RUNN calibration in Zwift?

My RUNN used to be within about 0.5kph of my treadmill when I first got it, according to Zwift. Now it’s 1.1-1.3kph faster than the treadmill (12kph treadmill set and Zwift over 13kph through the RUNN). It feels about right, like it feels faster than 12kph, but how can I check accuracy (as in is the RUNN correct?). I have factory reset the RUNN and updated the firmware and the result is the same.

If I was to calibrate the RUNN via the Zwift calibration page in Zwift, is that calibrating the RUNN to Zwift (so should be then spot on), or would it be calibrating the RUNN to the treadmill? (Which I don’t want).


I’ll be interested how you get on with this, I’m very similar. After the sensor saying it’s happy with my calibration it then tells me I’m running about 0.5-07kph faster than the treadmill.

I want to believe it’s because I’ve really been running faster than I thought all this time and I suppose it’s a possibility as my best times outdoors are as good as my treadmill running. But I just assumed the tech wasn’t that good…

It’s probably the treadmill drifting. They are notably a) imprecise and b) unreliable in measuring speed. I’d trust a Runn 10 times before a treadmill. It’s a simple device - measures the delay between two marks with a known spacing.

If you want to make 100% sure, place a mark on the belt, measure the belt length (total length, top and bottom), run at a set speed, and have someone count the passages of the mark over a set time. You can calculate the belt speed from that. Don’t do it without someone running on the treadmill, high probability the speed will be far off without a load.

1 Like

Correction: this is actually wrong. It measures the delay between the two built-in optical sensors for each mark. You could put a single mark and it would still work; and the spacing between the marks is irrelevant (except they recommend no closer than 18 inches).

1 Like

I’ve posted much about the accuracy of the Runn.

By default it’s calibrated to 5mph so if you do a reset of the device and set your treadmill to 5mph it’ll give you some indication of how close it is.

However…dependent on the quality and power of your treadmill it could slow significantly under load. For example under load that 5mph on the treadmill could be reported as 4.5mph on the Runn and therefore Zwift.

You may also find that a calibrated Runn at 5mph reports different speeds to the treadmill as the speed increases.

This is why it is best to calibrate the Runn in game with the middle speed at the normal pace you would run constantly at. So I calibrate mine at 8mph which is my steady pace.

This assumes the treadmill is right and the Runn is off. Considering a) how the Runn measures speed vs b) how the average treadmill does, I’d venture this assumption is wrong.

I’m inclined to agree with the comments regarding the likelihood that a factory reset RUNN is more likely to be accurate compared to the actual treadmill, especially given that 12kph on the treadmill, as showing 13kph in Zwift via RUNN feels a lot more like 13kph outside.

I had a play with the Zwift calibration last night and realised it doesn’t do what I assumed (I assumed it would ignore the treadmill and calibrate RUNN to Zwift), it looks like what it will do though is calibrate Zwift to the TREADMILL speed via the input from the RUNN, which I do not want as that assumes the treadmill is correct (which it definitely is not).

What I was hoping it would do, which seems like it could easily be implemented? Is when you do the Zwift calibration, it should give you a real-time zwift speed on screen during calibration, so when it says run at 12.5kph for 40 seconds and you set your treadmill to 12.5kph zwift should be able to calculate, given the input from the RUNN, how “far” a zwifter should have ran in 40 seconds, then it can calibrate the RUNN to Zwift (and not to the treadmill), if that makes any sense at all?

Though I am yet to convince myself that’s even possible, as in theory I keep thinking you could self calibrate based on a pace bot, so if Zwift pace bot 11.9kph is on screen and you need to set your treadmill to 10.3kph to match the screen pace bot of 11.9 then you know that treadmill 10.3= zwift 11.9, but even then I’m not sure that’s a correct way of judging it as, lets say you have a faulty RUNN than reports speed 2x as fast as true, so if your belt is going 6ph and RUNN is sending 12kph to Zwift, you’d match the pace bot and therefore assume your running at 11.9/23 kph but your actually running 6kph.

Its doing my head in, If it was only 0.3-0.5kph different I’d be happy enough, but at over 1kph difference at 12kph I feel I really want to satisfy myself the RUNN is closer to real life than the TM.

Like Rober_C says, probably just have to do a belt measure & count to satisfy myself that the RUNN hasn’t gone skewy.

There is nothing to calibrate - Zwift receives speed through its connection with the Runn. It takes that number as-is, same as with, say, cadence from a sensor.