3 ms·
I went one step further and actually implemented an MCMC model and ran it for a while. It did turn a slight profit on low stakes games for the 3 months that I r
by phreeza 11y ago
I went one step further and actually implemented an MCMC model and ran it for a while. It did turn a slight profit on low stakes games for the 3 months that I ran it. I stopped and did not move on to higher stakes games because I live in Germany and there appears to be no simple way for me to legally participate in these games and then withdraw my money. I am keeping the scripts ready for when the fad invades Germany, though!
I agree on the stats part. Scraping them from various sources was definitely the most painful and limiting part of the project, and I was only able to obtain a limited amount of data even then. If I could get a complete and large dataset to work with, I would be tempted to give it another shot, just to see how far more data takes me.
- ssharp 11y agoWhen I was looking at this last year, there were several paid options to get direct access to datasets and/or API's, without having to scrape. Out of curiosity, what sport(s) were you modeling?
- phreeza 11y agoI was modeling football (soccer). All the commercial data sources I found had no publicly listed prices, only 'contact us for a quote', so I couldn't be bothered because it seemed likely to be too pricey.