Thomas Gossmann

Unicyclist. Artist. Developer.

Diagnostics: Time-Analysis for Unicycle Freestyle Routines

Analytics in freestyle are rare, mostly the only given measurement that’s been done during practice is counting dismounts and see if this number decreases as the competition approaches. There is so much more room for more analytics on freestyle routines to help during practice to measure a certain aspect of the routine and with multiple measurements over the whole trainings plan, you can document and compare these results to see if there is an improvement or not. Also to use those analytics as hint to change something in your routine, because it’s not catchy as you thought it is. Here is another test that can be done on a routine, which is measuring the time which is spent on different aspects in ones routine.

Before I start getting on topic, let me first explain a little how those tests can be created. All these tests mostly have in common that they quantify a certain aspect of the discovered subject (E.g. you can count the number of tricks in a routine). It must be explained which input data you can gather (the method) and all possible further calculations that can be done with that (the result). Lastly you can interpret the results for further usage (the discussion). Additionally when creating such a test, you must discuss its credibility in regard to objectivity, validity and reliability.


For freestyle, we are given a maximum time limit we can use for routines, the difference is how we use our given time. We can use the time to do a 3-minute stillstand or a firework of tricks, which both might be impressive but not very spectacular in front of an audience, that expects art and entertainment. There are four different aspects in which you can spent your time.

All time that is spent for tricks
All time that is spent on body technic as defined here.
All time that is spent on moves, where the audience is able to correctly guess what’s coming next before it’s already happening (e.g. too many hops in a hopping combination or too long glides, where a viewer already guessed the spot on which the tipspin will occur). Typically those moves drain the attention of a viewer and bore them the longer this actions happens (excluded are those actions, which should happen at a certain time, e.g. a specific moment given by music – then the audience is eager to see if a rider matches action and music).
Void is everytime when the rider doesn’t act with it’s purpose on the self-given theme, e.g. the simplest case can be when the rider is just riding without doing anything, although this also can happen during a trick (e.g. during one-foot, especially when the overall skill-level is much higher). Void happens every time, when the rider is out-of-sync between his moves and the theme of the routine but more when the riding doesn’t catch the audience attention anymore (So very technical riding will for sure catch the audience attention and is as such excluded from this – however, when the tricks continue too long the audience will loose the theme, which needs to be prevented).


You need two things: A video of a routine and a stopwatch. Now watch the video at maximum four times. Each time you decide to measure one of the previously defined aspects. Start the video and hit the start/continue on the stopwatch when an action of the described aspect happens and push pause when that action of that aspect stops. At the end you get the total time, which the rider has spent for each aspect of the routine.


Note down each of the measured times and the actual time (RT) of that routine (not the maximum possible time) in a table and the first thing you can do is to calculate the percentage of each aspect in the routine. Here is a fictional example:

Total Time: 240 seconds (4 minutes)

Aspect Time [s] Relationship [%]
Tricks 140 58,3
Artistics 50 20,8
Filler 20 8,3
Void 70 29,1

That’s already a nice overview of what the routine may look like. We could also draw a graph to visualize the results. There is still more to calculate from the given input data.

Effective Use (EU)
The time which has effectively been used. It is a simple calculation: EU = Tricks + Artistics – Filler
EU Offset (EUO)
That’s the offset between the used time and the given time to see how much profit resp. waste the routine brought: EUO = EU – RT
EU Offset Index (EUI)
The index is the EU Offset in percent related to the actual time of a routine to make that number comparable across routines that have different durations: EUI = EUO / RT * 100
T-A Ratio
The ratio between tricks and artistics. A simple division: TA = Tricks / Artistics


With the calculated data, there are three values to keep a closer look at: The T-A Ratio, Effective Use and the Void and Filler value

1. T-A Ratio

When we design our routines we try to reach a 50:50 ratio between tricks and artistics. For sure, hitting the exactly that ratio is quite impossible, so everything between 40:60 and 60:40 is in an acceptable range. With another 5% of additional range at both ends (35:65 – 40:60 and 65:35 – 60:40), which I call the danger zone. That means a routine can work out to be catching and can keep up the tension while leaving that boundaries a routine becomes too much one-sided, either as X-Style or a theatre/dance (with a unicycle as prop).

2. Effective Use

Even we are only given a specific maximum time, it is about how we as riders use this time. Beginners will be tending towards a routine where they either present artistics or tricks but not both at the same time or even doing nothing by just following choreographed paths which a coach teached them (= the void). An expert rider is able to show artistics while executing tricks and this overlap allows them to use more of the given time (I personally see a huge potential here).

3. The Void and Filler

A time spent on the void is very boring for an audience and decreases the attractivity of a routine. Similarly a filler action also bores a viewer, because he already knows what action is ahead, which eliminates the surprising element. Both values should be zeroed out while practicing.


The credibility needs to be related to objectivity, validity and reliability. Also very important to adress are measurement errors.

Measurement Errors

As the time for each aspect is stopped using a timewatch, there is a delay when an action happens and when the user hits the play button on the stopwatch. Furthermore, this delay can sum up and alter the actual value. At the moment I’d say this measurement error appears to be in the range of 2 seconds, which is negligible as it will not have a significant impact on the total values (There is no data available yet to support this assumption).


Objectivity is realized through validity, unless the user is measuring wrong on purpose.


Validity is tricky. Especially when it comes to artistics. Some actions may be valid to time others won’t and this will differ across various users. Both measurements can be valid.


Reliability is established when a person measures the same times of the same video in multiple measurements (within an acceptable margin).

Real-Live Examples

Last but not least, I did some measurements on real routines to share with you and see what these values may look like, I collected the results in a google spreadsheet for you to inspect. As you can see, Kazuhiro and Janna (2014) both landed in the danger zone (at both ends) when it comes to the T-A ratio, however both routines were fantastic and had a great mixture, whereas Katrine actually was showing an X-Style to music with a costume. Interesstingly Janna improved very much from 2012 to 2014 and greatly improved her Effective Use time by decreasing the filler value and increasing artistics, which she demonstrated in parallel to executing tricks (She didn’t won, because the current rulebook favors tricks over artistics by a 2.5:1 ratio).

How to use this tests in your practice?

This test is very simple to use and to get the best out of this, you should define certain dates in your trainingsplan, when you will take a video and measure the mentioned times and calculate all the respective values. The results will help you to see if you are on track or if you need to modify your routine and help you with changing the choreography in the right direction. Let’s assume you have 4 dates for measurements in your trainingsplan, you can compare each result one-by-one over time and even plot them in a graph to visually inspect changes during your training and realize your improvements.


The tests are very simple yet powerful and provide you with rich data about your routine. It’s also very simple to use in your training. Additionally, the results are good candidates that can (with still unknown others) used to calculate an attractivity index for a given routine – I hope I once will read about such a test YOU described. For now, grab a video of a routine and try this test for yourself and play around with the results. Feedback is always welcome.