When it comes to Apple products, I've always been a slow adopter. My first Apple device was the last traditional iPod, the 120GB Classic. My first computer was a 2010 MacBook Pro, and I did not get an iPhone until the wonderful 6 launched. The only time I was ahead of – or rather on – the curve was with AirPods, which I got already in their 1st generation.
In light of that, it's no surprise that my first Apple Watch was the 4th generation.
But why did I, a previously militant proponent of mechanical watches, get something as far from something with a nice automatic movement as it gets?
In the last 2-3 years I became more and more interested in the whole concept of quantified self, particularly when it came to working out. In November 2017 I started going to the gym regularly, doing mainly indoor cycling, HIIT, and yoga. On the weekend I would also continue doing rides on my racing bike. Measuring my progress on the bike(s) was easy - for indoor cycling every bike has a computer which gave me a nice summary after each ride. Outside I would use Strava on my phone to track myself using GPS. Doing so for other workouts was much harder, though.
At the beginning of last year, when one of my colleagues got the watch, I got really intrigued. Of course, I've heard about Apple Watch before, by it was only when I saw the Activity app that I got really intrigued. After couple of weeks of deliberating, in March of 2019 I finally caved in and bought a 44mm GPS-only Series 4 Apple Watch (Nike+ edition).
Since then, I've really enjoyed using the watch, especially the Activity app, helping me track all my workouts without the need to use my phone or other accessories like my HR strap. One thing I missed, though, was a long-term perspective to it, as most metrics are limited to single workouts or single days.
Earlier this year, I got the idea to try to get my Activity data out, and do my own analysis with charts and stuff. When I started looking into options to get the data, I was expecting a typical Apple walled garden with a lot of ways in but very few ways out. I was really surprised to find that the Health app on the iPhone has an export feature, which creates an archive of all of the data, including activity and workout data. Even though the raw export is a bunch of somewhat clunky XML files, thanks to Mark Koester and his qs_ledger translating them into more digestible CSVs was pretty easy.
To get a better understanding of the last year, I looked at the data and analyzed some of the main metrics to see when and how I was active and what can I work on this year to close even more rings (or close them more) and fulfill my need for gratification. And because just looking at the data is boring, I also decided to write about it and build pretty colorful charts to show you my last year as seen by my Apple Watch.
Before looking at the specific data, it's important to see what was actually the sample it's size. For the purpose of this analysis I looked at data between March 18th, 2019 and March 18th, 2020. When it comes to statistically significant days in this range, I considered a day as active when I burned more than 10 kcal.
Out of the 367 days I was "active" 96% of the time, with only 14 days where I would consider myself not active.
Those were usually sick days when I was at home and did not even put my watch on. In August there is also a 4 day period when I was on holidays in Slovenia and did not bring a (functioning) charger. The gray days are those outside of the observed sample.
Now that we've established the sample size, let's start with the basics - activity, in the world of Apple Watch known as active energy.
The watch, on it's own, measures active energy as calories burned while doing something that has at least the intensity of a brisk walk. In reality, this is pretty much anything apart from sitting around on your couch or walking at a leisurely pace.
The first thing that I looked at was how did my average week look like between March 2019 and March 2020 in terms of activity and average active energy burned.
What this chart clearly shows is that in the last year (or 367 days) I had a pretty fixed workout routine, always working out on Monday, Wednesday, Friday, and Saturday. Tuesday and Thursday were my laziest days, on average not even reaching my daily goal.
Given this pattern, if I were to take it one step further, how did my average week look like on an hourly basis?
As before, a clear pattern emerges - clear cut 60 minute workouts on Monday evening, and Wednesday and Friday morning, with Saturday and Sunday being more blurry, since my schedule was not that regular and I would often combine and alternate workouts with bike rides of various lengths. My most "on fire" hour was Wednesday between 7am and 8am, averaging 376 kcal in active energy, which is almost half my daily goal.
Overall, over the course of last year I averaged 1033 kcal of active energy per day, with the max being August 25th, the day of Cyclassics in Hamburg, when I burned over 3777 kcal.
Now, where does all this energy comes from? The main source of active energy for me were Workouts, that I explicitly tracked with the Watch. This can be anything from an Outdoor Walk and Indoor Cycling to High Intensity Interval Training or even Dance, which comes handy during those 72h stints in Berghain. Intuitively, this makes sense since a 45 minute HIIT class can easily get me to 600 kcal, which otherwise represents a solid couple of hours of regular activity.
During a workout the watch uses the built-in optical HR sensor to get an accurate reding of the heart rate, instead of relying on only the movement sensors (gyro and accelerometer) to approximate calorie burn. This means the calorie reading during workout are more precise.
On top of that, during a workout the Watch shows the relevant info on the screen, which can be configured depending on the Workout type.
For each workout type I compared not only how much I burned, but also how long I spent doing that, giving me a good idea about how efficient each workout type was.
|Energy burned (kcal)||Time spent (min)||avg kcal/minute|
As someone who loves cycling, it's no surprise that most of my workouts are cycling-related, either a full-blown bike ride, or at least an indoor cycling session at BECYCLE. In total I spent over 104 hours doing some form of cycling, with a little over 1/3 of that outside.
The least frequent workout type was Other, which represents all the odd occasions I needed to get a couple more calories to hit my goal and would proceed to jump around and do pushups, jumping jacks, or squats to hit the goal.
When it comes to workout efficiency (bang for the buck or kcal for the minute), outdoor cycling remains the king, burning over 14 kcal per minute of workout. Indoor cycling remains close 2nd, with running closing the top 3. To my surprise, HIIT was only fourth, which might be caused by the fact that unlike with other workouts, there are always lower intensity warmup and cooldown periods included in the time of the workout, therefore lowering the overall kcal/min ratio.
