Weekly Tea

Hi all!

We will have a special tea this Thursday to determine the winner of the 2023 Excel Prediction Challenge.

What is this challenge? It's a simple Kaggle-style prediction competition with a $50 prize for the first-place entrant. Tomorrow at 9:30am I will send out a Google Sheets link with two tabs: one with Xs and Ys ("training data") and one with Xs only ("test data"). To enter the contest, you should copy/paste the data into your own Google Sheets page, do some analysis, and send me back your work with a new tab containing your predictions on the test data.

The catch is: you may only use Google Sheets built-in functions to analyze your data and make predictions. You may not use R/Python or any special Sheets plug-ins, and all your work must be documented in the Sheet itself. The evaluation criterion is mean-squared prediction error.

Submissions will be due at 3:30pm, halfway through tea time. Hope to see you then!

Kevin and Samir

“We should not forget Excel. It may come as a surprise to us that a large proportion of the people conducing statistical analysis use Excel for this purpose. Economists, managers, biologists, chemists and others, start looking for a solution by a search for an Excel solution to their problem.” –Yoav Benjamini