Forecasting: 2017 French Presidential Election
2017 French Presidential Election Poll Results versus Social Media Influence (SMI) Ratings for the main candidates, data from February 2017, prior to April 2017 first round voting Source: ZettaCap and OpinionWay
ZettaCap’s Social Media Influence or SMI forecasting method was used in the 2017 French Presidential Election. When our first forecast was made, in December 2016, we declared that Macron would be France’s ‘Next President’. Considering the state of the race at that time, such a forecast was completely out-of-consensus. The betting market, Predictit, placed the probability of Macron winning at only 8% while Fillon was at 61% and Le Pen at 32%. In other words, the markets did not view Macron as a serious candidate. Polls likewise had him as a relative longshot. Depending on the poll at that time, he was considered to be in third or fourth place, and normally polling at about half the level of the lead candidate.
On top of these negative indications, Macron was considered a political novice and had just started a new political party. Such a situation would not bode well for a forecaster backing him. However, ZettaCap’s SMI consistently pointed to Macron as the candidate with the highest influence on social media and our prediction was that he would ‘surge’ in polls – which he soon did.
As April’s first round approached, SMI picked up considerable, albeit late, interest in Melenchon, the far-left candidate, and predicted a substantial boost for him in the last stretch of the campaign. In fact, Melenchon performed exceptionally well after ZettaCap pointed out his sharp SMI increase, shocking many. ZettaCap also was the first known quantitative forecast to predict that Macron would win the first round of voting, while the consensus saw Le Pen winning the first round while losing the second.
For the second round, SMI’s forecast beat every major poll. Though having the highest known percentage vote forecast for Macron, it was still slightly under the actual result.
Important because:
- Social Media Influence (SMI) making an early and out-of-consensus forecast,
- First known forecast of Macron becoming president,
- First known forecast of Macron gaining the most votes in the first round of voting,
- Most accurate forecast for second round of voting, beating every known poll.