As before, I've also looked at the distribution of working out over the course of an average week. Unsurprisingly, it almost exactly mirrors the active energy chart, with peaks of more than 30 minutes of working out at the time of my classes.
The third of the main metrics that the Apple Watch measures (and evaluates in terms of daily goal) is standing. As opposed to the other two goals, which are measured as absolute values, standing is defined as the number of hours in a day where I was standing (or active) for at least one minute. The daily goal is 12 hours, which is usually not that hard to achieve, especially since the watch reminds you to stand up 10 minutes before every full hour.
Since the release of watchOS 6 in September 2019 the watch also started tracking the absolute time I was standing, which I find more interesting to look into than just the hourly goal. As with active energy, I've compared the median minutes stood per day of the week.
It's hard to say whether my figures are good or not, since I couldn't really find any reputable benchmark on daily standing. Also, the Internet is flooded with resources about daily stand-ups, which makes it really hard to find something non-tech related.
What is clear, however, are once again the patterns, with a clear correlation of workout days with higher amount of standing (I'd bet that workout counts as standing regardless of whether I'm riding, doing situps, or running). This time, however, weekend takes the lead, with more than 4.5 hours of standing on Saturdays.
Now that we've covered the three basic metrics (Activity, Workouts, and Standing), we can look how Apple Watch ties them all together.
The whole concept of Activity is focused around gamification, which is represented by the Activity rings. There is a ring for each metric, and your goal is to close the ring each day by (over)achieving your goal. The daily goals are always 30 minutes of working out (or activity), 12 hours standing, and customizable amount of active energy. I have mine set to 770 kcal per day.
On top of the daily goals there are also awards, which are split into different categories:
I'm not going to brag about all the awards I've collected 😏, and rather focus on the numbers. So, how often did I hit my goals?
|Goals hit in days||Goals hit in %|
First of all, the Move goal. My goal, 770 active calories per day, seems to be exactly the amount that it easy to hit with a workout, but hard to hit otherwise. The 60% hit rate is therefore quite a good representation of my regular week with 4-5 days of workouts.
As expected, the Workout goal was the easiest to hit - on a regular day, 30 minutes of brisk walking is pretty much my commute to work, so hitting it 90% of the time wasn't really challenging.
The Stand goal is usually a bit trickier. The hourly notification is easy to ignore or miss when I'm focused on something, and missing couple of hours in a row makes it tough to catch up in the evening.
One thing that I have not yet really figured out is how precise the overall measurement is. As I mentioned before, outside of Workouts majority of the measurements are based on the motion sensors (accelerometer and gyroscope). This can work quite well for things like counting steps, but e.g. standing is, based on my experience, more tricky, and it is also the most inaccurate of the three metrics. Sometimes it's enough to just quickly walk to get coffee to appease the Watch while in other cases it would remind me even after I spent 20 minutes of the hour cooking and doing the dishes.
The whole topic of active energy is also hard to evaluate. From my understanding, the non-workout active energy is based on some average energy burn, which might be adjusted using regular HR readings (every 30 minutes or so). Unfortunately I am not wearing a second HR monitor the whole day, so I honestly have no idea how close are the figures that are generated outside of workouts close to reality.
What I can compare, however, is the precision during workouts, or at least use my outdoor cycling workouts as a benchmark. From my experience the accuracy of the built-in optical HR monitor is pretty good, as the active calories burned during a workout are almost always very close (within 5%) to the reading of my bike GPS, which uses a electric HR chest monitor. I attribute the difference partially to the technology, and partially to the difference in measuring the time of the activity.
I'd be curious to see whether with more active workouts (like HIIT) there would be more of a difference, since they involve a lot of movement and flailing around.
Electrical chest monitors use two electrodes in the strap to measure the heart activity just like ECG mesurement. Optical, on the other hand, shine a green or red LED through the skin and measure the light refracting off the flowing blood.
What seems to work pretty fine is the exercise time. Aside from me telling the Watch that I'm working out right now, the movement sensors in the Watch are smart enough to realize when I'm actually moving and not just reaching for the remote to pick the next TV show on Netflix. Also, just waving my arm in the air usually does not help. Believe me, I've tried.
So, how did the watch change my behavior? It's hard to evaluate that based on the numbers, as I have very little to compare to. After all, the main reason to buy the watch was to have some way to measure myself.
I feel like Watch definitely played its designated role of annoying motivator, pushing me to stand up more, walk more, and be more active to avoid breaking my streak. As I mentioned earlier, there were multiple occasions in which I squeezed in late night workouts to close my rings and keep my streak, to visible entertainment of my girlfriend.
Having the Watch and measuring my workouts also increased my awareness of how much does my body burn. In the past I would consider all workouts equal - it would just be a way to tick the mental box of doing something to stay fit. While that still applies, I also realize the differences in energy output of the different exercises and workouts. In practice, that means that a 5k run is definitely not an excuse to just lay down and snack for the rest of the day, while after a ride I'm probably not going to do anything else.
My goal for the next year is to try to increase my activity throughout the day to even out (or catch up with) those workout spikes a bit. Despite closing my rings I still spend most of my time sitting at my computer, which is definitely not very healthy. I want to find a way how to move a bit more during the work day and maybe even increase my daily activity goal to keep myself motivated.
Is it going to work? I'm not sure, but I know that the Watch will for sure pester me and motivate me to do so. Because what's not in the Watch does not count.