This Bot Can Tell You How Good You Are at Wordle
I like to think I’m good with Wordle (Worrying about Wordle is one of the biggest failings of my personality and I’ve come to terms with it, thank you very much) so I was thrilled to see the New York Times have a new tool called WordleBot , which can evaluate your performance on any previous Wordle.
If you’ve played Wordle recently, you can visit the tool in the same browser and it will know how you made your guesses, but you can also upload a screenshot of Wordle you’ve played in the past. I’m going to introduce you to the bot with some old screenshots, but then we’ll dig into the strategy for today’s puzzle number 293, so either stop reading or go solve it now if you don’t want to see spoilers.
How to use the Wordbot
Solve today’s puzzle first and then follow this link . He will guide you through the whole process. If you need an analysis of an old puzzle or someone else’s puzzle, take a screenshot. (After you solve the puzzle, the Wordle stats screen will appear. Press “x” and then take a screenshot of the grid. You can also return to this page anytime after solving it, as long as the next puzzle hasn’t come out yet, and it will remember your guesses.)
Then scroll through the result screens. The first gives you a total score for both skill and luck. The rest of the steps are through your guesses and the bot compares your choice to what it would have done.
How good is WordleBot in Wordle?
By the way, the bot is cheating. It counts how well you narrow things down, but it knows what 2309 words are on the list of official solutions . In fact, there are over 13,000 five-letter words in the English language; many were simply omitted to make the game more interesting, such as four-letter plurals of words that would not normally be there. Some obscure words have also been omitted, as well as a bunch of words that I don’t consider very obscure: FUTON will never be a solution, the bot told me, and PORKY won’t, although both are valid guesses.
On the other hand, at least the bot doesn’t know which puzzles have already been posted. When I’m playing, if one of the choices is SNOUT but I remember that SNOUT was the decision last week, I can mentally cross it out. The bot was not given this instruction.
How to learn from WordleBot (spoilers ahead!)
The most interesting thing about WordleBot is not that it will tell you how skilled you are, but that it can give you tips on how to become a better person.
Today’s puzzle was easier (my two colleagues and I solved it in three attempts). The bot, the smug bastard, says he could solve it in two. (Begins with KRAN.)
I got an overall skill rating of 93. I liked my opener (STONE) and my second choice CHAMP reduced the number of options to two. My third guess was successful.
I then ran staff writer Sarah Shawfety’s solution through a bot. This gave her an overall skill score of only 85, mainly because her second SNARE guess reduced the solution pool to only three options instead of two. Associate editor Joel Cunningham scored an 84 for skills, in which case he appears to have been banned for being too lucky. His opener left only 9 possible solutions, and after CLAMP there was only one. (We all beat the typical Wordle player, says the bot: average skill score is 79.)
Out of curiosity, I tried today’s puzzle again in a different browser and struggled to play poorly. (As we learned from my Antiwordle post earlier this week, I may not actually be very good at Wordle, but I’m definitely bad at being bad at it.) I liked my starter, BOGUS, and my subsequent, PARSE. But after that, the little bot just shook its head at my wrong guesses. At that point, I narrowed down my choices to just four options. My next two, RAISE and SPARE, were “blank” guesses that didn’t add to what we already know. But that’s not really the case: both gave clues about the position of letters that were previously yellow, and both included gray letters that I could rule out as possible. However, I ended up with only 22 skill points, solving the puzzle in five guesses, which is fine.
So what have we learned? Among other things, we received evaluations from our starters. CRANE – bot selection; he liked Joel’s STERN and Sarah’s ROUTE. All this on average narrows the pool to 90-100 options. (However, this number can be more or less, depending on the actual decision of the day. Sometimes you just get unlucky.)
So go ahead and ask the bot what he thinks you did. You may be furious at the answers, but you will probably learn something.