The science behind the Spoked algorithms

Dr Nic Berger
19 min

We want to share our point of view with you. What are we trying to achieve and how are we doing that?

“The most important thing: we want to build an online coach that is as good as having a real-life coach and an exercise physiologist at your disposal at any time during the day.”

You see, they are not the same. Coaches are of course very good at what they do, and they often use a combination of experience, what they have learnt and intuition. An exercise physiologist on the other hand understands exactly how the body works and what is the best way to improve performance. Working together they are a powerful combination. And finally of course we have you, the athlete. We’re continuously communicating with our athletes to make sure we show them what they need to and want to see. Importantly, also what they don’t need to see, as sometimes this can become overwhelming with so much data and information to choose from.

"We use facts backed up by science, and hard data to inform practice.”

This means that any of the questions or changes in the plan are based on scientific findings, not just because we think they should be there, or ‘because it worked for me’ (it is frustrating how pervasive this kind of thing is). I will reference and link the relevant papers so that those of you who are interested in reading more about each topic will be able to. Crucially, (and you’d be surprised how often this is lacking), the plans will be flexible and use common sense. And this is what we have created at Spoked: coaches, exercise physiologists and athletes working together to create true custom plans.

Now you might be wondering why you should pay for such a service and not just use another site or plan that might be cheaper (hard to find though to be honest!), or even free. The quick answer is: they aren’t very good.

“A recent review of physical activity apps showed that they were of limited quality and had shortcomings in their effectiveness (Bondaronek et al., 2018).”

Sure, they will set out some structured training plan and if by a random stroke of luck your body responds exactly like the plan says then you will improve. Here’s the very large flaw though: we aren’t all the same. We are not predictable robots and have busy lives that can affect us, our performance and recovery in lots of different ways. Therefore, the thing that sets us apart from other services is that we use science and common sense to check how you’re doing every day; just like a real coach would (Davis et al., 2019). Herein lies the issue with pretty much any other online training platform: they are lacking the ability to constantly assess and adjust your training. Roughly they are split into fairly expensive options where a real-life coach will interact with you, or they are cheap(er) / free and very inflexible. I won’t name all our competitors here, but suffice to say there currently only is one other ‘adaptive’ training plan. Their plan is purely based on your ability to hit certain numbers, but is lacking a multitude of other factors that can affect your performance or ability to train. So I will use this blog to highlight our point of view and explain what sets us apart from every other training app out there, and why we have chosen to ask and track specific things about our athletes.  

My job at Spoked as the in-house exercise physiologist / sport scientist is to check the plans, inform all the feedback loops, decide which key variables affect the plan and constantly keep updating the algorithms to match the newest findings. This is key, as research is constantly evolving. Ultimately, using this approach has resulted in an adaptable plan that suits your needs.

Something a coach will ask you each day is how you are feeling (Davis, 2019).

This obviously changes as the day progresses, so we ask it in the morning and again just before your session. A coach / physiologist examines your ride data to assess whether you are hitting those numbers; or aren’t (Mujika, 2017). Spoked does that as well. This isn’t at random and the questions are based on the information most valuable to coaches. In future posts we’ll dissect the individual areas we have incorporated into our algorithms and why they are important. Some of the features I mention in this article aren’t live yet, but are planned for our next version update. I will highlight if this is the case in this post.

So, as well as subjective feedback, we also monitor how well you are able to follow and complete the set sessions. We do this in a variety of ways. Firstly, we check that you are actually able to achieve the power or heart rate we have set you. These are based on your maximal power numbers you have recorded previously for specific periods of time (10sec, 30sec, 1min, 5min, 12min, 20min, and 60min), called ‘benchmark’ scores. If you consistently ride higher or lower scores, then we will adjust them so they are in line with what you can actually do. We also ask you how hard you felt the session was. If it felt a lot harder or far easier than planned, then an adjustment is needed (Perez-Landaluce et al., 2002).

“Then we check if you spent the correct amount of time in the prescribed zones; basically, how well you stuck to the plan (Borresen and Lambert, 2009). For this we give you some leeway and a score showing you how well you rode the session.”

