top of page

How to Fix Bad Data in TrainingPeaks

Training data spikes can be bloody annoying. Spikes can happen for a lot of reasons, such as wind, poor calibration, dying batteries, or just a bad day for equipment. Whatever the issue with your sensor, this spikey data will find its way through to training management software like TrainingPeaks that I use.


Bad data can affect things like peak performances and change the dynamics of the Performance Management Chart. This can lead to coaches and athletes making the wrong decisions on when things are going great or going poorly, so it’s important to identify and fix spikes in the data that are not real data.

[Scroll to data fix video down below]



One thing to remember about bad data: it is not TrainingPeaks’ fault. Bad data is usually coming from your sensor; TrainingPeaks is the vessel for analysis.

The other thing to remember: TrainingPeaks can help you!


Firstly, it’s important to note that not all data is bad; just because you performed poorly in a session or race does not mean it should be removed! Actually, I’d encourage you to keep as many of your poor performance in TrainingPeaks as possible–this is for when you are ready for a coach.


Check out all my TrainingPeaks Plans HERE


The second thing to note is that it can be difficult to identify “bad” data and just bad performance. It may take many hundreds or thousands of files to sift through before you can tell the difference.


However, once you notice or find this bad data, you will want to get it out of your TP account.

TrainingPeaks has a nice little feature that allows you to fix a data channel and remove it completely from a workout. This means that peaks, averages, and even whole rides can be recalculated to keep your charts ticking along nicely. As an example, you could remove the entirety of heart rate from a ride when the sensor’s battery died, or even simply just remove 5 seconds of a power spike at 3,000W so your 1 minute peak power is legit.


These little tweaks can save your TSS from a ride, which will recalculate all measures in your PMC (Performance Management Chart).


Since this trick is not completely obvious (and I always forget how to do it!), I wanted to make a video for everyone so it’s easy to sort out.


This is just my thinking and how I use TrainingPeaks, but you can probably find out more information in their help forum.


Either way, here are the 7 steps and a video of me fixing an athlete’s ride on video. In this video, I explain what I am looking for and how I identified the bad baits in this case.



How to fix TrainingPeaks data in 7 easy steps: