Well folks, it’s November 29th, so it’s time to have a detailed look at all of our socially exciting athletic data from the entirety of calendar year 2018 to see what patterns we might be able to discern. No better time to do it then now, before we all get drunk at holiday parties!
Oh – wait, what? The year isn’t over yet?
You mean there’s still over a month remaining to pretend to be fit? Nah!
Don’t let that get in the way of a good PDF file. Nobody runs or rides (and god forbid, not swim) in December anyway.
Wait, hold on – you’re telling me the ‘2018 Year in Review’ report isn’t actually for calendar year 2018?
You say it’s for Sept 1st, 2017 till Aug 30th, 2018?
What the bloody hell kind of athletic accounting system is this? Are we sure there isn’t motor-doping involved here too?
In all seriousness, Strava’s apparent calendar confusion aside, the report is actually pretty interesting – no matter the precise timeframe that it encompasses (and yes, all of the above is actually true). The company releases the ‘year in review’ report each year, inclusive of…well…a semi-nebulous slice of time. In fact, one year I actually got sent a massive printed version of it to a hotel I was staying at. This year I just printed the first page of 37 included pages and stuck it on a random book I had sitting around (apparently I’m quoted in it, but I don’t read Finnish).
Still, I long for the good ol days of giant printed PDF files. It brings me back to working in corporate America with copy machine rooms and the like. Anyways, onto the details.
The Stats:
There’s all sorts of fascinating stats in there. You could play trivia all day long with it. 37 pages of PDF stats, most pages include 5-15 stats that you can throw-down on unsuspecting fitness friends at a moments notice. For example, Lululemon has more members in their Strava group than Zwift. Bet you didn’t see that coming:
And of course, people not in a Strava group are three times lazier than those in a group. Or, at least socially lazy. And of course – if it’s not on Strava, it didn’t happen.
But in reality, the stats do support that those in groups both spend longer durations doing the activity and also go longer distances. This isn’t a surprise, as half the time I go out with friends we get lost – whereas when travelling solo my navigation skills are impeccable. Thus naturally I’ll go longer with friends, because I’ve made the mistake of letting them navigate.
Speaking of social, there’s a number of socially interesting things in here. Some of these are items are obvious to those that are on the platform – but they’re also things that brand ambassadors, sponsored athletes, Instagram husbands and the like should be paying attention to. Most notably, pictures equals likes. This isn’t a surprise to anyone that’s attached a photo to their activity, it gets twice as many likes according to Strava – and I certainly believe it.
What is most fascinating is that 32% of all snowshoeing activities include photos. I mean – I kinda get it. If I go out and put on giant oversized shoes and flop around in the freezing cold for hours, I’d want some photographic proof posted as well.
Next, we’ve got the most used emoji by country and state. First off – I’d love to have a detailed analysis of the Belgium/France/Netherlands ‘sunny’ emoji usage. Because quite frankly, having lived/living in two of three of those countries, there’s no part of ‘sunny’ that I’d consider a frequent occurrence.
Perhaps people are making titles talking about the lack of sun for every workout?
And Vermont – what’s up with the creepy black heart emoji?
From the international consortium on how this emoji should be used:
“A heart shaded completely black. May be used to express morbidity, sorrow, or a form of dark humor.”
I mean, at least I get you North Dakota – I’d totally agree with the snow emoji as well. Though again, the key takeaway from this slide is actually that emoji’s double the number of kudos on your activity. Thus I’m adding emoji to all my uploads from here on out (plus the photo as well).
Next, some interesting stats on the most popular days for workouts:
The May and September dates definitely make sense – both time periods that globally tend to have pretty good chances of great weather, especially in the northern hemisphere.
The inclusion of July 4th in the US as the most popular day for sport is interesting given the holiday must be driving people outdoors. However, as an American, I don’t actually associate July 4th with fitness. Rather, with hot dogs and trying not to blow your fingers off. Certainly the November 23rd (Turkey Trot that year) makes sense, and maybe July 14th encourages folks to get out on the bike due to watching all that Tour de France coverage.
A quick check of the other countries reveals some similarities:
Brazil – May 1st: Public holiday
Japan– Sept 16th: Public holiday weekend
Spain – Dec 31st: Next day is public holiday
Germany – May 6th: Sunday in between two holidays, often used as bridge holidays
However, the dates for the UK and France don’t align to anything. They’re just random outliers where it must have finally been sunny outside.
On a more serious note, I thought it was interesting how Strava presented various gender differences in stats. Many slides highlight the nuanced differences to how men and women differ on uploads or other metrics. However, they never explicitly point out the gender share breakdown in percentage between men and women using the platform. Likely because it would unfortunately show that far less women use Strava than men (despite Lululemon’s stellar standings in group membership).
The above slide is the closest we get to actual hard numbers, and even that isn’t telling the whole story as it skips a large number of sport types. Still, it does tell a story. If we total up those three sports (run/ride/walk), we get:
Total women uploads: 149 million
Total men uploads: 636 million
In other words: 77% men, and 23% women.
That said – the differences in speed, or duration aren’t actually that much different between men and women. Only 5-8 minutes shorter for rides, and slightly slower for runs
However, US women do actually tend to run for a longer duration than US men. Whereas the inverse is true more globally. A different slide helps paint that picture that the average distances are slightly less for women than men.
We do also see that race participation in the US is growing slightly faster for women than men, and if we look at year over year stats, then women have nearly doubled that of the men from a growth perspective. This slide also shows that for running, people are clearly favoring Sunday over Saturday for their long runs.
There are really so many ‘WTH?!?!’ stats in here though. For example, the bike commuting stat is without question the most interesting to me. It notes that the longest bike commutes are from those in Mississippi. I can guarantee you that if you asked me to name the top commuting state, heck, the top 25 states for longest bike commutes – I’d never have guessed Mississippi. This is furthered by the fact that the Bike League of America ranks the state towards the bottom – 40th of 50 states for bike infrastructure. Same goes for Oklahoma – 2nd on the Strava list, but ranked 46th in the bike infrastructure list.
I’d love to know what types of routes those are. Are they dedicated bike paths? Are they on roads just on shoulders? I mean – going 14 miles (22km) is a heck of a long way for a one-way commute – well above the global average:
But it also gets into some of the nuances of how people use the platform. For example – one of (if not the most) bike friendly nations on earth – the Netherlands – doesn’t rank towards the tops of the ride commuting list. But that’s logical, because most people living here (including myself) don’t bother to start the Strava app or record commuting rides. For example – I haven’t uploaded a single commute to the office since living here in Amsterdam. Despite commuting 7 days a week to/from the office (yes, even weekends).
In any case, I’ll leave you with the most important stat: Beer
Beer wins for runners, and coffee wins for cyclists. At least in terms of titles. Approvingly though, #3 behind both of those is cake. The Girl would be happy.
With that – thanks for reading. And again, you can download the entire PDF report here.
FOUND THIS POST USEFUL? SUPPORT THE SITE!
Hopefully, you found this post useful. The website is really a labor of love, so please consider becoming a DC RAINMAKER Supporter. This gets you an ad-free experience, and access to our (mostly) bi-monthly behind-the-scenes video series of “Shed Talkin’”.
Support DCRainMaker - Shop on Amazon
Otherwise, perhaps consider using the below link if shopping on Amazon. As an Amazon Associate, I earn from qualifying purchases. It doesn’t cost you anything extra, but your purchases help support this website a lot. It could simply be buying toilet paper, or this pizza oven we use and love.
Have you considered that maybe Mississippi leads the way because it doesn’t have “bike facilities”? (Or to put it another way, every not-limited-access road in the US is a “bicycle facility”.) Around here, the worst of the “bike facilities” is a sidewalk, riddled with driveways and with stop signs in your face at every intersection, with signs on it. I would never dream of using it, especially since there’s a perfectly good road alongside the entire way (and Pennsylvania got rid of its “side path” law almost 20 years ago).
