Research from the company Bioracer Motion in collaboration with Ghent University investigated the subjectivity of different bike fitting services and the subsequent impact on advised position. Results showed that bike fitting is quite severely suffering from expert subjectivity.
Supported by a Flemish Innovation and Entrepreneurship grant (Baekeland mandate), collaborative research and development is done on the topic of data-driven bike fitting, using Bioracer Motion Sensing Data. Based on this work, the following publication was presented at icSports 2019 in Vienna: The Need for Data-Driven Bike Fitting: Data Study of Subjective Expert Fitting.
Before the results and insights of this paper are presented, it might be useful to have an idea what bike fitting actually is. This kind of service tries to setup a cyclist’s bike in such a way that it is optimally suited for the individual rider’s needs and ambitions. On a bicycle, there are quite a lot of adjustable parameters and determining the right values for these parameters is very important before somebody clips into his/her pedals. The reason for this is twofold: (1) cycling is a highly repetitive movement, and (2) the body is stuck at three contact points (handlebars, saddle and pedals). In addition, as most active cyclists are pedalling at roughly 100 rotations per minute with rides that can take up to 5 hours, the same, repetitive movement is repeated around 30.000 times per ride, which increases the risk of overuse injuries. The combination of a rather static position on a bike and the highly repetitive nature of cycling movements can cause big problems with even the smallest misalignment/mistake in the rider’s setup.
What’s in the paper?
A lot of people with different backgrounds, levels of education and experience aware of the fact that a proper bike setup is indeed a crucial aspect of the overall cycling experience, have started to offer bike fitting services at various locations. However, among those ‘bike fitting experts’, there are a ton of different approaches and philosophies. A very prominent difference is the technology they use to fit people on their bike. Bioracer Motion developed their own system which allows bilateral analysis of body movements during cycling. To date, it is still one of the only systems where the right and left side of the body can be analysed simultaneously. Of course, not everyone relies on computer aided motion analysis systems. Some bike fitters use saddle pressure technology, while others just film the cyclist or don’t use any technology at all (they fully rely on what they see). Apart from the difference in adopted technology for the bikefit, there are also big differences in the amount of time spent with the cyclist and the way intake, reporting and after fitting service and experience is done. Price also varies among various service providers.
All of the aforementioned differences led to the inevitable research question “Does bike fitting suffer from some kind of expert subjectivity?”. From previous experiences and relevant literature, it was hypothesized that it would indeed be plausible that one bike fitter would advise a different position than another. To confirm the hypothesis, 3 cyclists were subject to 9 bike fits at different bike fitting studios, spread across Flanders. The adopted principle was kept the same for every cyclist: they arrived with the same bike setup and they were not restricted in any way to answer questions asked by the bike fitter.
This experiment resulted in considerably large differences between the advised positions. The range in saddle height and fore-aft position for both test persons was around 3 centimeters. It was presumed that there would be differences between the various bike fitting studios, but it was beyond the author’s expectations to notice such large differences.
Bike fitting is quite severely suffering from expert subjectivity. The industry will need several changes to try to narrow down this gap. Firstly, education is and will always remain very important. There is no uniform education in bike fitting to date. However, different initiatives are undertaken to provide some kind of certification (i.e. International Bike Fitting Institute). Secondly, bike fitters have to be more aware of the needs of the individual. Some bike fitters advised a more aerodynamic position to the recreational rider as compared to what is advised to semi-professional riders.
The further Baekelandt mandate funded research will focus on a more objective way to conduct bike fits. As there is a lot of data gathered with different technologies, a data-driven way of bike fitting might be a solution to this problem. The conclusions of these kind of experiments are providing valuable evidence for the need of such a system (combined with better education).
In case you have any questions about this research or the topic in general, contact Jarich Braeckevelt of the Department of Electronics and Information Systems at jarich.braeckevelt@UGent.be.