Serious Wordle Players Get a New WorldeBot to Analyze Their Performance
The New York Times just released an updated version of its WordleBot, with a range of new features to help analyze your Wordle-solving performance as well as new opening word recommendations.
WordleBot was introduced back in April for NYT Games subscribers(Opens in a new window) as a tool for analyzing how they went about solving the daily Wordle puzzle, offering feedback in the form of scores for luck and skill on a scale of 0-99. It also offers tips on what you could have done differently to solve the puzzle faster.
WordleBot 2.0(Opens in a new window) promises to improve on that feedback. The number of common English words it knows has been increased from 2,300 to around 4,500, and the solving method used now offers better estimates. The bot ranks how likely a word is to appear in relation to how common it is based on appearances in The New York Times since 2010. One side effect of that is plurals and past-tense verbs are classed as less likely to appear.
You may be surprised to hear it also takes into account how often a word has been used as a baby name in the US to determine likelihood of appearing. This is because the more often a word is used as a name, the less likely it is to appear as a Wordle solution.
Regular WorldleBot users may notice their skill score changing compared to the original bot because the expanded word list means invalid guesses are judged more fairly, but guesses that "reveal the most information about which words might remain" are now rewarded.
Other tweaks include a fourth ratings column called "Info Gained," a comparison of how your choice compared to other player guesses in the same situation, and a step-by-step animation showing how Wordle players around the world solved the puzzle each
Read more on pcmag.com
