- Models will always remain partly unreliable
- Changing only the smallest factors can change your whole model
In the summer vacation I spend some time in developing a model that could calculate the difficulty level of all juggling tricks. I wanted to do this because I’m part of the rules committee of the Dutch Juggling Championships. We have to judge how difficult certain tricks are, but this is highly subjective. I wanted to create a more objective system. Already on high school I did some research about this. Now I took the time to make it in a working model in order to further test and develop it.
Still the outcomes aren’t reliable but would match the subjective values to some extent. The problem is to judge were the subjective values are wrong or the model values. I’m not going to invest more time in making it more perfect, because I know it’s never going to be fully reliable and judging the tricks every year doesn’t take so much time as it would for me to develop this model.
The user of the model takes some knowledge about juggling. In juggling you have a mathematical way of notation for certain tricks. The model could read these notations and give a certain difficulty value. For example try to input the following notation: (6x,4)(4,6x) and click on the numbers itself. I will display certain information about this juggle trick which are all calculated by the model.
In this side project I could use the knowledge I gained from the basic course Modeling I followed last year. I was already aware of the benefits of using a system. By making this model I know that it takes a lot of time to make your model even a little trustworthy. I’m not sure if I’m going to use this in future because for me it lacks to much reliability.
