After having scored only 78% accuracy in the first attempt to hack the Titanic data set on Kaggle, I tried to come up with a way that scored better than my last try. This resulted in the women-child-model which sored an accuracy of 82% without using any machine learning algorithm.