A big round of applause for our readers!
--------------------------------
A great many of our readers took a chance and bet on the presidential election results (see here). The polls predicted a runoff between Emmanuel Macron and Marine Le Pen; many of you listened to them and so qualified for the runoff.
Seven of our readers scored better than Ipsos: our warmest congratulations! All the polling institutes had clearly underestimated Mr Macron's score in the runoff, especially those that did not conduct a poll after the debate between the two rounds. Our winners managed to foresee this spectacular rise.
The prize goes to Élisabeth Pietravalle, with a score of 8.39. The average error per candidate is less than one point, which is remarkable. Jean-Bernard Eytard, with 10.09, and Hervé Delay, with 10.88, were barely worse.
Tangente thanks all the participants for their acumen!
-
Prizes, prizes!
--------------------------
A shower of prizes has fallen on a great many participants, offered by Tangente, by Éditions POLE, and also by our partners Hewlett Packard, Éditions Belin, Cassini, Le Pommier and Wartberg, whom we would like to thank.
The winner, Mrs Pietravalle, receives an HP Prime and one hundred euros' worth of Éditions POLE gift vouchers. Her two runners-up receive Jérôme Brunet's magnificent book on fractal art and a few gift vouchers. The next two will receive the Bibliothèque Tangente issue Mathématiques et Politique. Numerous trial subscriptions to digital Tangente, as well as books on mathematical culture or logic puzzles, will be given out all the way down to the candidate ranked 85th in the competition.
-
Polling institutes put to the test: Ipsos ahead of Ifop and Elabe
------------------------------------------------------------
Never during a presidential election has the role of polls been so called into question. One candidate even wanted them banned when the polls were not in that candidate's favor, only to shelve that demand once the tide turned in that candidate's favor…
They have been criticized, often rightly, for lacking transparency in how they are conducted: their method departs markedly from the mathematical theory of random sampling. Furthermore, adjustments are applied that the various institutes keep secret, just as a great chef would guard their recipes.
While these remarks deserve consideration, we tested the nine institutes that conducted polls for this election using the same rule applied to our readers. Far be it from us to compare the two! Our readers do not have the investigative resources of the institutes. That said, they certainly took their cue from the polls.
All the institutes got through the first round by predicting Mr Macron and Mrs Le Pen's qualification in the right order (only BVA put them level). Note the remarkable accuracy of Elabe, whose average error does not exceed half a point. In the runoff, all the institutes greatly underestimated Emmanuel Macron's score; the credit goes to Ipsos and Ifop, who conducted polls on the last Friday and placed the future president, admittedly too low, at 63%, but above their competitors.
Here are the results, in accordance with the competition rules:
