Battle of cheeses

Tournament via Instagram stories

Abstract

Dairy consumption in Brazil is determined by family income, consumer profile and their eating and consumption habits, as well as their origins, beliefs, culture and ideologies. In 2018, the IBGE revealed that Muçareala is the most consumed cheese in Brazil ($.809kg/person/year$), followed by Minas cheese ($.565kg/person/year$), Requeijão ($.285kg/person/year$) , Prato ($.267kg/person/year$) and Parmesão ($.074kg/person/year$). On the other hand, works scraping on Twitter revealed that the most commented cheeses on this social network are, in that order, Coalho cheese, Requeijão, Cheddar, Prato, Parmesão, Muçarela, Swiss and Gorgonzola. However, the citing frequency does not necessarily reflects preference. Not even the consumption faithfully reflects the preference, after all, one consumes the cheese that fits in the family budget. To estimate the preference among the eight most popular cheeses in Brazil, this work carried out a $2^k$ bracket experiment, with $k=3$ phases (quarter-finals, semi-finals and finals). The experiment was conducted in the \emph{stories} of Instagram, using the \emph{poll} feature, and promoted entirely random matches. The result differed from the literature for consumption and citation on Twitter. The resulting grouping was: in first place Parmesão (55.9\%); second, Muçarela (47.1\%); third, Prato (43.6\%) and Coalho (34.2\%); fourth, Swiss (45.5\%), Cheddar (41.3\%), Requeijão (32.3\%) and Gorgonzola (29.4\%). Future work should develop the inferential statistics for tournament bracket experiments.