As for pictures on snow, I’ve never been on snowshoes but cross country skiing it’s certainly possible (although nowadays you have to prepare with etip gloves) to take a picture while moving. I imagine that’s even easier for a snowshoer. On the bike it’s far more difficult, and I always stop in order to take pictures, so I don’t usually do it. (Back in the days when I used a dedicated camera with buttons I sometimes took pictures while moving, but they only occasionally turned out well.)
“Hey Siri, take a photo” – followed by volume Down Button – no etip gloves needed. That’s how I get my snowshoe running photos. However up here in Canada during winter, as soon as you pull out your iPhone, count on 1 picture at the most before it shuts down in the cold!
Exactly. Kind of the opposite of the “Netherlands dynamic” – if there, commuting is so part of the culture that people don’t even bother to record an activity, then perhaps for those in Mississippi who went “out of their way” to commute will want it captured in Strava.
Since I don’t use Siri on phone, watch, AppleTV, or HomePod, I didn’t even think of that. For that matter, my VIRB 360 understands “Hey Garmin” commands.
Your picture shows the other reason there might be more photos from skiers and snowshoers, it’s just so damned picturesque in the winter, especially when the sun is out :-)
My thought is that the stat is “average commute length” which isn’t based on how many people are actually commuting – so there could literally be 1 die hard guy somewhere in Mississippi doing a 14 mile commute every day.
Huh. I’ve gotta say, I never thought to take a photo with Siri, but that’s brilliant. Just tried it out on the bike. Perfect with gloves!
Yeah, and there’s probably some skewing of Strava users too. I would guess that there are fewer users per capita in more rural states, which probably means a higher percentage are more serious athletes, i.e., the ones more likely to make a longer bike commute.
Yep – I love winter running/snowshoe running. Surprising however how quickly you can overheat when snowshoeing even in -20C or colder temps.
On an Android the user can say “Okay, Google, take a picture.” and it will open the camera app then you would say “Shoot” or “Capture”, etc to take a photo. It will also do a selfie in the same way. I believe the voice option must be turned on in Camera for this to work.
On Android: Double Click the power button and the camera automatically opens.
Click Volume Button and picture is taken.
@Rob: But you’re still stuck with an Android phone
Well the bike commute distance per state is an average. So if there’s only one person who did a single bike commute that totaled 14 miles that’s all that’s needed.
>Next, we’ve got the most used emoji by country and state. First off – I’d love to have a detailed analysis of the
>Belgium/France/Netherlands ‘sunny’ emoji usage. Because quite frankly, having lived/living in two of three of
>those countries, there’s no part of ‘sunny’ that I’d consider a frequent occurrence.
Perhaps because you haven’t been living in Belgium yet? ;-)
Seriously, the summer 2018 has been the wamiest on record in Belgium, and also one of the brightest and dryest ever. And given the weather we had in the spring and automn (so far), the whole year is going to fare well in that ranking. I doubt it has been much different in the Netherlands or France.
So… About that Black Heart in Vermont… First, we’re a low population state so small things can have large effects in general. Second, things come to Vermont late (we were one of the last states to get television, for example) so we may be lagging behind the rest of the world in emoji usage. And third, we consider ourselves to be fairly reserved… You know, taciturn. It’s possible that, as a small population of people who may not be heavy emoji users, that result may be heavily influenced by something like the Richard Tom memorial ride – an activity for which the black heart emoji might make sense.
The reason why Spain’s most active date is December 31st is not because January 1st is public holiday, but because there are many popular races that day in many cities across the country. It’s “San Silvestre”.
“San Silvestre Vallecana” is the most famous of all of them, gathering more than 40.000 runners in Madrid.
I live in Mississippi, and we do not have the longest average bike commute. While there is some bike commuting, especially along HWY 90 on the Gulf Cost, that has to be a mistake. What we do have is the 444 mile long Natchez Trace which runs from Natchez, MS to Nashville, TN. It is a very popular road on which to bike long distances. You can see people biking and racing on the Natchez Trace nearly every weekend. Only a portion of the Trace has a dedicated walk/run/bike path.
