Can you replicate a Wingate test on an indoor trainer?

I’m studying sport and fitness and the current topic is anaerobic and aerobic fitness. We have been provided with a list of anaerobic field tests that you could use to determine anaerobic fitness; these are not just cycling related. However the Wingate test makes a lot of appearance.
In the test the resistance to the ergometer is proportional to the bodyweight of the athlete. Is there any way that this type of test could be reversed back out to a test that could be performed on a indoor trainer. My gut feeling is that resistance mode is probably going to the best method.

I guess the 2 difficulties are:

  1. Working out how the weight resistance applies to a variety of trainers and resistance settings

  2. Working out how to replicate the effect of the instructor lowering the test weight basket at the exact moment the cyclist starts their effort.

Sadly I’m nowhere near clever enough to answer either question, but I’d be delighted if someone else is…

Full disclosure: I know nothing about any of this. But reading what you posted, it seems like those cycling ergometers are just very antiquated power meters. So indoor trainers actually should work better. It also sounds like that test is just a 30 second max effort?
So I guess use the warm up to determine a reasonable gear and then hammer it out for 30 sec in resistance mode :man_shrugging:t2:
Again this is just guessing based on the descriptions. No actual knowledge or experience :sweat_smile:

p.s. I love this part of the test protocol: “Note: Test administrators must provide verbal encouragement through the test.” :joy: :joy:

I remember doing the Wingate test during my Bachelors degree back around 1992/93.
It was damn hard (I was 115Kg back then - due to Rugby and lots of time in the weights room).

If you do it on a trainer in ERG mode then how will you replicate the Wingate test protocol?

You can use a workout builder to create a wattage target profile for ERG mode but how will you correlate the correct wattage target for bodyweight?

Thats not how (IIRC) the Wingate test works - it is basically a normalised absolute power test that shows not only peak power (resistance is normalised for bodyweight)* but also shows how quickly you reach that peak power and then how quickly it tails off.

  • on the bike that we used back in the day, it was basically a loaded cable that increased pedal resistance and you loaded actual barbell plates on depending on bodyweight - there was a table of reference to calculate what weight load for what bodyweight.

Works a treat though - my power was waaaay up there but tailed off super fast - others in my cohort had a very different curve.

I’d be really interested to understand how you go about this - just from a curiosity perspective :+1:

I think this could be done pretty easily on a trainer that has an appropriately fast reaction time.

I think the best way to do it would be in free ride mode on Zwift, on a stretch of road with a slight uphill gradient.

The reason I think Zwift would be good is that the resistance increases with speed - which means that the pedaling gets exponentially harder the faster you pedal (power required to overcome aerodynamic resistance varies with the cube of speed).

This would help prevent riders taking the test from “spinning out”. Eg it’s almost impossible for people to “spin out” on an air assault bike - the air resistance from the fan increases too quickly. So same principle here with riding in Zwift.

Side note: I think an air assault or airdyne bike equipped with a power meter would be the ideal bike for this.

The reason to do it on a slight uphill grade is to have some fraction of the resistance to be proportional to rider weight. Then the faster the rider goes, the more aerodynamic resistance takes over - and resistance is related to speed, not rider weight. Bigger riders will be able to ride faster. But the cube relationship between power and speed means not a whole lot faster, hence meaning less chance they’ll spin out. E.g. 30 mph vs 25 mph is 1.73x the power to overcome aerodynamic drag.

The appropriate gear can be chosen so that the rider is riding at, say, 100 rpm cadence at the power/speed where you estimate they might max out.

TR resistance mode has a linear power curve. So I don’t think TR would be the ideal platform for this, as there would be higher risks of riders spinning out (don’t get the benefit of the exponential power curve vs speed).

Edit: this approach wouldn’t exactly replicate the wingate protocol you linked to, but it would produce a very similar power curve that would measure max power output In the initial few seconds, and then decay over time.

