glcd
Member
- Joined
- March 19, 2016
- Posts
- 21
Hi guys,
I haven't really posted here before but I've been working on building a simulation model for ESC - this takes into account betting odds, past voting, and also allows your own judgement, in order to simulate a potential outcome for this year's contest. It allows allows you to save down the daily betting odds which may be useful/interesting for future years.
I thought I'd share the current version to see if it was useful at all (building models like this forms part of my day job so thought it might be an interesting application of my expertise as I'm trying to get into the betting side of ESC at the moment).
It's in a bit of a development phase at the moment so any feedback/comments on improvements or glitches are welcome! The main issue is that it's built using the old voting system - the simulation uses historical results as part of its analysis and as there's no historical data for the new voting system that just couldn't be implemented in any meaningful way. I think the difference between old/new voting systems applied to past years have shown that there are rarely major differences between the two so hopefully this is reliable. Any suggestions for how to try and adapt this for the new voting system greatly appreciated! For next year, once we see how this all works it should be much easier.
(Also Romania are built into the model as I started on this before their disqualification but I've changed it so that it's impossible for them to receive any points so don't worry about that).
The model is available for download on (it's all on an Excel file) the following link:
https://www.dropbox.com/s/fv834d1obvvjc5f/ESC16_Simulation_v.1.1.01.xlsm?dl=0
(It looks like you need to download it to run the simulations rather than opening it in a browser - it's not riddled with viruses I promise! You'll also make sure you 'Enable Content' to allow the simulation to run!)
Thanks in advance for any feedback!
I haven't really posted here before but I've been working on building a simulation model for ESC - this takes into account betting odds, past voting, and also allows your own judgement, in order to simulate a potential outcome for this year's contest. It allows allows you to save down the daily betting odds which may be useful/interesting for future years.
I thought I'd share the current version to see if it was useful at all (building models like this forms part of my day job so thought it might be an interesting application of my expertise as I'm trying to get into the betting side of ESC at the moment).
It's in a bit of a development phase at the moment so any feedback/comments on improvements or glitches are welcome! The main issue is that it's built using the old voting system - the simulation uses historical results as part of its analysis and as there's no historical data for the new voting system that just couldn't be implemented in any meaningful way. I think the difference between old/new voting systems applied to past years have shown that there are rarely major differences between the two so hopefully this is reliable. Any suggestions for how to try and adapt this for the new voting system greatly appreciated! For next year, once we see how this all works it should be much easier.
(Also Romania are built into the model as I started on this before their disqualification but I've changed it so that it's impossible for them to receive any points so don't worry about that).
The model is available for download on (it's all on an Excel file) the following link:
https://www.dropbox.com/s/fv834d1obvvjc5f/ESC16_Simulation_v.1.1.01.xlsm?dl=0
(It looks like you need to download it to run the simulations rather than opening it in a browser - it's not riddled with viruses I promise! You'll also make sure you 'Enable Content' to allow the simulation to run!)
Thanks in advance for any feedback!