Sorry Ray.. but this is just 100% WRONG!
“However, the dates for the UK and France don’t align to anything. They’re just random outliers where it must have finally been sunny outside.”
June 26th is *my birthday* !! :)
Hey, it’s mine too! We must be kind of a big deal ;-)
And amazingly.. My daughter’s too! :)
Looking into the file I was not surprised that the most popular running segment in the US is in Brooklyn’s Prospect Park. There are many races in the park, and it’s full of people running on Sundays.
One thing I was looking for in the file, but didn’t see – lists of the gear people use. I.e. the devices used to upload to strava, as well as the shoes/bikes people log with their workouts.
Lululemon having the largest group isn’t surprising. They host several Strava challenges throughout the year (like the 40/80 challenge in January and the Ghost Race in October) where you can earn free or discounted gear. People love free stuff!
Tulsa has over 80 miles of bike/run trails in the metro area and there are a number of people that will do 10-15 mile rides in to work using the trails. They are generally safer than riding the streets early in the morning or night due to the higher volume of traffic and lack of sunlight.
link to incog.org
Missing statistics: Percentage of users preferring metric or imperial units ;-)
Funny that Strava claims CO2 savings by all the cycling and running commutes. Of course- it depends on what foods you eat and what kind of car you drive- but overall, vehicles produce less carbon emissions per mile than humans exercising. Your 5 mi running commute will burn, let’s say 500 Kcal. If you eat beef to replace those calories, that’s 5Kg of CO2 equivalent emissions. A mid-sized car will output about 200g per mile of CO2, or only 1Kg of CO2 emissions for the same commute. Dairy protein is about 5 times more efficient than beef, so then you’re on par with a car.
Granted- cycling is a lot more efficient. And being a vegetarian is more efficient still. But- carpool with 4 people in a Prius, and that’s hard to beat, even with aero wheels.
Morey000:
I was aware there was a carbon cost to cycling / running, but I hadn’t realized it was so significant. Than you for highlighting it. I did a little research and I think your estimates are a little high however. Look at the Harvard piece at the link below. It looks like the “average” american diet produces 2.5g CO2 per kcal.
link to keith.seas.harvard.edu
Ah yes, the all beef diet fallacy. Does anyone still fall for this junk?
That sounds about right. Using protein sources exaggerates the effect. For 2.5g/Kcal which sounds about right, that would put the 5mi jog at ~ 500kCal x 2.5g/Kcal or 1.2Kg of CO2… which is still higher than my mid-sized sedan which would put out 1.0Kg. But, I’m mostly a vegetarian…. and then we should subtract your basal metabolic rate for that same time… but you could drive a more fuel efficient car… etc etc. The point remains (Despite Mark not ‘falling for this junk’) that there are a lot of reasons to cycle, run or walk to work- but saving CO2 isn’t really a good one. Admittedly- cycling is 2 to 5 times more efficient than running. But then again- so is carpooling in a Prius.
What about tones of CO2 generated to make a car in the first place?
Bullshit!
…and unfortunately it has been spread to public at TEDx also…
If plants are making polysacharides and O2 from CO2 and H2O, it doesn’t mean that human beings are able to degrade it to CO2 and H2O.
For example:
During glycolysis, when pyruvate is made from glucose, only 1 molecule of CO2 is produced.
During Krebs cycle, when isocitrate is made from citrate and isocitrate is transformed to ketoglutarate, only 2 molecules of CO2 are being produced.
Here’s the key problem with that argument.
Where does the carbon, in that carbon dioxide that’s produced (either by driving, or by cycling), come from?
In the context of driving, it comes from the petroleum products that are burnt (diesel, petrol, LPG, whatever). Those products ultimately are dug up from the ground: the carbon was sequestered underground, away from the ecosystem, for thousands of years; digging it up and burning it releases it into the ecosystem, increasing the amount of carbon in the system.
