Wordle and a World of Unrealistic Expectations

Feb 06, 2022 21:28



Like many other people, I've become a big fan of the new web-based puzzle game, Wordle.

There's a lot of discussion about Wordle going on in social media --- not just people sharing their results, but a lot of meta-discussion about the game itself and why it's so popular.  Some of the observations that others have made:

  • Less is more.  You're only allowed to play once per day, which builds anticipation from day to day, and also keeps people from getting overwhelmed from binging on it.
  • Simplicity wins.   The game itself feels like a throwback to the early days of the web, with its primitive graphics, and simple game play.  The game is simple enough for novice programmers to write versions of their own (and many are doing so, as evidenced by all the variations available).
  • Everyone loves a pure love story.  If you haven't read the story ... the author created the game as a gift to his romantic partner.  It was only much later that he decided to make the game available beyond that ... first to his family, and then to the public.   (Of course, everyone fears that purity will wane, as the author sold the game to the New York Times.  Time will tell.)

But one of the things that I've found interesting in the discussions about Wordle is all the discussion about everyone's daily performance.

If you haven't played the game --- the goal is to guess that day's five-letter word in six guesses or less.   Every guess gives you clues about what letters are and aren't in the word.   At the end of the game, win or lose, you're given a little graphical representation of your game that you can post (spoiler-free) to your favorite social media outlet.   And that prompts discussion about how "easy" or "hard" today's word was for them.

That discussion, of course, centers on the "score": namely, how many guesses it took to get the final word.  Implicitly, there's a common assumption: a high number of guesses indicates a hard word, or a difficult problem-solving process, or the inadequacy of the solver.   Conversely, a small number of guesses indicates an easy word, or an easy problem-solving process, or a brilliant solver.

(Unless, of course, you get the word on the first guess, in which case you're cheating.   And, yes, it's easy to cheat at Wordle.)

But it occurred to me ...

... why do we ascribe value to how quickly we solve the Wordle?

First of all, Wordle is a lousy measure of intelligence.  Being good at Wordle requires a very particular set of skills, which don't necessarily transfer to a lot of other life tasks.   So taking a lot of guesses (or losing entirely) doesn't mean you're less intelligent that someone else who solves it more quickly.

But more importantly ... why do we rate ourselves on the number of guesses?   The goal of the game is to find the mystery word within six guesses.   If you get the win, why does it matter whether you got it in six guesses or in four?

I think about this a lot when it comes to teaching --- and, in particular, teaching CS.

  • One of the biggest struggles faced by those of us who teach the first algorithms course (colloquially called "CS1") is the diversity of backgrounds of our students.   Some students in CS1 have been programming for years --- either recreationally or professionally --- and simply haven't gotten the academic "credential" that says they know the stuff.   Other students have zero background, which is completely fine for a first course.

    So you go into a classroom, and you assign a little in-class exercise that you expect will take a novice student 30 minutes to complete.   The all-star programming veteran finishes the assignment in 10 minutes and leaves.   The novice student finishes in 30 minutes, with equally high-quality work --- and leaves the class feeling dejected that it took them three times as long to do the work.   But that's not fair; the 30-minute student met my expectations, and shouldn't feel bad that they haven't the preparatory privilege that others have.
  • Yet again this week, I had a meeting with a student during office hours that reminded me of a rant of mine from earlier this year.   The student is struggling, but didn't want to come talk to me about it because they didn't feel like their questions were "worthy" of my time.   Why should it matter if a student gets help (from authorized sources) to complete their work?  The goal is to learn, and to complete the work as a means to learn.  If the goal is achieved, why should it matter how we got there?

I think we spend a lot of time setting up unrealistic expectations for ourselves --- or accepting unrealistic expectations from others --- and then feeling bad when we don't meet those unrealistic expectations.

Maybe we need to re-think how we define success.

In a second-season episode of Star Trek: The Next Generation ("Peak Performance"), the B-plot of the episode involves Data, the android second officer, going through a crisis of confidence when he loses a game called Stratagema against a grand master.   The episode ends with Data resolving his crisis of confidence by ... well, I won't spoil it for you:

image Click to view



So, if you're playing Wordle ... maybe it's time to re-think your goal.  If you get the word within six guesses, celebrate!  Maybe the revised goal will give you different solving techniques that will help you "win" if you aren't trying to get there in the shortest path possible.   (I have a couple of those techniques.)

And as for the rest ... just because someone else gives you a way of rating your performance in life doesn't mean that you have to take it.

Previous post Next post
Up