Researchers in China use computer modelling to predict Trump election victory

Former US president Donald Trump is “more than 60 per cent” likely to beat Vice-President Kamala Harris and recapture the White House, according to computer modelling by academics from Fudan University in Shanghai.

The forecast was issued by the university’s Centre for Complex Decision Analysis (CCDA) on Sunday, just two days before Tuesday’s election showdown, amid a flurry of weekend polls that showed the race still neck-and-neck in the final sprint.

The researchers – led by Tang Shiping, from Fudan’s School of International Relations and Public Affairs – used a computational simulation to predict vote shares and electoral college votes in eight key states.

The chosen states – which all registered a gap between Republicans and Democrats of less than 5 per cent in the 2020 presidential election – included the battleground states of Arizona, Georgia, Michigan, Nevada, North Carolina, Pennsylvania and Wisconsin.

The eighth state selected for the modelling was Florida, where US presidential elections once were decided by the slimmest of margins. In recent years, the sunshine state has proven reliably Republican.

The technique, called agent-based modelling (ABM), mimics the actions and interactions of individual agents within a system to analyse complex behaviours and predict outcomes.

  

​Read More