This is important for two reasons: if you are going too hard then the following training sessions might not be suitable anymore. Similarly, if you didn’t go hard enough then the training response won’t be as planned (Borresen and Lambert, 2009). It is also a very nice and clear way to show you whether you have followed the prescribed plan, as that isn’t always easy and can take some practice. It will naturally take a certain amount of time for you to be able to switch between zones, i.e. we don’t expect you to immediately jump from 100 watts to 300 watts, but appreciate this will take a few seconds, which we have factored in. When it comes to using heart rate, the lag between applying a higher power output and your heart rate response is measurable (Bearden and Moffatt, 2001). Basically, it’s delayed, but because we know that, we will take that into account and allow for the required time delay in our calculations. This might seem like a small thing, but is actually very important when it comes to us assessing how ‘well’ you have ridden a session. And over time you should be able to see how much you are improving your ability to follow a plan by using our clear data display graphs.      

“Naturally we put a lot of emphasis on the correct training, but what if I told you that what you do between training sessions is actually more important (Tuomilehto et al., 2017)?”

One thing most of us are guilty of is being very particular and committed to training, but maybe slacking on what happens in between. Most of us would really not want to miss a session, or even still do it if we’re not feeling like it, but many would easily skip a meal or sleep a bit less if required. Actually, in the long run that will have more of an impact on our bodies and performance than missing the odd training session (Bishop et al., 2008). Recovery is a very large area that is continually being researched. You could follow the best and most suitable training plan that our brilliant coach Rich and all of us here at Spoked have painstakingly put together, but not reap the corresponding rewards if you don’t follow the best pre-exercise and recovery procedures. So we start by asking you when and how long you want to train and then design your week so that you aren’t doing sessions back to back that might mean you haven’t recovered properly. The adaptations from any training stress develop in the time periods between the training stimulus, and recovery is often overlooked. The very basic supercompensation model shows that if you leave the correct amount of time between sessions you should always begin your next session slightly improved (or at least not in a reduced performance state; Shea, 2010). Recovery nutrition is a very big part of this and, along with tapering, is a very complex issue and something I am looking forward to presenting to you in more detail in future blogs.

Our fluid plan uses your daily feedback and ride numbers to track whether the training is too hard, too easy or on the money. A couple of very good or bad sessions won’t necessarily change your plan, but performing consistently above what we have set, or not being able to reach your targets will trigger a re-evaluation of your plan. This is in tandem with your daily feedback on how you feel and how you slept.

“Many of us are chronically tired and don’t sleep enough without even realising it (Mukherjee et al., 2015). But sleep is the most valuable recovery and training tool that you have at your disposal and it’s free; miss it at your peril!”

We will ask you every day how long and, crucially, how well you slept. Based on when you went to bed, how long it took you to fall asleep and how often you woke up, we calculate your sleep quality score. 100% is the desired goal, but research has shown that the average is far lower than that (Tuomilehto et al., 2017). Elite athletes often score well, but they mostly don’t have full-time jobs as well. We have a cut-off point and if you fall below it, we will ask if you want to maybe change your session to a less intense one if you are feeling the effects. You probably will be fine, but the option needs to be there in case you aren’t. If this happens repeatedly, you don’t get a choice and your training intensity is reduced. There is clear evidence that reduced sleep quality affects performance. In fact, this is such a big and fascinating topic that we will be writing a separate post about this, and especially delving into sleep trackers and whether they work as well as they claim (spoiler: they do not! [Ameen et al., 2019]).

In addition to asking how you slept, we also want to know your physical and mental freshness. Why would we want both?

“Research has clearly demonstrated that mental fatigue, even if physically fresh, affects your ability to perform (Marcora et al. 2009)”.

Importantly, it increases your perception of effort, so if you are trying to follow a session based on RPE you will probably not do as well, or perceive it as harder compared to being in a good mental place; not very pleasant. Mental fatigue is real, and having to concentrate hard for prolonged periods of time (sitting in front of a computer for example) will lead to it.