In the context of cycling, it comes from the food we eat. Where does that food get the carbon from? The atmosphere. Photosynthesis takes in carbon dioxide from the air, and water from the ground (in some manner or another), and turns it into carbohydrate molecules – sugars. Eating that food and using it as fuel merely returns carbon back to the atmosphere from whence it originally came. The net change in carbon levels in the overall ecosystem is zero.
You could have a car that emits a single gram of carbon dioxide from petroleum products over a distance of a thousand kilometres, and it would still be a bigger problem from a climate change perspective than the cyclist who puts out however many kilograms of carbon dioxide over that distance – because that car is ADDING to the carbon load, from a source that was sequestered away, instead of using carbon that was already in the system.
Biofuels – creating diesel from vegetable oils, for example, or synthesising our hydrocarbon chains from carbon dioxide taken from the atmosphere – are fine (at least from that perspective, although they do carry their own issues.) But that is not the vast majority of fuels.
It isn’t carbon dioxide _per se_ that’s the problem, in other words. It’s adding carbon dioxide that is created by burning carbon that was sequestered away.
Yes, I have problem with your argument…”Eating that food and using it as fuel merely returns carbon back to the atmosphere from whence it originally came.”
First of all, we must distinguish between Carbon, carbon-dioxide and carbohydrates.
In anorganic chemistry is possible to produce a perfect burning with sufficient oxygen supply. But it is not a case in everyday life.
Just to compare diesel with benzin:
Diesel contains long chain carbo substances with a few double bonds. If you put it to sunlight, it won’t evaporate. If you burn it, you will get a small amount of CO2 and lot of carcinogenic ash.
Benzin contains short chain carbo substances, easily evaporating and if you burn it, you will get lot of CO2 and a very few/no ash.
In human body food goes under metabolism and it’s degraded products we shit out. That’s why we have digestive system.
To that degradation process we need O2, so controlled oxidation could take place. The CO2 is a result of that oxidation and we breath it out.
That is how life goes. What is being taught nowadays at chemistry?
I’m still not clear on why Strava doesn’t add p-values, error bars, or even the word “Statistics” to this data. Perhaps it’s because people like me don’t log our Chocolate Chip cookies after a workout and they’d have to throw out a lot of data points. With all that math out of the way it can be enjoyed as infotainment.
No need need for error bars when not sampling Etc
As someone who lived in Oklahoma damn near most of my life, I can attest: there’s a weirdly awesome bike culture there, especially in Stillwater and Tulsa (and to a lesser extent OKC). Check out the gravel scene or Tulsa Tough for a little snapshot into it. For sure, there’s almost no bike infrastructure – most of us just ride on the sides of roads. (I now live in DE, where people apologize all the time for the aggressive drivers here, which just makes me laugh because biking here is SO much easier, but I digress…) But couple that with the fact that in that part of the country, things are just so much more spread out, and I’m not surprised at all to that the average commute is that long. When I commuted I biked 5-6 miles just to campus one way. Did not include getting around the rest of town.
I agree with others comments about the U.S. State statistics being skewed by a few enthusiasts (since it’s an average).
And, for rural states largely lacking bike facilities, or where urban sprawl means a greater distance between work and home, bike commutes are generally longer and more circuitous.
My Kansas commute is 12 miles each way, and that’s among the shortest of my bike commuting work peers.
1 retired co-worker did 28 miles each way, year round, for the duration of his career. All on a relatively high traffic highway shoulder, where he was struck by highway-speed motorists 3 times in his career, and managed to survive.
Also, Strava here is largely used by the Ironman or ultra crowd.
We’re home to Garmin, and very few people have ever heard of KOM’s, so you’re really only getting the super committed fitness folks, or people who really want to show off their net mileage.
Ray, that random book is coming in english as well! You will be the first one to know!
Haha! I started trying to find what page I was on, but after like 5 pages, I was like…umm…this might take a while!
Great, looking forward to the English version (as long as it has equal numbers of pictures).
Hi Ray… I’ve posted in the Strava comments area a couple of times… but no Joy… would it be possible for Strava or a third part (i.e. you) to look at the strava data and segment it by decade of life or age and gender and publish real world power data based on age and gender? Looking for FTP, peak power, or basically reproducing the Coggan’s chart by age and gender. This is amazingly hard information to get.