Edit #2: wingate tests are brutal. I read somewhere about someone doing a 3 minute wingate test. 30 seconds is bad enough.

Thank you for the feedback.
Looking at the metrics that are collected during one of these tests it appears to be a full all out effort. The weighted basket is there to provide resistance to the bike, so as long as there is a way to replicate this in proportion to the rider then it should be possible. As mentioned all the data would then be there to compare.

If I understand correctly the original intent of the “weighted basket”, it was to adjust the resistance to the rider weight so the rider would not spin out during the test.

If you want to adjust resistance to solely rider weight, you can do so by using resistance mode in the trainer (eg using TR), and setting a high slope eg equivalent to riding up a steep hill.

IMO, a better approach is to take advantage of exponential power increase vs speed, as it prevents riders from spinning out better than just adjusting resistance to weight.

Example: say you have two 75kg riders, Rider A with an FTP of 175W and Rider B with an FTP of 350W. If you set the resistance just based on weight, Rider A night max out at a cadence of 70rpm, but then Rider B would have to hit 140rpm cadence to max out. That’s a pretty big difference.

If you adjust resistance based on speed, the cadence difference between these two riders would be Rider A at 70rpm, and rider B at 88 rpm.

Anyways - lots of options for how to do this. But should be possible with the right trainer.

Yes, you should be able to replicate a Wingate test on a smart trainer. The original test was on a mechanical ergometer (a modified Monark cycle ergo) with a friction based resistance via a set weights hanging on a rope wrapped around the wheel. The weight is normalised to body weight (usually 7.5% BW). The resistance is set and power increases with increasing velocity (angular velocity or cadence). The trick is to mimmick that electronically.
I suggest using “resistance” mode. The Wahoo Kickr has 0 - 100% in resistance mode. The question is what percentage of resistance mode is equivalent to the 7.5% of the individuals BW required for the test?
This may require trial and error or calculation of the torque at the bottom bracket of 7.5% BW (since this does not change) and finding the equivalent torque curve for the smart trainer resistance mode.
If there are no published torque curves for the smart trainer (I don’t think there are), you could do some steady state tests at the lower end i.e. get power and cadence --> calculate angular velocity (rads-1) to then calculate torque Nm on several different resistances. Then extrapolate on a percent resistance vs torque curve. Then find the percent resistance that matches the 7.5% BW torque.
In Trainerroad build a workout for the test e.g. 5mins in erg mode, maybe 50W warmup, then switch to resistance mode for 30 second at the percentage that matches the 7.5% BW, then 5 mins erg mode easy spinning to recover. (Note: The resistance doesn’t have to be 7.5% BW, That is just what the original test has. You can adjust for different populations and cycling experience).
You can then analyse for peak power, mean power, fatigue index, time to peak power etc etc. The only issue is that everything is rounded to 1Hz (1 sec). Ideally, 10Hz or more would be good but not a deal breaker.

1 Like

I don’t believe a validation study has been done yet performing a Wingate test on (eg.) a Wahoo Kickr vs. the Monark. I think the Wattbike has been validated.

Cursory searches:


Would be a great simple study to run… when labs start opening up again.

As others have said it would just be a matter of calibrating the electronic resistance to approximate the target power peak output & cadence.

We have done Wingates on the Kickr in the lab for pilot testing with no issues, but not for anything publishable. Just have to estimate approximately the right resistance & gearing combo for the athlete. Even having to shift here and there doesn’t seem to throw off the data much. If anything you could say it has greater ecological validity for trained cyclists.

A while back I was interested in exactly implementing a Burgomaster protocol on my trainer…so I was trying to figure out exactly what you are. The advice I got most often was ‘Just put your trainer in slope mode, move the resistance up to 80% or so, and go all out for 30 seconds’. That works ok.

But, having done a Wingate test, and having done an all out 30 second print on my trainer…I would say the Wingate test is a darn sight harder. So there you go.

1 Like