Research on Predicting Wordle Results

Abstract

Worldle is a very popular puzzle game, but the number of players is decreasing. In order to reduce this phenomenon, this paper intends to solve it through relevant methods,such as ARIMA model, Credibility, Multicollinearity model; Ridge regression algorithm; Fuzzy synthetic evaluation method. First, the data in the game is cleaned up, mainly including the modification of the number of letters, the number and percentage of special letters in the word competition result report, etc. Secondly, the use of the establishment of the relevant model, and to be analyzed related factors and circumstances, to obtain clear results. It is worth noting that the model itself has certain advantages and disadvantages. Finally, based on the above steps, put forward four feasible suggestions, hope that the game can be adopted, the game complement.