Lama Lab Test of the Trainer Difficulty Setting

Lama Lab conducted a test of the trainer difficulty setting on Zwift where he rode up a hill at 100% difficulty setting and then again at 0% difficulty setting to demonstrate that the difficulty setting had no effect on the power output. Here are his results:
100% Trainer Difficulty Lap #1 - 354W [2:18]
0% Trainer Difficulty Lap #2 - 352W [2:20]

I understand that on the second lap (0%) the force applied on the pedals was less while the rpms increased keeping the power the same. So it took the same power and the same time to climb the hill proving that the difficulty setting had no effect.
But if he applied the same power and the resistance was less (gradient) on the second lap, wouldn’t he had gone faster? Does anybody know how Zwift calculates power and speed?

I’m sure someone will give you a more detailed answer than this, but the general calculation for speed in Zwift has to do with the wkg the rider is putting out. So, for the Lama’s second lap in your example, the time would be the same with the same wattage output, because all of the other ‘in game’ factors are the same, also (i.e. rider height, rider weight, course, bike, wheelset, etc.)

Now, IRL he might have gone faster on that second lap (presuming 0% gradient vs whatever climbing had to be done at 100% on the first lap), but in Zwift the power output is the primary determinant of speed. Therefore, at the same wattage output with the trainer at 100% difficulty the ‘rear wheel’ might be rotating more slowly than with the trainer at 0% difficulty (seemingly indicating a slower speed at the 100% level), in the game the avatar would be travelling at exactly the same speed over the same terrain.

This is one that I feel like confuses people regularly (most often with people commenting about how the speed on their Garmin/Bolt/Whatever is different than their avatar speed. So, just keep in mind that the in-game physics might be different than what you experience when actually pedaling the bike on the trainer.


I assume you refer to the video he made to show how trainer difficulty impact going down hill

Zwift does not calculate speed based on resistance, it use the power number.

Zwift does not calculate power from a smart trainer It just use the number that the trainer send.

Calculating Speed from power is a complex calculation.

Zwift probably use something like this.


In real life, you can’t change the perceived gradient without changing the actual gradient used to calculate your speed as an output of your power. The only equivalent thing you can do in real life is shift gears.

The in-game physics only cares about power as a variable input. How you generate it (at what cadence, trainer wheel speed) is irrelevant. So yes, it creates a strange disconnect, which you can clearly experience any time you’re climbing in the 15-20% slope range. You have the imoression of slipping in place.


Trainer difficulty only affects the trainer. Ride up a 10% gradient at halfway slider, and the trainer will be told to emulate a 5% gradient.

As far as the game is concerned, you’re still riding up a 10%.

So whatever your slider is set at, if you’re doing 250W Zwift uses that to calculate speed up a 10% gradient.

The only difference is that your trainer makes it feel easier or harder, and that can in turn make it easier or harder for you to put out power (most people find it’s easier to apply power when the resisting force is greater).


Well stated. The trainer receives a slope/resistance that’s half of the one displayed, but Zwift still calculates output speed with the full slope.

1 Like

Thank you! That was the disconnection in my analysis, that even though the resistance (gradient) provided by the trainer is less, Zwift doesn’t use the new gradient but the original gradient for its speed calculation. Now the numbers make sense. Now I can sleep again :slight_smile: Thanks again!


I did a diagram.


That graphic should be in a sticky message at the top of the forum. Lots o’ confused people asking about this.

Great job.


Now make one for descents!


Done. :smiley:


Now Wahoo only needs a fix in the firmware for the kickr climb…

if (gradient from zwift<0;double gradient)

or sth like that :smiley:

Are you certain about the last column for descents? I was under the impression that the 50% factor extended to the simulated speed calculations as well.

I thought otherwise. It’s not what Shane’s saying in his video, as far as I understood it.

There is an example where he U-turns on the KOM - Zwift says 7% and his Wahoo says 3.5%; but he’s descending at 40-50kph at the bottom. I don’t know what Shane weighs, but that doesn’t seem like a 3.5% speed to me. Then he does it at half Trainer Difficulty and still hits 50+kph.

Indeed, it would make no sense for Zwift to make people slower on descents when they don’t make people slower or faster on ascents.


I thought the point was the same as with the 50% factor in general: to have the rider make more of an effort going downhill.

The main reason I’ve been wondering about this is because the ascent and descent numbers in the ride data don’t match when riding a loop, the ascent is always bigger but the factor varies. (Of course there can be other explanations, such as different thresholds for what counts as a climb vs a descent.)

It’s not about making the rider make more of an effort, really. Lowering trainer difficulty makes your trainer simulate less of a hill. In turn that means you have more resistance to push against, so can make more of an effort.

Trying to put down power on a 10% downhill is harder than doing it on a 2% downhill or on the flat.


Eric Min in the latest Zwift cast said that the 10% downhill will feel like a 5% so you don’t spin out. “its a little complicated” then he took a note about it presumably so they (Zwift) can figure out a way to explain it to all of us…
Nothing was mentioned about the virtual speed/decline, so we can only speculate.

I would think if Zwift simulate up hill as the real incline then they will do the same with downhill.

The slider is just there to change what the trainer does the hill in Zwift never change, only the amount of resistive force that your trainer apply.

To calculate speed Zwift use Power, incline , mass and a bunch of other variables. The incline is as it is defined in the route profile. So Zwift don’t care how you generate the power you can have a lot of resistance and low rpm or high rpm and low resistance, it is your problem how you generate the power.

1 Like

Yeah, that interview was what I was thinking about as well. How I understood it was that the descent factor does not behave the same way as the trainer difficulty setting because its purpose is different. Looking forward to the official explanation…

Even if the purpose is different the inner working should stay the same. Zwift use power to calculate speed using the real incline as per the route profile.

Zwift is a training app so there is no point in going downhill with no resistance. Zwift knew that with steep descents you will just have to sit and wait for your averar to get to the bottom before you could pedal and that will be boring and not training. So they gave us some resistance to push against.

1 Like