From the beginning of my return to serious cycling and structured training a decade ago, I determined I would need two things to improve,  power data and an indoor trainer.  Even though I had been relying on a heart rate monitor during my running years, I had read that there were better ways to measure training effort on a bike.  Also, I learned there were more productive ways of training than merely grinding out kilometres on the road.

Relative to the cost of a new bike, power meters were expensive. Ten years ago, $2,500 bought you a decent intermediate level carbon road bike with 10 Shimano 105 gears but the cheapest power meters were single-sided crank meters costing a further $700.  Having laid out that expense, I sought an indoor trainer and decided my best option then was a $300 Kurt Kinetic liquid roller wheel that was essentially mechanical but had an upgrade so it that could transmit power data to an app via blue tooth.   

My initial indoor training comprised watching Youtube videos of cycling in the Alps whilst trying to follow the power profiles of a free British Cycling training program that I had printed out on paper.  I joined the Zwift platform early on but back then, Zwift didn’t have much in the way of training programs so I moved to using Trainer Road in addition.   At least then I could follow the power trace on a separate computer screen and ride in the Zwift virtual world.  For Zwift and Trainer Road, I was able to use the power data from my left pedal crank. 

Having data from just on crank wasn’t an issue unless there were particular training instructions to concentrate on just my left leg for example  when I would notice big differences in the reported power for the same perceived effort.  At one point, I had enlisted a personal trainer to analyze my riding and give me feedback.  He put me on his Watt Bike that recorded effort from each pedal.  The results showed that, like most people,  I had an imbalance of effort between left and right sides and I always wondered how that would impact my true power output on a bike where only data from the left pedal was recorded?  If my left leg generated less effort, wouldn’t my true power be underestimated by a single side power meter?  How did more modern “smart” trainers differ in this respect?

The key feature of a smart trainer is that it links in with the terrain and conditions being projected in the virtual world on screen.  If the rider begins to climb a virtual hill in the same gear and cadence (RPM), then their power will need to increase or they will slow down as on a real hill.  On my non-smart trainer if I continued in the same gear and cadence my power would remain the same and my speed would drop on a climb.  I would need to increase cadence or gearing to generate more power to climb at my previous speed.  This is contrary to the real-world situation where you would likely reduce gearing ( go to a bigger sprocket).

Some current high-end trainers have an additional features  where the bike front forks are raised to simulate the actual incline of a hill,  much in the same way that running treadmills can incline to simulate running up a slope.  Also, there are trainers that simulate the lateral and fore and aft movement of the whole bike. Lateral movement is nothing new as Kurt introduced this feature on a machine more ten years ago but it was a feature that never really caught on.   Similarly, there are fans to simulate windspeed that vary with the video scenario.  The jury is out as to whether these extra features are really worth the expense.  The harsh reality is that you are on a training bike in your basement, not  riding to a mountain pass in the Andes.

Having researched all that was currently available I needed to sift through the features that would actually make a difference to my training experience.   Unlike most indoor riders, I suspect, I was someone who could stare fixedly at a mechanical power reading and stop watch on a gym fixed cycle and follow a printed interval program. I did find blasting loud music through headphones helped.  Oddly, headphones, much like pedals, are left as an individual choice with indoor training systems even though some systems, like Pelton, have a screen for the video.  For my video screen I have always used the television that was superseded  from my sitting room  and I powered by an old laptop with a gaming chip.     

So, here are the potential features of indoor cycle trainers:

1. Power at the drive hub displayed.
2. Power at the drive hub automatically varied by the virtual terrain.
3. Power balance between the left and right pedals or cranks displayed.
4. Cadence displayed.  
5. Lateral movement of bike to simulate sway.
6. Fore and aft movement of bike to simulate surge.
7. Raising of front forts to simulate incline.
8. Variable fan speed to simulate speed of bike.

For the visual display, there are several options:

1. Wall mounted television
2. Stand mounted proprietary computer screen
3. Stand mounded laptop

The quality of the perceived ride is dependent on a number of the trainer’s other  specifications:

1. Weight of flywheel – generally it’s agreed that a heavier flywheel is better to simulate inertia of wheels on a real bike.
2. Wheel on or wheel off.   Higher end trainers are all wheel off versions where the bike rear wheel is removed and the rider drives a resistance unit.  The resistance unit may have a cassette or a single sprocket like the Zwift hub system that simulates gears.  In cheaper systems, like my Kurt Kinetic, a trainer tire is fitted to a rear wheel and drives a roller by friction.   

Finally, the quality of the data out will impact the overall cost.  A system that claims power accuracy to within +/- 1% will cost at least twice the price of a machine delivering +/- 2%.  Cheaper machines are likely to be only accurate to 5% which become significant if you aim to train in zones very close to your sustained maximum power output or Functional Threshold Power.

Given this plethora of options, how can I choose what I need and not just go for what I want because it is new and available?    Firstly, I’ll look at power data and control. All machines give power data and, for me, an accuracy of +/- 2% (about 5 Watts at my FTP)  is probably adequate.  While it may be interesting to know what the balance of output between my left and right legs, there’s not much I would be able to do with that information so it’s not something I am going to regularly need. 

I guess the key question regarding power is whether I want “smart” control with the power demand automatically being changed by the training program, or in the case of programs like Zwift, by the virtual terrain.   People I have spoken to insist it is much easier to follow a training program, particularly in the so-called ERG mode where the program sets the power level the rider has to maintain.  If your cadence slows, power is maintained by forcing you to increase crank torque.  The advantage here is that you are no longer having to chase the power level by pedalling faster  or changing to a higher gear ratio.

Owners of smart trainers report that you certainly know you have hit a hill in the virtual world when riding in the resistance mode of a smart trainer.  This advantage alone justifies a smart trainer in the eyes of many athletes.  Whether you always stay in ERG mode for intense training intervals is up to you. Some athletes suggest using resistance mode for intervals at VO2 max. Either way, the option is open to you with a smart trainer and, as someone said,  I would rather use the small amount of extra brain oxygen needed to use  dumb trainer in my leg muscles!  

As for simulating the incline of a hill by lifting the front forks, like many, I am as yet unconvinced.  Unless you are deliberately riding killer gradients, the few degrees of inclination of most hills are barely noticeable  in terms of bike position.  According to Bike Calculator , a cyclist generating 150watts should be able to maintain 28kph on the flat but to travel that fast up a modest 2% (1.12 degrees)  hill the power would almost need to double to 285Watts.  Now, it would be hard to feel a 1 degree slope if you were stationary but you’d quickly notice the extra power needed if you wanted to maintain speed!   

A fan emulating wind speed also seems an unnecessary extra to me. Cooling fans are important especially in summer of it you are in a windowless basement.  However, if you live in Calgary, you only need to crack a window in winter to get the temperature down.

So, do I need a smart trainer? I think the answer is yes because I ride indoors three or four days a week in winter on structured interval rides and group endurance rides.   I don’t think I could justify spending over $1,000 but there are several excellent sub-$1,000 trainers out there.  In Canada we do appear to be paying a premium for these machines though compared with the prices charged to Europeans and Americans.  The most popular mid-level trainer in the UK is currently selling for 450GBP (840CAD) including tax whereas in Canada it is  945CAD with tax.  The US price is around 582USD or 810CAD. Short of going abroad and importing the trainer myself, I am resigned to paying the premium.