Strava stole my (estimated) power!!

[Crog Welly]

Recently switched to a QX100 from a DF gravel bike for daily commuting.

Strava uses some black box algo to estimate power output and calories burned. For the same heart rate (not speed) Strava estimated power is 30% less on the QX100. For the same heart rate!! Not speed.

In the week prior to the switch Strava estimated power used at between 1.05 and 1.21 watts per heart beat with an average of 1.12 watts per heart beat. After switching to the QX100 the range is 0.7 to 0.9 watts per beat with an average of 0.79.

The only possible data that the algo could have used to change the estimate is that the brand of bike in the bike definition on Strava was Cruzbike. The other inputs are speed (from the iPhone Strava App) and heart rate (from a Polar H7 HR strap). I have changed the brand to some other brand, a DF brand to test this hypothesis. I will keep you posted.

Anyone else notice this or something like this?

Crog

 

[jond]

Recently switched to a QX100 from a DF gravel bike for daily commuting.

Strava uses some black box algo to estimate power output and calories burned. For the same heart rate (not speed) Strava estimated power is 30% less on the QX100. For the same heart rate!! Not speed.

In the week prior to the switch Strava estimated power used at between 1.05 and 1.21 watts per heart beat with an average of 1.12 watts per heart beat. After switching to the QX100 the range is 0.7 to 0.9 watts per beat with an average of 0.79.

The only possible data that the algo could have used to change the estimate is that the brand of bike in the bike definition on Strava was Cruzbike. The other inputs are speed (from the iPhone Strava App) and heart rate (from a Polar H7 HR strap). I have changed the brand to some other brand, a DF brand to test this hypothesis. I will keep you posted.

Anyone else notice this or something like this?

Crog

Why yes I have.

It’s why I pay no heed to strava estimates. Bike weights anyone. Aero anything cda

Generally power is disseminated by watts/kg so adjust your total weight on your profile appropriately for a more reliable estimate

Heart rate is fraught with issue for estimation see dehydration see sickness see I’m excited I’m bored etc etc.

I do have two power meters which I once religiously subscribed to for many years. As they have passed like my powers I feel liberated to not know actual power lol.

If metrics are important to you then get a power meter. They have never been more affordable and have great advantages for training racing touring tootling along.

 

[Bill K]

Take a look at your uphill segments (assuming you have some) and compare your two bikes.
Strava should be much more accurate estimating low-speed climbing where aero has much less effect.
Also, it is likely that your heart rate on a recumbent for the same perceived effort is lower since you are in more of a sitting position than standing.
Or maybe you are just excited to be riding a Qx100!

 

[anotherbrian]

Recently switched to a QX100 from a DF gravel bike for daily commuting.

Strava uses some black box algo to estimate power output and calories burned. For the same heart rate (not speed) Strava estimated power is 30% less on the QX100. For the same heart rate!! Not speed.

In the week prior to the switch Strava estimated power used at between 1.05 and 1.21 watts per heart beat with an average of 1.12 watts per heart beat. After switching to the QX100 the range is 0.7 to 0.9 watts per beat with an average of 0.79.

The only possible data that the algo could have used to change the estimate is that the brand of bike in the bike definition on Strava was Cruzbike. The other inputs are speed (from the iPhone Strava App) and heart rate (from a Polar H7 HR strap). I have changed the brand to some other brand, a DF brand to test this hypothesis. I will keep you posted.

Anyone else notice this or something like this?

Crog

Strava uses the Type (Road, Mountain, TT, or Cross) of bike, the specified Weight, and the rider weight in determining the estimated power. There are large differences between bike types. Are you sure the bike type and weight were the same?

I never noticed a correlation between estimated power and heart rate on Strava when I was doing estimated vs. actual power comparisons (and their estimate was quite accurate over a 40mi ride as far as average power and calories). I would first upload the full file with power data to my main account, and then strip the power and upload it to an alternate account and compare the two. I use Road for my S40 and TT for my M5CHR.

And re-reading your post ... Power is all about speed and elevation differences and nothing about heart rate. An out of shape rider might use more calories to make the same amount of power as a fitter person, but the power required will be the same provided they're the same weight/shape as each other.

 

[Crog Welly]

Anotherbrian, I will play around with bike type, that might well be the difference. I think the boffins at Strava are pretty sophisticated. I will likely spring for a power meter to settle my mind.