New bureaucracy takes the form not of a specific, delimited function performed by particular workers but invades all areas of work, with the result that - as Kafka prophesied - workers become their own auditors, forced to assess their own performance. (Fisher)

I collect a lot of data about myself. I track every workout and every kilojoule. I run time-tracking software. I wear a smart watch. Each day of my life is associated with at least 200 data points, many of which are averages or totals derived from a near-constant stream of personal information: my heartrate data becomes a resting heart rate estimate; every second of activity on my computer becomes a sum total of ‘productive minutes’ for that day. And when I say ‘200’, we’re also not counting the softer, natural language stuff. I journal most days, for example, and I do it on a computer.

It’s a lot.

To be fair, many of those 200ish data points are collected with very little input from me. In a technological ecosystem overrun by free-to-use but best-in-class products and services with near-universal adoption, bleeding personal information is the norm. It’s a painful cliché, but if you’re not paying for a product, you are the product. And while I may work hard to imagine a life outside of market pressures, a company like Google rarely does. They’re incentivised to collect as much data about me as possible, because every datapoint could (one day) be worth a fraction of a fraction of a cent for a shareholder. The only difference between me and a slightly less obsessed casual user is that I use a service called to collect as much of my own data as possible into a single location. And, of course, I augment the deluge of passively-generated data with other things like mood and social data.

Back when I was an eye-twinkle, this process of obsessive integration was called building a ‘Quantified Self’. Now we just call it capitalism.

For a tech-savvy, health-conscious millennial like myself, this is the world. Yes, I monitor myself because I can. But I also monitor myself because, in a very real way, my ‘self’ is a bounded product to be sold to potential employers (or lovers, friends, etc.). My health and productivity determines the value of that product. It feels like my responsibility to optimise.

While it’s certainly true that I have self-centred reasons for wanting to be the best and healthiest ‘me’ I can be, I’m also – like many of my generation – a product of endless (unpaid) internships and (fragile, underpaid) casual contracts. Many of my rational, self-interested reasons have been co-opted by, and reduced to, financial/status pressures. In the situation that I’m most familiar with – that of the research assistant and graduate student at a major university – there’s also an absurd tension underpinning all of this:

While University mission statements are stacked full of commitments to integrity, valuing staff and caring; the USA’s adjunct workforce is one of low pay stress, limited health insurance, and contract precarity. The situation is mirrored in much of the UK, as the number of PhD graduates compete for limited work, with a shrinking pool of permeant posts. Amidst this situation, staff, as well as students, are pointed towards the ever growing portfolio of well-being and resilience programmes, when one might suggest that the source of the stresses causing the need for such interventions also lie within the ability of Universities to fix. Adjuncts don’t need wellbeing courses, they need proper contracts and pay. (Rivers & Webster)

I collect data about my health and my productivity with obsessive fervour, trying to optimise the tiny fraction of my life that is in my control. It’s totally rational for me to do this, and it’s been quite successful. Present Me is happier and healthier than Past Me. I’ve given up flexing against it, because what little I control can make me measurably and substantially happier. But it’s strange.

Dimensionality Reduction is a Kind of Survival

It’s true that what gets measured gets managed. Unfortunately, when you measure 200+ things, ‘management’ can smear a little too thin. If you want to be better – healthier, whatever – you need to focus on the essentials.

The traditional solution to this problem is to ignore most of the datapoints and obsess over a single metric, like steps. For me, though, this seems counterproductive. I can be pretty … focussed … at the best of times, so it’s likely that I’d end up over-optimising that one thing at the expense of everything else. It’s also the case that ‘Rational & Reflective Me’ wants to be happy and healthy and well on every measureable front.

So, rather than ignoring most of the data and focussing on one tree at the expense of the forest, I’m going to compress each day of activity into a smaller set of metrics. I want a set of numbers that describe my current wellbeing in a way that’s simple, meaningful, and still optimisable.

If you’re looking to describe yourself with only a handful of numbers, there’s only one system that matters: Dungeons & Dragons.1

While most of the details of the d20 system are impractical abstractions for my purpose, the core notion of ability scores is pretty useful. There are six numbers that represent a person’s general ability to take any action they might want to take: Strength, Dexterity, Constitution, Intelligence, Wisdom, and Charisma. Any time you want to do something and there’s uncertainty over whether someone could do it successfully, you roll a d20 and add your ability modifier to the roll. If the number you get is higher than the difficulty of the task, you ‘pass the check’ and successfully do the thing. A normal person in the world is around 10-11 on each ability, which translates to an ability modifier of 0. The best in the world – the peak possible human – can get to an ability score of 20 (mod: +5). A legendary, superhuman beast can get as high as 30 (a +10). Only a carrot would end up with a 1 (-5).