This shouldn’t be mistaken for being ‘soft’ or anything silly like that. This is training optimization and ultimately will lead to far more gains than just grinning and bearing it. The old school has no place here. Fatigue in general is a very contested topic, and I will present the main points in a separate post.

In the next paragraphs I will outline all the changes we have planned for future versions, but aren’t quite available yet. When I came on board last year, Rich and I sat down and he asked me to make a list of all the things that we should include and what we should ask our athletes. I was impressed by his drive to create such a sophisticated platform, and although my list was long and resulted in A LOT of work, he agreed to include every single one (much to our coder Lee’s dismay, who has to write the codes and algorithms for the app. He loves it really!).  

For example, for a big update in our next version we have planned to assign riders into 3 categories based on their ability, experience and current results. Nothing new about that, but what we have taken into account is that the ability to perform high intensity and long duration work varies significantly between these groups. Recovery periods need to be longer if you’re a beginner, the training cycles are shorter (2 weeks on, 1 week off instead of 3 weeks on, 1 week off), and the taper for an event can be extended (Thomas and Busso, 2005). Importantly, you can decide whether this works for you or not.

“Some riders adapt far quicker and can go for higher intensities with less rest, and others might be the opposite, despite their experience and training.”

This doesn’t mean you won’t improve as much; it just means the improvement is optimised for you, and importantly you won’t be digging yourself into a hole and end up fatigued and injured, or both.

We have lots of female athletes that are following our plans, and their menstrual cycle may affect their ability to perform at the same level as other times during the month (de Jonge, 2003). I still remember my lecturer who taught one of my favourite topics of Environmental Physiology warning us about core temperature variations at different parts of the menstrual cycle. If conducting any kind of work in the heat and assessing core temperature this needs to be taken into account, as the core temperature rises more at certain phases of the menstrual cycle. Severity of menstruation can also have other effects, although these do not affect everyone in the same manner (Bruinvels et al., 2016). Therefore, in the next version of Spoked, when setting up your training plan, we will ask about the timing of the menstrual cycle and if it has impacted performance in the past. We use this to alter the sessions if needed, but also to give advice before training and competition, especially in hot environments.

We here at Spoked appreciate that you might not want to always just follow a set training plan, might do other sports, go on holiday and sometimes might be ill. A rigid plan will not allow for variation as a result of any of these, but we are making sure that these will be taken into account. Although most of us primarily love cycling, a lot of us do other sports (they do exist I heard….), and if they are intense enough to cause some fatigue or soreness, we will factor that in and adjust your cycling training accordingly. Especially the impact of any other activity you class as ‘intense’ (running, hiking, gym etc.) will result in a reduced intensity cycling session that follows (Barnett, 2006).  

“Cross training is beneficial, but needs to be included in the weekly training load.”  

Who doesn’t love a café ride or just a nice ride out with your mates without any set power or goal? Everyone does, right?! Or you might commute to work via bike. Therefore, we’ve factored in the ability to set ‘free rides’ when you design your week. You tell us how long and hard you think they might be, and we adjust the rest of the training to match your goals. Once you’ve completed three free rides, we analyse them, and you tell us how hard you think they were. Harder than you thought / planned? No problem, the following hard session will be reduced as a result.

Going on holiday? Good for you! Some of us actually go on non-cycling trips, which will sound bizarre to some of you, but trust me, it happens. All you need to do is tell us how long you’re going for, if you think / know what kind of activity you might be doing, and we will pause or amend your plan. This way, when you’re back, we will ease you back in instead of going all-out straight away.

Unfortunately, we all get ill sometimes. The severity will dictate if we need time off and how much. Maybe you can still do some light training, but maybe you need to be off the bike for a whole week or more (Weidner and Sevier, 1996). Our platform will guide you through the options and importantly give you information about why it might be bad to continue training, and reassurance that not all gains will be lost.

“Did you know that in well-trained athletes a 2-week break does not really cause a significant drop in performance (Chen et al, 2021)?”

