Predicting the Guess Distributions and Number of Reported Plays for Wordle
Faculty Sponsor(s)
Kate Lorenzen
Subject Area
Mathematics
Description
We present our results from this year's Mathematical Contest in Modelling, relating to the popular word-guessing game Wordle. Given historical data for the number of reported plays for Wordle taken over the course of a year, we use a random forest regression model to accurately predict future results. We also predict the guess distribution for a given target word using past data using regression, and make a classification scheme to categorize target words by difficulty using a k-means model.
Recommended Citation
Perez, Brendan and Liu, Boyu, "Predicting the Guess Distributions and Number of Reported Plays for Wordle" (2023). Linfield University Student Symposium: A Celebration of Scholarship and Creative Achievement. Event. Submission 46.
https://digitalcommons.linfield.edu/symposium/2023/all/46
Predicting the Guess Distributions and Number of Reported Plays for Wordle
We present our results from this year's Mathematical Contest in Modelling, relating to the popular word-guessing game Wordle. Given historical data for the number of reported plays for Wordle taken over the course of a year, we use a random forest regression model to accurately predict future results. We also predict the guess distribution for a given target word using past data using regression, and make a classification scheme to categorize target words by difficulty using a k-means model.
