A Year Of My Heart

A year ago, I decided to get all analytic on my training. Mainly I just like tech and pretty representations of data. So I bought a heart rate sensor. And now it’s been a year of me using it almost every time I train. Which means I can look at a year in the life of Frances training, with all the … whatever that reveals.

What does it reveal, Frances?

Well, other Frances. I trained 156 times — that I recorded, let’s say 170 because I pretty much did not train without it unless I forgot either sensor or phone. For a total of 190 hours — there’d be a few more in that for the times my phone battery died. For a measly distance of 1481 kilometres — of actual training rides, not including cross-town, Kreuzberg-Wedding type stuff, so maybe double that at least, no wonder I spend so much on my bike and it feels like it’s constantly in need of repair. Hey, just like me! (Wow, there’s a realisation, right there.) About 1/3 of that was ballet, another third cycling (mostly road at the moment, but some cyclocross), 1/6 bouldering, and the remaining 1/6th a mix of yoga and core training.

Oh, and supposedly I burned around 121,000 calories, which is about 60 days of eating 2000 calories a day. I’m not really convinced about this. I think it’s more of an imaginary number, and not the mathematical kind.

What else? Speed, both average and top are derived from iPhone GPS. I’m not sure how much dispersion there is in this, but I suspect it can easily be 5km/h or more in either direction. My next gear purchase (after … umm … new brakes and probably new rear derailleur pulley wheels) is a speed/cadence sensor — which probably means also a proper cycling head unit instead of phone …

I seem to unintentionally train in 9-10 week blocks, then give up in despair for a couple of weeks, then, like a goldfish circling its bowl, forget all that and get right back into it. Knowing that this might be my natural rhythm though, it could make sense to train in 9 week blocks with a week off, if for nothing else than keeping my enthusiasm. Also I doubt I’ve been training like that this year, my rhythm’s all over the place.

My maximum heart rate seems to be constant around 190 (excluding the huge jumps into the 200s that were either the battery going flat, the sensor getting jostled, or actual random heart weirdness from having stupid fun training in -10º weather). I dunno, I have no context or expertise for reading anything into these figures, other than I seem to like training if it involves a degree of discomfort and some suffering — which I didn’t need a heart rate sensor to tell me.

So, a year of data. What to do with it? No idea! Will I keep using it? For now, yes. It’s become automatic to put it on. I don’t really use it during training, though I’d use it for cycling if I could find an iPhone mount that could hold my ancient 4S. But mostly I do it on feel, and that corresponds pretty closely to the various heart rate zones. I do do regular post-training gawks, to compare how I felt with actual data — and knowing that data across sessions gives me a bit of a feeling for where I’m at on a particular day or week. And one other thing: I train a lot less than I think.

Worth it for seeing a year of training all pretty like that? Yup!

Polar Flow and H7 Heart Rate Sensor — One Year Weekly Training Report
Polar Flow and H7 Heart Rate Sensor — One Year Weekly Training Report
Polar Flow and H7 Heart Rate Sensor — One Year Daily Training Report
Polar Flow and H7 Heart Rate Sensor — One Year Daily Training Report

Ballet & Tech (A First Attempt)

Me (on and off for the last couple of years): “It would be awesome to have a power meter or something so I can go all data on my training…”


Has any dancer ever measured a performance with a fitbit or pedometer? How many steps? How far do they dance? PLEASE will someone do this?

Me (in Jo Siska’s ballet class on Wednesday): “OMG Jo! Look! Data!”

Inaccurate data. But that’s what this is, a test of how to get meaningful and accurate(-ish) data on what goes on when I’m dancing.

When I was living in Wedding, part of my training routine was morning cyclocross rides in the forest around Flughafen Tegel. Last year when I inherited an (old, 4s) iPhone and stuck Trails app on it, I started to see what the intangible feeling of each ride represented. A couple of things were missing though, one of which I finally prodded myself to buy this week – a Polar H7 heart rate sensor (yeah, I got the pink strap). The other is one of those crazy expenses I’m unlikely to throw euros at unless I have around four thousand of them spare for a new bike: a power meter.

Power meters tend to be the province of bike crank arms, pedals, or hubs and cost about double what normal people spend on a whole bike. And none of them are objects you can take into a dance studio. Slightly getting there is the rpm2 shoe insert power meter, still no good for dance though. Which leaves the very new Stryd – and very cheap, not much more than a Fitbit (which I’ll get to later), and about the same size as the H7 – a power meter for runners.

Before all that, Wednesday. In the studio with my heart sensor on and my iPhone beside the barre, cos it uses Bluetooth to sync. That’s several problems right there. First, doing ballet (or generally dance) training with an iPhone lodged somewhere is not so practical, which means a pedometer is going to count exactly zero steps. Second, Bluetooth is possessive, it likes quasi-line-of-sight and proximity. Bouncing around ten meters down the studio with heart monitor facing away from it is going to generate some highly improvised heart rate info. If, for the sake of science, I slip my iPhone into my trackie pocket, I’ll get pedometer info, but any GPS-based data capture (speed, distance, location) is comically useless, having an accuracy of greater than 4 meters. I was dumping my heart info into Trails, which is a fine app for cycling training, and much of the time it had my location not even in the same building, plus my altitude changed by 24 metres.

Thursday on my morning training ride around Tempelhoferfeld, I used both Trails and Polar’s Polar Beat. The data resolution of both is pretty good, Polar Beat is more fine-grained, and neither had a problem with my phone being in the back pocket of my jersey. I’ve been doing enough cycling with data recording to know what looks right.

