CNC's response to Atlas Intel regarding survey accuracy

The Centro Nacional de Consultoría (CNC) presents its response to Atlas Intel.

1. No polling firm, including Atlas Intel, reports its results based on the total eligible voting population, regardless of the length of the explanation or the sources cited.

   
   

2. The figures published on the cover of SEMANA magazine do not appear anywhere in their self-assessment.
   

   

   
   

3. If you want to evaluate the accuracy of the candidates' predictions, you have to consider the estimated vote in the primary elections. For example, Atlas estimated 12% in the primary election and 44% for Paloma Valencia, meaning they estimated Paloma would receive 5% of the vote. The CNC estimated 28.8% in the primary election and 39.8% within the primary election, meaning Paloma would receive 11.1% of the vote.

   In reality, Paloma received 3,236,286 votes out of 20,900,614, or 15% of the vote. Using this metric, the CNC's error is 4.1%, and Atlas Intel's is 10%. If the evaluation is done this way, which is how it should be, the true picture emerges.

That's the reality, don't overthink it.

Pablo Lemoine

President

Centro Nacional de Consultoría S.A.