The calculation model for average scores to analyze the variation of Wordle game

Abstract

Wordle is a very popular puzzle game offered by The New York Times. Every day, the puzzle editor provides a five-letter word as a riddle. we regard the associated percentages of (1, 2, 3, 4, 5, 6, X) as statistical probabilities and propose a calculation model for average scores to analyze the variation of reported results. At the same time, we establish a probability model based on normal distribution to estimate the reported scores. However, the reported scores are mainly related to the difficulty of the solution word. To address this, we discuss the attributes related to the difficulty of the solution word and establish index variables. Based on the index variables we propose, we establish a linear regression model to predict the reported scores.