In some cases it has actually shown to be beneficial, as lots of athletes are somewhat over-reached / slightly overtrained. Strength or maximal power output is almost entirely maintained, and the very small drop in aerobic fitness is related to those changes inside your body that reappear very quickly. The best example for this is a slightly elevated heart rate after a break, which is related to a decrease in plasma volume (basically how much of your blood is water). You can lose approx 500ml in a couple of weeks, and as a result of this your heart rate will be slightly elevated. This is typically gained back within 4 days of commencing training again (Mujika and Padilla, 2000).  

If you managed to stay with me until here I commend you! This was quite a long post, but we wanted to make sure we explain in good detail our point of view on how to create the best possible training plan for you, and what we’ve done to achieve this. As I mentioned, many of the items in this post deserve their own feature, and in future blogs I will tackle many controversial topics and often-asked questions.

Until then: happy riding from your in-house physiologist Dr Nic!

PS: feel free to send me any questions, thought or suggestions via my email at


Physical activity apps:

Bondaronek, P., Alkhaldi, G., Slee, A., Hamilton, F.L. and Murray, E., 2018. Quality of publicly available physical activity apps: review and content analysis. JMIR mHealth and uHealth, 6(3), p.e53. View research.

Rating of perceived exertion (RPE):

Herman, L., Foster, C., Maher, M.A., Mikat, R.P. and Porcari, J.P., 2006. Validity and reliability of the session RPE method for monitoring exercise training intensity. South African Journal of Sports Medicine, 18(1), pp.14-17. View research.

Perez-Landaluce, J., Fernandez-Garcia, B., Rodriguez-Alonso, M., García-Herrero, F., García-Zapico, P., Patterson, A.M. and Terrados, N., 2002. Physiological differences and rating of perceived exertion (RPE) in professional, amateur and young cyclists. Journal of sports medicine and physical fitness, 42(4), pp.389-395. View research.


Bishop, P.A., Jones, E. and Woods, A.K., 2008. Recovery from training: a brief review: brief review. The Journal of Strength & Conditioning Research, 22(3), pp.1015-1024. View research.

Heart rate response / lag to exercise:

Bearden, S.E. and Moffatt, R.J., 2001. V̇ o 2 and heart rate kinetics in cycling: transitions from an elevated baseline. Journal of Applied Physiology, 90(6), pp.2081-2087. View research.

Hunt, K.J., Grunder, R. and Zahnd, A., 2019. Identification and comparison of heart-rate dynamics during cycle ergometer and treadmill exercise. PloS one, 14(8), p.e0220826. View research.

Jeukendrup, A. and Diemen, A.V., 1998. Heart rate monitoring during training and competition in cyclists. Journal of sports sciences, 16(sup1), pp.91-99. View research.

Zakynthinaki, M.S., 2015. Modelling heart rate kinetics. PloS one, 10(4), p.e0118263. View research.

Coach / athlete interaction:

Davis, L., Jowett, S. and Tafvelin, S., 2019. Communication strategies: The fuel for quality coach-athlete relationships and athlete satisfaction. Frontiers in psychology, 10, p.2156. View research.

West, L., 2016. Coach-athlete communication: Coaching style, leadership characteristics, and psychological outcomes. View research.

Data analysis and training quantification:

Borresen, J. and Lambert, M.I., 2009. The quantification of training load, the training response and the effect on performance. Sports medicine, 39(9), pp.779-795. View research.

Jobson, S.A., Passfield, L., Atkinson, G., Barton, G. and Scarf, P., 2009. The analysis and utilization of cycling training data. Sports medicine, 39(10), pp.833-844. View research.

Mujika, I., 2017. Quantification of training and competition loads in endurance sports: methods and applications. International Journal of Sports Physiology and Performance, 12(s2), pp.S2-9. View research.

Sylta, Ø., Tønnessen, E. and Seiler, S., 2014. From heart-rate data to training quantification: a comparison of 3 methods of training-intensity analysis. International journal of sports physiology and performance, 9(1), pp.100-107. View research.

Mental fatigue and performance:

Brownsberger, J., Edwards, A., Crowther, R. and Cottrell, D., 2013. Impact of mental fatigue on self-paced exercise. International Journal of Sports Medicine, 34(12), pp.1029-1036. View research.

