Started messing around with Slovenian football league tables after my buddy kept losing bets. Figured I’d try predicting match winners using nothing but the standings. Grabbed my laptop at the kitchen table with cold coffee – here’s how it went down.
Scraping the Raw Data
First I needed current league stats. Searched “Slovenian PrvaLiga table” online. Found one site showing positions, games played, wins, draws, losses, goals scored/conceded, and points. Copied-pasted the whole mess into a spreadsheet. Took three tries because the formatting screwed up halfway through.
Stared at the jumbled data. Realized I had to:
- Clean up extra spaces making cells look funky
- Fix headers like “PTS” cutting off when pasted
- Check numbers – saw Celje had GF/GA swapped originally
Testing Prediction Methods
Wanted simple ways any dude could use without math PhDs.
Attempt 1: Pure Table Position
Pitted #1 Celje against #10 Radomlje. Easy right? Higher position should win. Checked last five actual results… Celje won only twice. Total mess. Table rank alone sucked for predictions.
Attempt 2: Goal Difference Power
Sorted teams by goal difference instead. Top two were Celje (+40) and Olimpija (+23). When they played each other? Drew 1-1. My method crashed harder than a Windows 98 computer. Needed deeper digging.
Attempt 3: Home/Away Split Check
Added new tabs for home/away performance. Saw Domzale won 70% at home but were trash traveling. When predicting Mura vs Domzale:
- If Domzale home? Tipped them based on home form
- If Mura home? Backed them since Domzale away was awful
This actually lined up with 4/5 real outcomes. Finally something worked!
Putting It All Together
Tested by “predicting” last weekend’s games with my three methods:
- Pure position: Got 1/5 right (lol)
- Goal difference: 2/5 (still bad)
- Home/Away split: 4/5 (decent!)
Moral? Football’s random as heck. But checking how teams perform home vs away gives you a fighting chance. Still lost imaginary money on that one wrong pick though.
