This was my Wordle result for today. For 3,00,000 (at least) people like me and me, playing Wordle and guessing the word of the day has become a daily morning ritual. Developed by Josh Wardle, a software engineer, who previously worked with Reddit, Wordle has caught the internet’s fancy. In the latest development, the game has been acquired by New York Times for a ‘seven-figure sum’.
Here we try to decode the charm of the ‘slow game’ in the age of internet algorithms.
Story behind Wordle
Wardle developed this game for his partner Palak Shah, a crossword enthusiast. In an interview with New York Times, Wardle said that he and Shah got hooked to Spelling Bee and daily crossword puzzles. Wardle had created a similar prototype in 2013, but it was scrapped for being ‘unimpressive’. Called the ‘slow game’, Wordle players get just one word a day. As per Wardle, the ‘enforced sense of scarcity’ that was partially inspired by Spelling Bee leaves people wanting more.
Wordle’s popularity grew mainly because of Twitter. As per Siobhan Murphy, Twitter’s communication lead, between November 1 and January 13, 1.3 million tweets mentioned Wordle. The conversation around the puzzle has experienced a daily average growth of 26 per cent.
Data analytics in Wordle
Players have to guess the five-letter word of the day, which is the same for everyone, in six attempts. The feedback is usually given in the form of coloured tiles. Grey tile refers to the incorrect letter; yellow tiled are those letters which appear in the given word but have been placed at the wrong position by the player; a green tile means that the player has guessed the correct letter at the right position.
Many people have come up with data analytics solutions to solve Wordle puzzles. For example, Andrew Fox of SAP has used data visualisation techniques. Fox wrote that there are 2,315 possible solutions that are publically available. Data visualisation techniques can be applied, along with data augmentation and filtering to increase the green tiles (list of correctly guessed letters).
Ravi Gupta, a data scientist with Disney, gave another solution where data analytics can be used to solve Wordle. He used the lexicon of all English language words (about 3,70,000) and sorted five-letter words out, leaving 16,000 words. He then calculated how frequently each letter is used in these 16,000 words (A is the most common– 10.5% of all letters; letter E is a close second at 9.8%, followed by S at 8.2%). This cumulative distribution showed that seven letters – A, E, S, O, R, I, L) accounted for 53 per cent of all the 16,000 words. This exercise helped him list the 21 best words to use as Wordle guesses, like the aisle, arise, laser, etc.
Wordle bought by New York Times
Wordle buyout has been met with mixed reactions. While a section of people are positive about Wardle making money out of the widely loved game, others are sceptical of the corporatisation of the otherwise simple and easy to use game.
Wordle’s charm lies in its simplicity. Its ‘anti-algorithmic’ approach, meaning no advertisements or notifications or nudge to suggest it to other friends, make the whole experience very unadulterated. Ironically, many people and publications initially appreciated Wardle’s decision to reject monetisation attempts. The Guardian even wrote, “…the game’s rejection of the capitalistic systems that define so many video games today has a refreshing, innocent quality. Long may Wordle remain so pure.”