Marcora, S.M., Staiano, W. and Manning, V., 2009. Mental fatigue impairs physical performance in humans. Journal of applied physiology. View research.

Smith, M.R., Marcora, S.M. and Coutts, A.J., 2015. Mental fatigue impairs intermittent running performance. Med Sci Sports Exerc, 47(8), pp.1682-90. View research.

Van Cutsem, J., Marcora, S., De Pauw, K., Bailey, S., Meeusen, R. and Roelands, B., 2017. The effects of mental fatigue on physical performance: a systematic review. Sports medicine, 47(8), pp.1569-1588. View research.

Supercompensation model:

Gambetta, V., 2007. Athletic development. Champaign, IL: Human Kinetics View research.

Shea, Jason. II, G., Example of overtraining, under-training, and ideal training to recovery situations. View research.

Thomas, L. and Busso, T.H.I.E.R.R.Y., 2005. A theoretical study of taper characteristics to optimize performance. Medicine and science in sports and exercise, 37(9), pp.1615-1621. View research.

Sleep and recovery:

Mukherjee, S., Patel, S.R., Kales, S.N., Ayas, N.T., Strohl, K.P., Gozal, D. and Malhotra, A., 2015. An official American Thoracic Society statement: the importance of healthy sleep. Recommendations and future priorities. American journal of respiratory and critical care medicine, 191(12), pp.1450-1458. View research.

Tuomilehto, H., Vuorinen, V.P., Penttilä, E., Kivimäki, M., Vuorenmaa, M., Venojärvi, M., Airaksinen, O. and Pihlajamäki, J., 2017. Sleep of professional athletes: underexploited potential to improve health and performance. Journal of sports sciences, 35(7), pp.704-710. View research.


Chen, Y.T., Hsieh, Y.Y., Ho, J.Y., Lin, T.Y. and Lin, J.C., 2021. Two weeks of detraining reduces cardiopulmonary function and muscular fitness in endurance athletes. European Journal of Sport Science, pp.1-8. View research.

Mujika, I. and Padilla, S., 2000. Detraining: loss of training-induced physiological and performance adaptations. Part I. Sports Medicine, 30(2), pp.79-87. View research.

Mujika, I. and Padilla, S., 2000. Detraining: loss of training-induced physiological and performance adaptations. Part II. Sports Medicine, 30(3), pp.145-154. View research.

MUJIKA, I. and Padilla, S., 2001. Muscular characteristics of detraining in humans. Medicine & Science in Sports & Exercise, 33(8), pp.1297-1303. View research.

Sleep tracker accuracy / effectiveness:  

Ameen, M.S., Cheung, L.M., Hauser, T., Hahn, M.A. and Schabus, M., 2019. About the accuracy and problems of consumer devices in the assessment of sleep. Sensors, 19(19), p.4160. View research.

Haghayegh, S., Khoshnevis, S., Smolensky, M.H., Diller, K.R. and Castriotta, R.J., 2019. Accuracy of wristband Fitbit models in assessing sleep: systematic review and meta-analysis. Journal of medical Internet research, 21(11), p.e16273. View research.

Miller, D.J., Lastella, M., Scanlan, A.T., Bellenger, C., Halson, S.L., Roach, G.D. and Sargent, C., 2020. A validation study of the WHOOP strap against polysomnography to assess sleep. Journal of Sports Sciences, 38(22), pp.2631-2636. View research.

Menstruation effect on performance:

de Jonge, X.A.J., 2003. Effects of the menstrual cycle on exercise performance. Sports medicine, 33(11), pp.833-851. View research.

Bruinvels, G., Burden, R., Brown, N., Richards, T. and Pedlar, C., 2016. The prevalence and impact of heavy menstrual bleeding (menorrhagia) in elite and non-elite athletes. PLoS One, 11(2), p.e0149881. View research.

Common cold and exercise:

Weidner, T.G. and Sevier, T.L., 1996. Sport, exercise, and the common cold. Journal of athletic training, 31(2), p.154. View research.

DOMS and recovery:

Barnett, A., 2006. Using recovery modalities between training sessions in elite athletes. Sports medicine, 36(9), pp.781-796. View research.