These are typically the cheapest and lightest types of smart trainers, but they can cause wear on your tyres (though specific trainer tyres are available to mitigate this issue), their power measurement is generally less accurate and the ride feel often isn’t as good as direct-drive trainers. While the NEO 2T did have some initial accuracy issues when it was first announced, those were largely resolved in subsequent firmware updates to my satisfaction. In all, the Zumo is super portable thanks to its carry handle, foldable legs and general lightweight build. At the same time, it’s very stable thanks to its wide footprint and the resistance is very nicely controlled and feels very smooth despite the smaller flywheel. As with other wheel-on smart trainers, the Flow Smart is louder and does report power with less accuracy than direct-drive models. But if you’re on a limited budget, it will still do a respectable job.

I feel like the Saris H3 is the little trainer that could. In past years the H1 and H2 have always fallen short on the list. But last year the H3 finally climbed into the top recommendations list. With the H3 the company made some minor technological tweaks, but also substantially reduced the price. And that combination put it over the edge. The Zwift Hub is a competitively priced direct-drive smart trainer from the virtual cycling and running platform, Zwift. The only other niggle is that the trainer doesn’t fold up for easy storage – the legs are simply fixed in place with bolts. This won’t be a problem if you have a dedicated pain cave, but if you need to be able to easily stow it away, this might be a dealbreaker. There’s once again been almost no appreciable shift in this category this year, so things stay basically the same as last year. And, there’s really only a few entrants in this category anyway. Only Tacx, Elite, and in some areas Magene compete in this realm as a true smart trainer (ones where you can control resistance).ANT+ Speed/Cadence Profile: This broadcasts your speed and cadence as a standard ANT+ Speed/Cadence combo sensor The H3 sits at the top of Saris’s smart trainer range and builds on the popular H2. Reducing the noise levels was one of Saris’s top priorities and it’s certainly achieved good things with the H3 – at just 61dB at 20mph (measured on an iPhone app), it’s very quiet indeed. Like I always say – for me personally, it’s hard to separate the fact that I’m riding indoors from outdoors. It’s still a trainer, and I’m still looking at a wall in front of me. My brain can only turn off so much of that. Still, much of the road-like feel is driven by the flywheel, and be it physical or virtual, flywheel sizes tend to be measured in weight. This impacts inertia and how it feels – primarily when you accelerate or otherwise change acceleration (such as briefly coasting).

How much you have to spend will ultimately dictate what kind of smart trainer you get, but it’s certainly possible to put together a good indoor cycling setup on a tight budget. It can simulate gradients of 16%, which covers more than 99% of Zwift's roads. Only a couple of stretches on a few of the most difficult climbs push higher than 16%.

Advertisement: Save up to 50% on Wahoo bestsellers

With a comprehensive specification for its price, the Pinnacle trainer impressed in use, offering easy setup, good ride feel ,and quiet operation. Hovers around 60db in my tests. While average in comparison, it's still quieter than drivetrain noise and not a problem with neighbours or family members. I’ve just got my replacement Saris H3 from ‘LordGun’. Original returned due to woeful under reading of power resulting in workouts being very challenging to impossible. It also was inconsistent depending on gear selection/flywheel speed. Comparisons done against 2 sets of Favero Duos and Giant Power Pro cranks recorded on DC Rainmaker Analyser. Replacement is closer in power but still inconsistent on two metrics. 1; Depending on gear selection/flywheel speed the actual power required varies for a given target/reported power. 2; Depending on the actual target power itself the deviation is inconsistent. Note: The Elite Zumo is not currently distributed by any retailers in the US (it is in Canada though, and in many places in Europe). This is up to individual retailers for whether they carry a given product, which may have different margins than the rest of the Elite lineup. Similarly, for challenges such as a ‘virtual’ everesting, the rules stipulate that the realism must be set to 100%. If you’re planning on using the Alpe du Zwift for your attempt, then you’re going to fall foul of ‘the rules’, as the maximum gradient there is 14%.