ERG mode - is slow trainer reaction for sprints a myth?

After reading the “do you switch off ERG mode?” thread, I thought I’d zoom in on yesterday’s Ebbets workout, as I noticed that the 5 second bursts seemed to be fast-reacting and on-target.

So my example, zoomed in below, is a 5 second burst from 233W to 477W on a 2017 Kickr Snap in ERG mode. You can see that the actual wattage (yellow line) ramped up to the 477W target 1 second early and fell back to 233W also 1 second early. The net effect is a full 5 seconds at the target., which was pretty much the pattern for the whole workout.

Note that the following 233W “valley” is matched perfectly by the Kickr Snap, I think this is helped by not running a super-high gear (I used 50-16).

When the following valley is much lower than 233W, say, 110W, the flywheel speed means I get a 0W reading for a few seconds before the resistance “catches up”, so for workouts with bursts that are interspersed with very low valleys I just select a lower gear to reduce inertia and minimise this zero-effect.


For the record (for other Kickr Snap users) I use a Vittoria 23c turbo tyre at 100psi and have 2.25 turns of the knob after tyre contact. Spindown is usually around 12 seconds. I have never once had a slippage. All in all I’m pretty impressed by the smart trainer.


Do you have a power meter? If so, don’t link it to TR, just record the data with your normal head unit. Then run the test again and compare the readings from your smart trainer to your PM. I found there to be a huge difference between my smart trainer and PM. I have since returned it to the shop.

I don’t have a power meter, so it’s irrelevant how accurate my trainer may or may not be to a power meter, although incidentally DC Rainmaker found the Snap to be very close to his power meters.

Without a power meter, as long as the trainer/tyre pressure (etc) settings for each workout are consistent then all is good.

The point I was making is that the claim that smart trainers in ERG mode are incapable of changing resistance in very short timescales (eg for sprint bursts) seems to be a myth, at least on my sample of n=1.

1 Like

That’s unfortunate. My power meter was reading up to 20% lower than the smart trainer when it had control. It wasn’t irrelevant as the gap increased the higher the wattage. It was not a constant x%. If I’d only had a smart trainer I’d think I was doing V02 efforts but in fact I’d only be working at threshold.

I found that powermatch didn’t work very well either hence why I returned it.

I did Ebbets on my kickr core with erg mode on and didn’t get near the power targets.

Thought I’d give it a go but will be turning off erg mode for short sprints after this.

Check the sprints at the bottom I was quite consistently out by 70watts. It did overlap into my sweetspot after but I never hit the power numbers.

What trainer are you using? EDIT, sorry, I didn’t read you post properly!

That’s interesting, I wonder why the Snap would follow the sprint power outputs closely but the Kickr Core would not?

Interesting. What trainer was it, out of interest?

Just for complete notes.
Kickr core. In small chainring and middle gears at the back. So relatively low fly wheel speed.

Additionally I have powersmoothing off.
Was powersmoothing turned off on the snap?

Thanks for the extra info. I’m embarrassed to say that I don’t know if my Snap has power smoothing on or off. I operate simplistically with Ant+ only (for all devices) and an Android phone.

My additional problem with Bluetooth is that currently I cannot connect my Snap to the Wahoo App (I have to spindown via the TR App).

As a result I don’t think I can check the status of power smoothing although judging by the fact that charts are always very smooth in profile and the fact that the “Actual Watts” figure in top left in a workout varies from “Target Watts” by only one watt usually, I think I must have power smoothing on.

N=2 in my experience with the Kickr Snap

1 Like

Powersmoothing is very likely on for you then, it comes as on in default with wahoo trainers. I’d recommend you borrow a phone and connect it to that and turn it off.

My power with 7 maybe 10 second smoothing varies greater than +/-1 watts more like 5 watts. We aren’t robots power output isn’t smooth.


Based on the image in your OP I’d say undoubtedly you do have power smoothing on. I have a Kickr and if I don’t have Power Match on my TR output looks pan flat like yours despite my head unit is bouncing all over the place.

If you look at other’s workouts and you see a perfect target to actual curve they are likely using a Wahoo trainer in default smoothing.

Power Match does its best, but it must be a near impossible control sequence to write. I quite enjoy the challenge of riding in Power Match and keeping the target as close to actual. I’m often focussed on keeping my pedal stroke as smooth as possible and cadence perfectly regular, but even still, you can feel the resistance moving underneath the pedals.

@chrisbowiehill What’s your experience in the fast reaction time of your Kickr in matching target and actual watts in short duration high wattage burst/sprint intervals?

Wahoo power smoothing is turned on by default, once turned off the curves will look different.

1 Like

From my experience, if you use a small chainring up front and middle to large gear on the cassette, the slower flywheel speed allows trainers to react pretty fast, even in power match mode. However, if the flywheel speed increases they start to react slower.

Also you definitely have power smoothing on. No one can pedal that smoothly. If you turn it off just know that you get the same workout even if you are nowhere near as smooth as it appeared when the smoothing was on.


I’m guessing you saw my 5 second sprint results in the other thread, reposting here for reference to the two screenshots below:

Here are two examples of 5 sec intervals that I was “on target”

Sprint 5:

Sprint 11:

This was on a Kickr 2017, with power measured at crank (Stages LR) and gearing of 53 chainring and somewhere in the middle of Kickr cassette on the rear. Prior to that I had subcompact 50 tooth chainring. I have more experience with 50 chainring on the Kickr, and have had no real problems getting a fast reaction in the big chainring. Maybe I’ve got good torque and fast reactions? I’m a slow twitch guy that was a long distance runner in my youth and struggle a bit with VO2max work on the bike.

Everyone likes to talk about flywheel speed, without regard to all the gearing in between and ignoring Erg mode means the Kickr is actively adjusting resistance. Don’t forget you have a flywheel, a resistance control unit, Kickr gearing, and bike gearing. Its a complete system, trying to explain trainer behavior simply because of “flywheel speed” is like saying the earth is flat because I can’t see the curvature on the horizon :wink:

One reason we have a little chainring is to make it easy to go uphill. When attacking or changing speed, the little chainring also acts like a torque multiplier (less torque required to cover accelerations).

1 Like

My experience is much the same as @bbarrera.

The reaction time is definitely much better with a slower fly wheel, but then you’re training yours sprints for breaks from a pack on a steep slope.

For the first year of having my Kickr (2017) model I ran with a slow flywheel, but I’ve realised this doesn’t help me much in the flat crit racing I do.

About 3 months ago I swapped to running 52x17 and the response is definitely slower. I do a session which is 20x 6s @ 850W / 24s @ 130W. I regularly go through 1,000W early on with the 6s sprints. As I get towards the end of the 20 reps sometimes I don’t quite hit 850.

Would be great if it was spot on, but bless, TR are doing their best :wink:

1 Like