Sharing an interesting TT experiment!

I wanted to test my ability to exert even effort by using my gears and keeping an eye on heart rate – almost like one would do on the road if you do not have a power meter fitted to your bike. (Note that comparison cannot be made as reliable on the road as it can be done indoors!)

  • Ride A: I rode London Loop as normal, using my gears as I would on the real road, standing up at slower cadence on the steeper inclines, trying for faster descents in high gearing, etc
  • Ride B: I rode the same London Loop but in “work-out” mode. I picked the one-hour flat ride (without ERG mode), settling in a gear somewhere in the middle of the range at comfortable spinning speed and at a power output level that I felt I could maintain at moderate effort and roughly “zone 2” heart rate

Comparing the two rides: total time taken to complete the London Loop did not differ significantly, but I am planning on repeating the experiment on other routes. The sample is rather small of course, so similar data from other Zwifters would be interesting to hear!

Wouldn’t be nice if one could also get the recorded average Watts from Swift? That would make the results/conclusions just so much more statistically significant!!

??? Not sure what you are trying to conclude with your “test”.

Average watts is easy to get if you use Strava, TrainingPeaks or any other site.

My dear Paul,

It is not clear to me what you are up to – in this and my previous posting.

If you consider yourself a keen gadgets nerd, then please go ahead and enjoy that elsewhere. I am a serious cyclist – hence training indoors on Zwift, which offers me excellent value for money.

My postings are technical and without hidden agenda. I am in pursuit of user-friendliness of software for practical applications to cycling.

So, if you are unable to make a constructive contribution, please refrain from offering negative critique.


Not sure what your test was trying to prove since you did not state that in your post. There was no negativity in my response at all, just questioning what you were trying to prove. 

Not sure why you are attacking my with such a condescending reply. I do most of my training on an indoor trainer on Zwift and train with both heart rate and power. 

for some reason you seem to have a chip on your shoulder. 

You may take the win .

I back off.

I see you edited I see you edited your original post after I called you out on it. Your test seems a little flawed in the way you did it. 

Wow dude you have somewhat of an attitude when someone calls you out on your mistakes.

At no time was I trying to win anything I just wanted you to be more clear on what you were trying to conclude. 

Training indoor with steady state power and with heart rate is very important to me since I am training for triathlon and must monitor both so I have enough for the run. 

I didn’t mean to piss you off, that was not my intent. 

I too am confused, you did two rides. But what was the difference? What hypothesis were you attempting to prove / disprove? What was the purpose of the test?

Thanks for your interest Mark. I’ll try to explain again. (BTW I have not edited my original post)


  • It is wide accepted that in time trial (or triathlon) rides we must minimise burning matches (i.e. we must endeavour to keep the power output constant).
  • We try to minimise power variance by changing gears when the gradient/wind requires that, and by keeping an eye on heart rate (assuming no power meter on the bike).
  • But we are not sure to which extent we are really successful in keeping effort/power output constant (e.g. on steep hills we tend to go harder than on steep descents, our cadence varies, position changes between seated and standing …. etc.)   

Hypothesis: TT cyclists are unable to keep power output constant by using gears

Experiment: Ride the same route twice on Zwift. First ride normal by using gears with varying cadence and positions to compensate for varying resistance caused by hills and descents, and try to keep effort constant (read “heart rate”). Second ride on the same route but by using one gear and constant cadence – at the same effort (read “heart rate”). This is done in Zwift’s “work-out” mode. Then compare the total times it took to complete those two rides, and compare the ave power output (watts) required by these two rides.

If the second ride wins by a statistically significant margin, then the hypothesis is true. If not it is false.

Hope this helps.

Again your conclusion is wrong. It really has nothing to do with how much time difference between the 2 rides. It is about how much energy is expended and how much you have left at the end of the ride, this is more true in a triathlon then a TT. 

Seems also your 2 rides had different elevation profiles so the skews the results even more. Plus power difference is more than likely more important than just watching heart rate.

Were is your data output?

Were is any and all graphs of the rides to compare?

Heart rate is just one variable, what about power?

You have give virtually no supporting data, just words on a page. 

I see what you are trying to do, but your process is flawed.

I will be waiting for your condescending reply.


Thanks for your reply.

I have not made a right or wrong conclusion yet, just suggested a method to possibly come to a conclusion. So far the sample was too small (one only), and the difference in power and time taken and HR small too, so no conclusion yet. So far the hypothesis looks false, so far.

No, I used the same route (London Loop) in both cases, so there is zero elevation difference to skew the results.

No graphs available for power, only ave. HR available from Garmin. These averages are adequate to conclude. One does not need graphs over time.

I get the HR readings and calories from my Garmin and the ave power and time readings from the Zwift dash board.

The reason why I have not published my figures as supporting data is not to confuse the essence of the topic with numbers - hence a page with only words. With only one experiment, the supporting data is meaningless.

If my process is flawed, then please support me by suggesting an alternative – using only Zwift and Garmin. Alternatively, you could carry out the experiment yourself. Then we have a sample of 2!

Average heart rate really means nothing, that is why a graph over time should be used.

Really an average of any of the readings is not a clear picture of what is going on. I can average 145bpm or 225 watts, but what does that really say?

If you are going to go further with this you will need to use more than just average heart rate. You should export your data to Strava, TrainingPeaks or another site so you can get a more clear picture of what is going on and when. 

You should have waited to post this until you have done more test. Also you could have asked others to do a test for you so you could get more data from more sources.

I believe Triathlete Magazine published the results of a test similar to this (not sure who did it) and it was concluded that a steady state effort was more beneficial and conserved more energy with a similar average power and heart rate over time. I think this is more or less the way Fromme rides and most of Team Sky and seems to be the way more training is being directed.

Lochi, your experiment is essentially flawed from the start, as stated above by Paul because heart rate is so variable and influenced by so many factors externally that it is basically useless. You need to be working based on power.