A lightweight simulation to optimize the famous mathematical problem:
Based on the Numberphile video about toilets
And the article on Wikipedia
When the aim is to achieve the highest possible score on average, it's visible that relatively low amount of pre-checks is required. For 100 total data points, it's enough to check ~5% to achieve the highest score possible, on average.
When the aim is to achieve the absolute best score, it's visible that it is required to sample between 35%-50% of the data points before making a decision. This is consistent with theoretical findings and analysis.