The main difficulty in predicting the Austrian Election is that there are practically no polls in the final week. This means a projection of trends has to be inferred with no real data to identify whether the trends have continued in the last week.
Well let’s have a go anyway. The model is built assuming d’hondt PR with a 4% cut-off. 10,000 simulations of the election are run based on the latest polls and the aggregated results displayed.
And now some probabilities