As another year comes to a close, it’s time for resolutions. I’ll spoil this for you: I don’t like resolutions. They typically don’t work — unless you’re a shareholder at a gym, in which case they work fantastically. Still, the week I have off from work is a time when I can’t help but reflect on the past year. I could reflect on the big moments: learning to ski, going to Newport Folk for the first time, running a half marathon, losing my grandfather — or the small: poker nights, finishing up my living room decoration, 52 Monday meetings with my team. That’s all worthy of reflection, but it’s all far too anecdotal for my neuroticism.
I need data.
I need numbers.
How did I do this year? What measurably contributed to my happiness? Maybe if I just solve for X I can find the meaning of life. The good news is that directly or indirectly I’ve been tracking a ton of different things in my life, many of which I can now compare for 2023. Is doing this going to help me improve my life? Probably not, but at least we can find some fun insights about Colin in 2023.
Let’s take a look at what I tracked.
- Day Quality
- If I Cooked
- If I Read
- Sleep Hours
- Number of Drinks
- Whether I left Home
- Traveling
- Phone Mins
- Caffeine
- Calories
- Fat (g)
- Saturated Fat
- Polyunsaturated Fat
- Monounsaturated Fat
- Trans Fat
- Cholesterol
- Sodium (mg)
- Potassium
- Carbohydrates (g)
- Fiber
- Sugar
- Protein (g)
- Vitamin A
- Vitamin C
- Calcium
- Iron
- Fitbit steps
- Weight
You’ll notice most of these are food-related – huge shoutout to Myfitnesspal for contributing that. I also have removed a few that were either irrelevant or that I don’t want to be posting online (Hel-lo future employers!) Still, it’s a sizable list to work with. Once I assembled this all in one place, all I had to do was apply a simple =CORREL() function to each column and I could see the relationship between them. When I showed this to my roommate he suggested I get out more, but I was able to reference the variable 6 to show him I do get out a decent amount. I don’t think he was impressed.
Here is the result of my labors.
the ark of the covenant is in here I just know it
I think the most important column is Day Quality. This is where each day, I’d record how I felt my day went on a scale between 1-5, 1 being bad 5 being excellent. Naturally, most days were 3’s, with fewer 2’s and 4’s, and even fewer 1’s and 5’s. There are a million things that can affect this, many of which I wasn’t recording. Sadly there isn’t a column for “woman next to me on the flight invaded my personal space.” There are still some very interesting insights here.
My prediction that my happiness totally revolved around my monosaturated fat intake has been demolished by this data, I will likely never recover. The biggest positive relationships are:
- Fitbit Steps +0.51
- # Drinks +0.37
- Caffeine Intake +0.36
This is a great time to note the difference between Correlation and Causation. We have to do some investigation and contextualization to make use of this information. In other words – just because I drink 10 shots of vodka and 15 coffees does not mean I’ll be writing a 5 that day. In the case of drinks, sure, they can be very fun and day-improving – but when I’m drinking it’s more likely to be a weekend, or a vacation, or an outing with friends, or a date. All things that themselves contribute to my happiness, probably more than the drinks themselves. This is also true of caffeine. Fitbit steps are a more direct link – Studies have found that achieving more steps a day “had significantly lower anxiety, depression, anger, fatigue, confusion, and total mood distress scores compared with measurements taken prior to the intervention.” I am no exception. But still, even this is more complex. I’m walking around more when bouncing between activities, which improves my happiness. Still, a 0.51 correlation is strong.
Lesson 1: Take more daily steps in 2024
The other variable that stands out concerning happiness is phone time – it has a significant negative effect (-0.36). This is another that works both ways, if I’m on my phone a lot I’m probably at home, and if I’m at home I’m less likely to have an above-average day. This aligns with phone time’s negative correlation to whether I left home and daily steps (-0.27 and -0.28 respectively.) That doesn’t change the fact that the less time I spend on my phone, the better. My objective should be to reduce time on my phone with things I enjoy more, as that’s what led to happiness in 2023.
Lesson 2: Reduce time spent on my phone
Given that steps is such a powerful connector to daily happiness, I think it’s worth taking a further look.
We can see some familiar faces here that we’ve already discussed – it correlates with day quality, and drinks because if I’m drinking I’m more likely to be running around, and it inverse correlates with phone use for the opposite reason. We do learn more. The more steps the more calories I eat and more specifically calories from fat (0.25/0.18), sugar (0.26), and carbs (0.26). This is of course because more steps lead to more calories needed and, well, the food options when you’re on the go are not exactly representative of a balanced diet. Protein (-0.05) is a major exception to this, so not only am I eating more fat/carbs, but also less protein at the same time. Not great.
Lesson 3: Achieve a more balanced diet, particularly when on the go
There’s also the opposite end – sleep (-0.34.) The less sleep I got the night before, the more steps I got in. This is a tricky one. I want more steps, but sleep is important too. I associate this mostly with wake-up time, if I got more hours of sleep the night before I likely woke up later, stealing some time I could’ve been making moves. So this lesson should encourage steps without losing sleep value as sleep is critical.
Lesson 4: Wake up earlier while preserving sleep time
From here, I’m going to call out specific correlations rather than showing the whole chart. Sort of a dartboard approach to this giant sheet of numbers. First is the strong correlation to drinking.
When I drink, I’m less likely to cook (-0.25.) Not super surprising. When I’m out, I’ll likely eat out. I should probably manage this moving forward to ensure it’s not a detriment. I’m also more likely to drink caffeine (0.23) a combination I should be mindful of.
The biggest concern I have with drinking is the correlation with happiness. Like I said above, this is not just because of drinking itself, it’s a lot about when I’m drinking. It’s a correlation, not a causation. I think for me, it’s more about improving my happiness when sober. In an ideal world, I should be able to find as much happiness on a sober Wednesday as a Friday on the town. That might be wishful thinking. Part of me knows this will exist in some form going forward because I drink on the weekends, but the other part of me wonders what my life looks like if this correlation is 0.
Lesson 5: Find more enjoyment during the week (preferably sober)
Phone time is another key variable to look at given its negative relationship to happiness. The most powerful correlations here are also the most confusing. Protein (0.70!) and Calcium (0.46!) have these huge correlations, more than most on this sheet, to phone time. Maybe this is because my phone use is at home and I have a lot more of these there? I do focus on protein foods at home and I love cheese. If not that, I really have no clue. Maybe Apple is manipulating me into eating babybel. I can’t say. The negative correlations of phone use are clear: Steps, whether I left home, and day quality. These are going to exist no matter what – I’m always going to use the phone more when I’m idle at home. Regardless, I can moderate this and I should.
I’ve decided to stop here with analyzing these, I think 5 lessons are enough for now. With this much data, it’s easy to overthink some things. I could sit here all day and try to analyze why I eat marginally more carbs on days I read, but that’s probably not worth it. This data is interesting, but it’s not everything. It could never capture all the little things that go into a great day or a bad one. Still – I know I started this by saying this likely won’t lead to much, but I did learn a lot. To some extent, whatever stories I’ve told myself are held to the fire on these sheets. They’re not perfect at all, and I’m taking shots in the dark with some of the causes (I still really don’t understand the Phone – Protein industrial complex) but they let me measure my efforts and learn a bit about what’s actually happening under the hood. So I’ll take these lessons and do the best I can, maybe you learned something too. Your data probably isn’t too different.
I’ll see you in 2024.
Colin