One of my pet peeves is the candidate effect on electoral results. A prominent example is Lukáš Vlček, former government minister. He gained substantial support around Pacov, the city where he served as mayor for 15 years. Within the party of mayors, his local effect is by far the most significant.

Hot spot analysis reveals some additional details. Surprisingly, the former government coalition lost support in the automotive heartlands of Mladá Boleslav (Škoda) and Kolín (Toyota) in the central part of the country, while they gained support in the Silesian "left-behind" areas around Ostrava

The support change map of the former government coalition shows a clear pattern. They gained support where they were already strong: in regional cities and their suburban areas. They lost support where they were already weak: in the peripheries along country and regional borders.

The plot shows changes in electoral segregation between government and opposition voters over the last 30 years, with increasing segregation post-2002. However, compared to countries like the US, spatial polarization still remains relatively modest.

This increasing spatial polarisation is evident when we calculate segregation indices. Czech people are increasingly surrounded by voters with similar electoral behaviour, reducing contact with the 'other side'.

Two main party blocs campaigned against each other: the parties of the former government and the (probable) new government. The margin of victory map reveals clear spatial polarisation between the blocs. Half of the Czech population lives in areas where one bloc won by over 20 percentage points.