knowledge regarding The Hobbit: The Desolation of Smaug.
This entire thing is a perfect blend of two major interests of mine: search engines and pop culture. For an example of something else I’ve written in this vein, check out the article I wrote and accompanying research about Game of Thrones and fan theories.
Before we continue this investigation, let’s briefly discuss what the Google Knowledge Graph is. If you are among the initiated with a thorough (or at least competent) understanding of Google’s Knowledge Graph, feel free to skip past this next section and jump back into the Hobbit stuff.
A brief summary and reminder of the Google Knowledge Graph
The best—and most meta—way to understand the Google Knowledge Graph is to type into Google “what is the google knowledge graph”?
If you do conduct such a search, you’ll see something that looks similar to the below.
If you’re unsure what you’re looking at, this is a screenshot of a search I conducted today to demonstrate what the Knowledge Graph is, what is looks like, and also what it is. On the left, you see conventional search results, from Search Engine Land, Yoast, and Wikipedia. On the right, you see an example of the Knowledge Graph itself, which, in turn, is defining itself.
As Wikipedia summarizes—and is excerpted by the Knowledge Graph itself—the Knowledge Graph is information collected and used by Google and “presented to users in an infobox to the search results”. Along with featured snippets, the Knowledge Graph is a major source of information for voice search inquiries, often provided via Google Home devices.
If the above is too meta and self-referential, here’s an example of Knowledge Graph information for Game of Thrones:
In this case, we see how the Knowledge Graph sources itself from various other parties, including the Amazon-owned IMDb, TV.com, Rotten Tomatoes, Wikipedia, and Google users themselves.
Now, let’s return to the matter at hand: the absence of an entire Hobbit movie within Google’s Knowledge Graph.
The Conspicuous Absence of the Desolation of Smaug
Now, to return to the matter at hand: the missing Hobbit movie.
As stated previously, I first discovered this gap in the Knowledge Graph when trying to write a funny tweet. My ensuing tweet storm began with this:
You can read the tweet thread here if you’d like. Either way, I am fairly confident—after testing this on various devices and browsers, and asking a multitude of people to test it themselves—that The Hobbit: The Desolation of Smaug does not exist within the Google Knowledge Graph.
Here are a few examples of searches for proof. In this first one, we see a search for “desolation of smaug” in which Wikipedia, IMDB, and Rotten Tomatoes all display info on The Desolation of Smaug… but Google Knowledge Graph displays info on the 2014 film The Battle of the Five Armies.
In this next one, we search for “the desolation of smaug” and see many options within the carousel of the Knowledge Graph… none of which include The Desolation of Smaug. The card itself is for the third film: The Battle of the Five Armies.
This search for “hobbit movies” shows the first and third movies but no mention of the second.
This one shows that “the hobbit 2” will surface correct results in traditional searches and an incorrect result in the Knowledge Graph:
This could go on for a while. For the sake of clarification, Google does surface information in traditional search results for “the desolation of smaug” and its variants. It will even display somewhat accurate results in its featured snippets and People Also Ask questions for questions like “what is the second hobbit movie”, as in this scenario:
What does become clear throughout these searches is that—at least as of now—the Google Knowledge Graph includes the first and third Hobbit films but not the second.
It’s not enough to know that the Google Knowledge Graph is wrong. At this point, we have to try to figure out why this is happening.
Is is Wikipedia’s fault?
The very first explanation would be that Wikipedia is inaccurate, as Google sources much of its Knowledge Graph information from Wikipedia. This explanation can be immediately eliminated in one immediate way: Wikipedia is accurate, regarding The Desolation of Smaug and its place within the Hobbit series.
It’s not inaccurate. It’s entirely missing.
This is not an issue of inaccurate information about The Desolation of Smaug. It’s not that there is inaccurate information about the second Hobbit. It’s that any mention of the second Hobbit is replaced by information about the third Hobbit.
Which leads us to…
Is this an act of protest against The Desolation of Smaug?
One argument that must be considered is that the powers-that-be at Google—or at least those responsible for the creation and maintenance of the Knowledge Graph—have conspired to try to wipe the second Hobbit from our collective memory.
In defense of this potentially being the explanation, let’s consider (and remind ourselves) that the novel The Hobbit should never have been adapted into a trilogy for the screen. It was undoubtedly both a cash grab by the studios and a mis-step by the artists.
However, in argument against this explanation, it’s also worth considering the following Metacritic scores:
- The Hobbit: An Unexpected Journey – Metacritic score of 58%
- The Hobbit: The Desolation of Smaug – Metacritic score of 66%
- The Hobbit: The Battle of the Five Armies – Metacritic score of 59%
Without having seen any of these films, it becomes apparent that, if one of them were to be erased from Google, the least likely contender would be The Desolation of Smaug.
Which now leads us to…
Tinfoil conspiracy theory: is Topher Grace involved?
There is one thing I cannot go without mentioning. Over the last few days, a series of news stories have come out about Topher Grace releasing a 2 hour version of the Hobbit trilogy, as seen here:
Of course, it’s unlikely that this is news story is directly or indirectly correlated with #SmaugGate. But it did demand to be mentioned.
How to Fix This
At this point, the question becomes how to correct this mistake. At the moment, I am not entirely incentivized to fix it, because a) I think it’s funny and b) I would like to see what happens if Gregg Turkington and his fellow Hobbitheads get involved.
However, my tweets thus far have already attracted the attention of Google. Specifically, the part of Google responsible for the Google Home, who presumably initially thought there was an issue with one of their devices.
From here, the process would involve either:
- Filing feedback in the manner they recommended or
- Getting someone involved with The Hobbit to officially reach out to Google by clicking the place in the bottom right of every Knowledge Graph card, where it asks if you “manage the online presence” for the item in question.
In the case of #SmaugGate, we would have begin the correction through either The Hobbit: An Unexpected Journey or The Hobbit: The Battle of the Five Armies, as we cannot, for obvious reasons, tell Google that we manage the online presence of The Desolation of Smaug.
Someone who does manage the online presence for The Hobbit would begin by clicking the above. Which does suggested that, yes, this is the partial responsibility of not just Google but the movie studios who allowed this situation to manifest itself.
And yes, all of this does matter
At this point, it could be tempting to brush all of this away. Who cares about The Hobbit? Why was there ever more than one Hobbit made? Is it not good for Google to eliminate such a misstep from our collective history?
The reason this matters, as far as I’m concerned, is that it points to further questions. If the Google Knowledge Graph is missing the second film from a billion dollar film franchise, what else might it be missing? To what extent are we relying on flawed, faulty, and fabricated sources for all of our information?
What if this were to happen, but with a political party? A country? A war? An important novel? A news story? A crime? A world event?
When did the truth become so malleable, so subjective, so flexible? Perhaps it has always been this way. What’s important—and why I felt compelled to write this—is that we hold our sources of truth accountable and that we learn which truths are true. Which is why, of course, the ultimate answer is to submit something to Google to make them aware of the gap in their knowledge.
And, in the mean time, please let me know if you find gaps of information on par with Google’s ignorance of Smaug and his desolation.
Update: Tuesday, 8/28
Google has updated the Knowledge Graph. I discovered this on Friday, August 24, although it’s worth noting that I had not checked Google in about two weeks before discovering this, so it could have been updated any time between 8/3 and 8/24. Upon my discovery of this update, I immediately tweeted about it.
Please note that I did not take any of the steps detailed above to fix this, aside from writing this article itself. A few people have asked if I think I’m responsible for fixing this. I can only assume that The Hobbit: The Desolation of Smaug would still be missing from the Google Knowledge Graph had this article not been written.