I’m Evan Savage and I’d like to welcome you to Quantified Savagery. I recently left my job at Facebook to focus on exploring the Quantified Self, and I’m super-excited to share those explorations here on my new blog.
In this first post, I’ll explain the Quantified Self, give you a sense of what I’ll be posting in the near future, and provide some helpful tips on reading this blog.
What is the Quantified Self?
I’ve had to explain this countless times to friends, family, and co-workers: why did I leave one of the world’s best employers to explore a field most people haven’t even heard of? I usually start by name-dropping Fitbit or Nike+ as prominent examples of personal data collection and analysis. I then do some semi-coherent hand-waving about the vast potential of data collection and analysis. All of this is really just an attempt to cover up the fact that I don’t really know.
That doesn’t really cut it as an explanation for a major life decision, though, so let’s look a bit deeper.
The Quantified Self community website has this tagline:
self knowledge through numbers
- self: you seek to answer your questions.
- knowledge: in doing this, you gain an awareness of your behaviors and motivations.
- through numbers: this process is driven by data gathered from sensors, journals, and any other tools at your disposal.
This is a good first-level approximation: you gather your data, analyze it, and interpret the analysis to become more self-aware.
But why is this suddenly important? After all, journal-keeping has been around roughly as long as written languages. The answer lies in technology. For the first time in history, over half the world’s population owns sensor-packed networked computing devices. We refer to these devices as mobile phones only by historical accident. In fact, they’re really powerful tools for speeding up this process of gaining self knowledge through numbers.
As I said before, though, this is only a first-level approximation. There are two main ways in which Quantified Self can achieve greater awesomeness: Quantified Mass and Qualified Self.
As Gary Wolf pointed out in his interview with On the Media, self-tracking doesn’t lead to self-obsession but rather to group-awareness. In asking our own questions, we find that these questions are important to others as well. When many people gather comparable datasets to answer the same questions, there’s an opportunity to extract insights that could benefit us all.
Scalable mass insights have massive power. Taking a 1% chunk out of the American obesity epidemic might not sound impressive, but that’s potentially a $6 billion impact on direct and indirect costs. And that’s just in the United States, which counts for a tiny slice of the global mobile userbase.
Many of the requisite data mining tools already exist, but they’re being employed to increase advertising click-through rates by 1%. The engineers building these tools aren’t indifferent to societal problems; rather, the datasets to solve those problems largely don’t exist yet. Once they do, the quantified mass can start driving these massive-scale incremental wins.
Not everything that counts can be counted,
and not everything that can be counted counts.
Our perception of life is rarely numerical. Much more often, it is visual, auditory, tactile, or experiential. The problem with data is that you can’t see or feel them. Even the best data scientists use data only as a means of telling a story.
Put another way, this awareness process starts with qualitative questions and ends with qualitative answers. Data is the intermediate representation, one we use for its unique ability to permit detailed analysis. Ultimately, though, we’re going to ask questions like
How can I improve my fitness?
and expect answers like
By finding training partners. By doing more engaging athletic
activities. By setting aside regularly scheduled time.
We’re going to need moral support. We’re going to give and receive advice. We’re going to have conversations and tell stories about our personal struggles with fitness.
These qualified aspects of self-awareness are arguably the most important to us. Data provide a stepping-stone, something we can build upon to address these aspects. By building systems designed for the qualified self, we can bring the benefits of the quantified self to everyone.
Back to the Quantified Self
That sums up why I’m so excited about Quantified Self: there really is an enormous potential here to revolutionize our lives on both the global/societal and individual levels. It’s also a fantastically diverse field, one that connects hackers and doctors and entrepreneurs and teachers and artists through mutual pursuit of insanely lofty goals.
The next few posts will detail my experiences dealing with panic disorder through self-tracking. I gave a talk about this to the Bay Area Quantified Self community, which you can view here for some initial context.
How To Read This Blog
Although this is a blog about personal data, one of my primary goals is to make the thoughts and insights shared here accessible to a broad audience. You can filter what you read with these categories:
- Non-Technical: expect to see insights, thoughts, discussions, descriptions of planned experiments, and post-mortems. I might link to articles, studies, or books that provide context, but I’ll try to summarize the relevant parts.
- Technical: expect to see code, statistical analysis, $ \LaTeX $ formulae, links to Github repos, and algorithm descriptions. I’ll assume familiarity with programming and mathematics, or at least a willingness to learn.
In addition, many of my posts will be connected to one or more experiments. For instance, my upcoming posts on self-tracking to address panic disorder will fall under the Panic category. For every experiment, I’ll attempt to post content in both the Non-Technical and Technical categories.
Of course, I’ll be glad to answer any questions you have, technical or otherwise!