I watched AC Milan lose the 2005 final, then I started reading the numbers
I spent my teens in FIFA career mode studying stats instead of playing, so I started building my own models. Fifteen years, a master's from KLU in Germany, data science specializations, and 38 papers later, those models are predictapp. I built it, Dex turns it into a clear read, and you make the call.
Why I built it
Most prediction content frustrated me. Either it was confidence with no method behind it, or it was so technical you needed a stats degree to use it. I wanted the read, not the hype.
So I did the work. Fifteen years of building, testing, and breaking models. A master's from KLU in Germany and data science specializations to do it properly. 38 scientific papers to get the methodology right.
predictapp is what I wish I had as a fan: the true probabilities, the fair price, the context, and the honesty about what the model gets wrong.
I built it, Dex reads it, you make the call
Dex is the lime octopus, my analytical partner inside predictapp. I explain the why. Dex surfaces the what: thousands of data points turned into one clear read, no hype. You decide what to do with it, and Coach Dex grades your saved bets so you can see your real ROI.
This is not generic AI guessing matches. It is a model I built, tested, and explain, with the misses included.
I publish the misses
Anyone can post their wins. I show the whole record. When the model says it is confident, here is how often it is actually right, measured on finished matches only. You should not take my word for it, so here it is.
What you get
- ✓The true probability of every market, each with its fair price
- ✓Both teams to score, goals, clean sheets, corners and more in one card
- ✓Team strength, form, and head to head context
- ✓A track record you can check, not just trust