In the at-the-table social world of D&D, there are also more specific rules of thumb about the way to benchmark these attributes in real world experience. The most common I’ve heard, for example, is that a character’s IQ is their Intelligence score times 10. That works. Probably the most interesting rule of thumb, though, is the relationship between Strength and a character’s carrying capacity as enshrined in the rules themselves. As detailed in the 5th Edition Player’s Handbook (and the SRD):

Carrying Capacity. Your carrying capacity is your Strength score multiplied by 15. This is the weight (in pounds) that you can carry, which is high enough that most characters don’t usually have to worry about it.

Push, Drag, or Lift. You can push, drag, or lift a weight in pounds up to twice your carrying capacity (or 30 times your Strength score). While pushing or dragging weight in excess of your carrying capacity, your speed drops to 5 feet.

Size and Strength. Larger creatures can bear more weight, whereas Tiny creatures can carry less. For each size category above Medium, double the creature’s carrying capacity and the amount it can push, drag, or lift. For a Tiny creature, halve these weights.

This is an intentionally simple ruleset, but it can lead to some strange real-world equivalences. Let’s think about the extremes. If a person has a strength score of 2, for example, they can put 30 pounds (13.61 kg) in their backpack and walk around freely, but they’re unable to lift something more than 60 pounds (27.22 kg). If a person gets a Strength score of 20, that means they can deadlift 600 pounds (272.16 kg) and walk around carrying 300 pounds (136.06 kg) worth of stuff without it slowing them down. The world record for a deadlift is a mind-bending 500kg (1102.3 lb), though, which means the current record holder (Eddie Hall) clocks in with a Strength score of around 36. By D&D standards, that’s a long way into the superhuman range. It’s possible that Mr Hall has had a pretty high level anabolic steroid spell cast on him, of course, so I’ll let the 272.16 kg benchmark stand. However, the average stat (10) would lead to a “dude, yeah, it’s chill, I got it” carrying capacity of 150 pounds (68 kg) and a max deadlift of 300 pounds. I couldn’t find anything but bro-science on this, but that seems pretty optimistic. I hiked with a 60 pound pack once, and I would not have described myself as moving freely.

So let’s hold these benchmarks loosely. If I’m going to think about myself as a Dungeons & Dragons character – as a way of coping with the strangeness of data-driven self-improvement and self-auditing under late capitalism – where should I stack up?

What Are My Ability Scores?

Here’s what I’ve done. First, I’ve taken a list of all the attributes that recognises. Then, I’ve augmented that with attributes I collect about myself (processed, mostly, via’s custom tags). I’ve then assigned each of these a range of possible attributes, along with whether the attribute has an effect on any ability score(s), and whether it’s good to have more, bad to have more, or best to stick as close as possible to the middle value in the range. You can see a .json file that represents all this if you want.2

With these arbitrary-but-workable benchmarks in hand, I’ve built a painfully simple script to do four things in order:

  1. Download the last 90 days of data from;
  2. Calculate an exponentially-weighted moving average value of those last 90 days;
  3. Compare each value to the range of possible good-to-bad values; and,
  4. Get the average bonus/penalty for each associated D&D ability score.

The result, dressed up a little, is available at It updates every 20 minutes. It also colours the ability scores green if they’re getting a boost and red if they’re trending downwards.3

After all this, the insight is unsurprising:

I’m pretty average, most of the time.

  1. For readers totally uninterested in D&D’s mechanics: look, let’s be honest with each other and agree that this blog is probably not for you. No hard feelings, but if you’re 900 words into a tongue-in-cheek post about late capitalism and the Quantified Self movement but you’ve never at least skimmed the D&D Player’s Handbook on a slow Saturday afternoon, I don’t understand you. I’m probably going hurt your feelings with a misplaced pun pretty soon. And even if I don’t, well, I just really don’t think that we see eye-to-eye about the joys of being human. 

  2. A lot of these are, necessarily, hand-wavy judgment calls on arbitrary ranges. It’s good to have more steps in a day, I decided, but that benefit caps at 15,000. Fibre is good for you, but you’ve gotta eat a lot: if you’re between 0-35 grams a day, you’re being penalised, and you don’t get any bonus points above 70. If you’re spending 120 minutes a day on distracting activity you’ll get a pretty whopping wisdom penalty, but it doesn’t get worse after that: three hours of youtube is worse than two, I suppose, but how much worse is it really

  3. Hence Sorry. Yeah. It’s dumb.