In a recent paper published at Journal Selva Brasilis, Jaramillo (2010) argues for the rejection of the Valdomiro´s law (which states that tight shirts and hair-gel are the main constraints to winning the world cup). This reply shows that Jaramillo´s main conclusions were driven by sample selection bias.
Jaramillo (2010) uses France (1998), Brazil (2002) and Italy (2006) to reject Valdomiro´s law. However, this dataset is not random, which points out to the existence of sample selection bias. In 1998 the best French player was from Argelia and was bald. There is no way to say that bald men use hair gel. It is the opposite; France (1998) confirms the Valdomiro´s law: the gayest Brazilian team in history lost (Leonardo ted bear was the leader of that team). In 2002 the Brazilian team back on track again won the world cup. What was the symbol of that team? Answer: the ridiculous hair cut of Ronaldo fenomeno. Again Valdomiro´s law correct predicted the elimination of the metrosexual English team in 2002. The conclusion of professor Jaramillo about Italy in 2006 suffers from the same bias.
In conclusion, Valdomiro´s law performance is confirmed for the whole history of the world cup. As soon as we apply Heckman procedure to correct for sample selection, Jaramillo´s results converge to Valdomiro´s law.
Doutor Valdomiro Pinto, Ph.D., ouviu falar no Heckman procedure e achou que fosse um tipo de carrinho comum no futebol Austriaco.
No comments:
Post a Comment