Which leads me to Fitbit, cos my flatmate has one. It stores the data locally so no need for a live Bluetooth connection. It does heart rate, pedometer, a bunch of other useful garbage, makes pretty data, syncs to phone, laptop, or to fitbit.com, and looks like a dainty watch strap.

So, Friday, ballet again. This time with a Fitbit and my H7 going to Polar Beat.

I’m siding with Fitbit when they say their data accuracy decreases outside fairly limited activities: both heart monitor and step counter are dependant on arms not windmilling for acquisition of useful data. Perhaps it requires repeated use to find the best spot on my wrist, but compared to the H7, Fitbit reported my average heart rate at ~20bpm less – I stuck fingers to neck and what the H7 shows is a good match. As for steps – and ignoring the first 18 minutes or so where I have no idea who it thought I was – it gave around 250 for the entire 40 minutes of barre, and 2200 for the class; obviously not counting a pas de bourée as three steps.

The H7 doesn’t do step counting – unless you pair it with their walnut-sized Stride Sensor somehow affixed to your foot. Its heart data though is magical. You can see every exercise through the class mirrored in my increased heart rate, and check out the centre adage starting at 40 minutes, where the curve is almost identical for both times, and the arc through the entire class, building intensity in small stages at the barre until peaking through the centre into longer and longer periods of maximum effort, before révérance-ing out. I can also look at sections, so if I select just the centre, then my average heart rate goes up to 167 and only once drops below 120. Lots of good data you can do stuff with. (And I can even assign training to Ballet, with a fancy Olympic-looking arabesque!)

But what about power? Or other stuff? Stryd for the power (and heart rate), and RunScribe for everything else? Would they even handle dancing? RunScribe would be awesome for visualising the mechanics of dancing, g-force, velocity, ground contact time, pronation – if it could handle the foot chaos. And then what to do with all this information? If it’s all just for a bit of woohoo! then Fitbit and its social network gamification of sleeping is fine. But if it’s for the purpose of improving performance, technique, being more diligent in how you train, that’s a whole other thing.

Ballet training with Polar H7 Heart Rate Sensor
Ballet training with Polar H7 Heart Rate Sensor
Ballet training with Fitbit Charge HR
Ballet training with Fitbit Charge HR

On fornication and genetics in the Breedster age

Fifty-two days ago an invitation to join Breedster , and became a community founder – one of the perks of compulsively subscribing to any site that says ‘join our mailing list’. Since then I’ve had 58 partners, 26 children (mostly dead now) and then April 21 happened, the year zero of the grid.

The net-art terrorists behind breedster released a ‘Zero Content Symposium’ today, On Fornication and Genetics in the Breedster Age, which brings the whole thing together in a bunch of pretty pictures, so we can finally see what we’ve gotten ourselves involved in, and what we’ve become.

This symposion is the retroactive manifesto of Breedster’s enterprise: it untangles theories, tactics and dissimulations to establish the desires of Breedster’s collective unconscious as realities in the Grid.

How to write a manifesto in an age disgusted with them? The fatal weakness of manifestos is their inherent lack of evidence. Breedster’s problem is the opposite: it is a mountain range of evidence without manifesto.

In terms of structure, this symposion is a simulacrum of Breedster’s Grid: a collection of blocks whose proximity and juxtaposition reinforce their separate meanings.

Breedster is a theatre of progress. Its protagonists are the “exterminating principles which, with constantly augmenting force, would never cease to act.” Its plot is: barbarism giving way to refinement.

From these givens, its future can be extrapolated forever: since the exterminating principles never cease to act, it follows that what is refinement one moment will be barbarism the next.

Therefore, the performance can never end or even progress in the conventional sense of dramatic plotting; it can only be the cyclic restatement of a single theme: creation and destruction irrevocably interlocked, endlessly reenacted. The only suspense in the spectacle comes from the constantly escalating intensity of the performance.

Topics include The copulogram as a means of visualising the social network: We are not our maps, The toroidal universe: New data, new debate, Meaningfulness and motivation: Microcommunities and mobs, Paranoia, hubris and hatred in the post 4/21 era, New ~insights in the epidemic potential of pathogenic causative agents in heterogeous communities through outbreak investigation by fluorescence spectroscopy, and well worth a look in understanding how it all came together, Questions and Answers, Adjourn

Breedster Grid
Breedster Grid
Breedster Copulogram
Breedster Copulogram

Google – turning news into art

Google is becoming a bit like an Iain M. Banks’ Culture Mind, sucking in information at such an incomprehensible rate, and adding to a database of unsearchable proportions. Perfect terrain for art to make sense out of. There have been lots of net-art projects recently that use Google’s ecosystem as the data source, and several of them focus exclusively on the news feeds where relevance is the poor second child to quantity.

newsmap won the ars electronica International Competition for CyberArts this year.

Newsmap is an application that depicts the permanently changing map of Google news. A visualization algorithm produces a visual representation of the enormous quantity of information gathered by Google. Beyond that, the graphic depiction enables users to get a quick overview of breaking international news and, in doing so, provides background information on its distribution and how it is processed further, and sheds light on the mechanisms of news production on a global level.

TooGle is similar but with pictures.

TooGle is a parser over the net focused on the most famous search engine of the net. When it runs, it gets the latest news of the biggest aggregator of news and it uses each word of the title to retrieve an images sequence from the largest container of images over the net. The result will be a short movie that use the words of the news, the images gotten and the log generated during the parsing.
The animated sequence is not time based; due random algorithm, the original frames will become remixed even more.
After a bit of time the result will become an unreadable sequence, an unreadable news… there will be news no more.

googlehouse makes information overload a house you can come home to. The house real-time construction with walls, rooms, drawn from images sucked from the google image search engine.