Over-Instrumented Running: What I Learned From Doing Too Much

I’m Thomas Blomseth Christiansen and I’m talking about doing some intensive run training. So in 2004, I ran aCopenhagen marathon and it was a grueling experience. I started out fast and then I finished painfully slow and it was quiet an okay finishing time. But I did a common mistake, even among the elite, my friend Ejnar told me, and he’s a pro-coach which I did a positive split. I ran too fast in the beginning and then I used kind of my energy and I hit the wall. And instead you should do a negative split, which is starting out slow and finishing fast. And some of the world records in running the long distance runs, they have said use the negative split, but it’s a difficult thing to do for even the athletes.

So this idea about negative split was marinating for some years, and in 2015 during the summer, I started experimenting. I did these split plans, and I had to code scripts myself and it took like two to three hours to get ready for a specific run. And what I did was I took post-it notes, and then I wrote split plans on the notes and attached them with rubber bands to my wrist so I could follow the split plan while I was running. Of course that was a kind of cumbersome process.

I learned a lot from it, so in 2016, I completed an app. Actually it’s in the app store for general release, but I made it as a tool for myself so I can just do the split plan in a few seconds and actually then go running. And then integration, and even so instead of having to write the split plan on post-it notes and put it to my wrist with rubber bands, I could send the split plans to a smartwatch so I would use that while I was out running.

And I just want to take you a step back and talk about the first principles of running, which is if you look at running from like a physics point of view there are two things to pay attention to. One is step rate, that’s how fast you move your feet and the other one is the step length. And then when you multiply those two, you get your speed right. So that’s running from a physics point of view.

So if you want to do these negatives, but you really have to control those two factors, and I had gained some proficiency in controlling like the split times. But then earlier this year I thought that I wanted to dig deeper into this. I wanted to be more better at controlling it, so I thought well I need to control the cadence and I am a musician to so I’m used to using a metronome. So I thought I could use a metronome for controlling the cadence, so I started doing experiments with a metronome while I was running. And when you do negative splits you have to change the cadence while you are running.

So what you see is my first one with the metronome, so we have the top one is slightly different cadence changing during the run. And then you have the other line, the actual data from the run. And the blue line is if I just ran, even pace the red line is the negatives and the green line is in between. So I did the negative split but I did it a bit faster than intended. Okay, what do you do then when you can’t control your process? You added some tape.

So I have the smartwatch, for the split plan, I had a Fitbit Blaze for the steps, I had an iPhone five for the metronome, iPod Touch for doing a sound recording because I wanted to get the steps for each kilometer.

So what you will hear now is what I was doing which was recording sound, and then doing a callout every kilometer. So I would shout into the iPod Touch to be able to capture the data. And then there was something strange going on, because if you look at my plan the steps per kilometer and then the Fitbit Blaze told me there was a few steps at the end and didn’t add up.

And this is an actual plan a week later, and you see that I actually ran the second last kilometer fastest, that’s a negative split and I was actually pretty close. But what was going on in those steps?

So I have on the Nexus 5 I had like a pedometer to so I was looking at the Fitbit data and the pedometer for each kilometer, so I was shouting at each kilometer, I was shouting these numbers and I found there was a difference. And then you know, this doesn’t add up at all, so like the Fitbit and the pedometer they have totally different reporting. And the same thing there are too few steps at the end and that was why I was having the highest cadence. So I think there was something going on with the step counting.

When I was running at a high cadence run of about 180 to 200, the step counting algorithm was counting my steps, so what do you do? You add some more instrumentation.

So I added some straps, and I added another watch, which also does steps and distance to get some more data, you know, when there is some uncertainty you want more data. And then I also made a home-made smart phone app to collect even more data, so I can hook into the sensors and get the real-time data. And also being able to control the metronome at a distance, because together with my friend Ejnar, the pro-coach, we set up a mission control where we could remote control a run.

So at the Copenhagen Marathon we had a very experienced female runner, running with this power pack and this in a back pack. So we had this 60,000 data points from her backpack during that marathon from these different devices. And Ejnar was in mission control while I was out running, where we did a 4K where I hit the planned time within six seconds.

So this is a 15 km course around around my parents cabin that I’ve been running a number of times now, running the same split plan. So you see the split plan and I have the timing, but I also have this plan step length, and then I have the plan cadence. And these are some numbers on my on-time performance.

So it seems like I’m improving, and you see the last one, I’m six seconds off 15 kilometers, which is not .4 seconds off per kilometer.

So what did I learn? So step counting it’s not really step counting. These algorithms have their own lives and a cadence of about 210, I had to move my feet really fast, that’s almost too fast. Controlling the step length is kind of hard. You have to really work on it and approach a steep slope is that you keep your cadence but then you shorten your steps, because your heart rate doesn’t go up to fast.

And then also like you can actually work with other factors in running than just the speed. And there might be one thing that I’m better at now than Kipchogi, which is controlling the pace and doing a negative split.

And then I have this quantitative understanding of my running now, which made it possible for me to have much better way of defining new challenges. So I’m looking into potentially a 90 minute half marathon, which is a crazy threshold. But now I know something about step place and the cadence, and what works for me, so I have some balance to work with.

So that was a crazy running time, thank you.