“Nobody runs or rides (and god forbid, not swim) in December anyway.”
Plenty of people swim in December. You’d almost have to be crazy to _not_ swim in December.
Now, June, on the other hand… _that’s_ insane.
Signed with love,
Australia.
Ha ha, nice one :)
The 31st December date in Spain is probably due to the “San Silvestre” popular races which are organised all over the country on that day. They have become part of the countries Christmas celebrations.
It’s interesting to know that I’m faster than average at running and cycling but the duration is lower than average across both sports.
I wonder if that’s a consistent theme across my life? Perhaps I should talk to my wife.
May 6th was the day of the wings for life world run. Not only does that explain why the date was up there in germany (the event in munich had 10.000 runners alone) but it also explains why that date was #1 world wide. so proud to be a part of that :)
You’re such a dork ??☕☕☕??????????…
In Spain there are quite a few New Year’s Eve races on 31/12, so that’ll be why you see peak activity in Spain on that day – here’s Madrid’s… link to sansilvestrevallecana.com
In looking at the state specific data… one area of possible error is that Strava seems to significantly under estimate the calories burned when riding dirt or gravel. Given the time and perceived effort when I ride both (havent used a power meter comparison yet), if you are riding in a state or with a preference with Gravel mostly, think the calorie burn and power used will be underestimated
Totally agree.. I rode up in the North GA mtns on a challenging route.. last Sunday..37 mi, 4700+ ft of climbing on soft/damp gravel. Strava shows I burned 1579 Calories. No way. Probably double given the biggest climb is 6 mi with 1300 ft of climbing.
Read this :
link to en.wikipedia.org
Then play with the data on strava :-p
Totally agree.. I rode up in the North GA mtns on a challenging route.. last Sunday..37 mi, 4700+ ft of climbing on soft/damp gravel. Strava shows I burned 1579 Calories. No way. Probably double given the biggest climb is 6 mi with 1300 ft of climbing.
@DCRainmaker:
Kudos on the band! I also got the “Strava employee” edition for my Vantage. ;-)
“Nobody runs (…) in December anyway”
Funny you say that. Each year I resolve to run through December as well, but as that time of year approaches I always catch a nasty cold (as a teacher I cannot really help but get sick once a year at least) that sidelines me for at least a week. Seeing all those numbers drop on TP, GCM, Runalyze depresses me so much that I call December off altogether, and usually hit the roads again come Jan. 1st or 2nd.
Oh my, if only I had joined a Strava club (consult Groucho Marx for reasons I didn’t) I could have done 1746 rides this calendar year, and there’s still a month to go. ?
Hi! On slide 36 – Run totals racing and fastest days I don’t understand the race participation growth figures. How is Men’s different from year over year?
Thanks for all the great content Ray!
Zlatko
The World has changed, Bethlehem is no exemption either…
July 14th in the U.S. was day one of the STP (Seattle to Portland) ride, ~10,000 riders take part, that probably has something to do with the numbers being so high that day.
The watch on the photo with orange strap is …?
Polar Vantage M with their orange strap.
I think the Dutch Sunny emoji usage can be explained very easily:
This year has known a very long sunny, dry, period, and that being abnormal makes people post emoji’s about it.
I wouldn’t post a rainy emoji as soon as I would a sunny one, since I think it’s normal for the weather to be dreary here.
In the warmer months I typically use my bike commute as a commute/training ride. Straight shot into Boston is just over 9 miles to my office, but it’s a less desirable route (with significantly more lights and traffic). Long route is 12.5 miles one way mostly on protected paths or redesigned bike lanes (and as an added bonus mostly along the ocean).
‘However, the dates for the UK and France don’t align to anything. They’re just random outliers where it must have finally been sunny outside.’
Probably people starting their New Year Resolution and joining Parkrun for the first time. Many venues recording record attendance that weekend.
I love Strava! Sometimes the I find the data more rewarding than the actual activity. Strange, I know!