creating stats from results
This idea started a couple of years back when I was playing Pro Cycling Manager and saw stats for riders that didn't seem right. Several community groups had created their own database with riders and stats, but they all were subject to the preferences and perhaps biases of the creators. This started to make me think if I could create the stats from the actual results over the last couple of years.
Long story short, I think there is definitely something there, but it still needs work. Here is one of my first post on this subject to start the sharing and discussing on this subject. Again by no means is this perfect, but it gives you an idea. For this overview I used the accumulate point in a year for a specific area. Filter for instance on a certain rider to see his stats evolution.
All predictions that are done on my twitter account with the according score's;
Today´s stats on 2-3-2017, 3-6-2017, 2-3-2018, 1-7-2018, 1-8-2018, 1-1-2019 &1-1-2020 in this file;
All stats are based on 3 years of results avg; 1x last year, 1/6 x 2 years ago, 1/12 3 years ago
I have a weak spot for data and cycling. This is an attempt to combine the 2 into 1