Show HN: Wordle Charts – insights about Wordle you don't need

24 pointsposted 5 months ago
by crtez
(wordlecharts.crtez.dev)

9 Comments

danpalmer

5 months ago

This is really neat!

I tried looking up the word that I always use to start and it wasn't in the data, which is odd because I know I had a 1-guess win, and I'm fairly sure it would have been with the word I use every day. That said I checked with another list of Wordle words and can't see it there. Maybe it was before the switch to the NYT.

CamperBob2

5 months ago

I looked up my least-favorite Wordle of all time, HUNKY, and it's not there either. It's not a complete list.

crtez

5 months ago

Yep, BESET (3/39/2023) is the first data point that’s available, as far as I can tell. Any reason you didn’t like HUNKY?

danpalmer

5 months ago

There are a lot of words in Wordle that are a bit US focused. I've never heard someone outside US media use the word "Hunky", so wouldn't expect it to feature in the ~2500(?) common 5 letter words that Wordles are/were chosen from.

Strands and Connections are way worse for this. I'm British, my partner is Australian, about half the time the games are noticeably US focused, and maybe 20% of the time we just can't get them because it's about sports we don't have or slang we don't have or some cultural thing that just doesn't translate.

I wish The Guardian had a better selection of these games because we find their crosswords much better for example.

CamperBob2

5 months ago

It's BS. Considered pejorative to Hungarians, so it never occurred to me to guess it. :-|

crtez

5 months ago

Thanks! And yeah, the data only goes back to the first (as far as I can tell) average that’s available—BESET.

There’s a couple of more quirks here and there that I’ve found-the NYT also tracks (for users that are logged in) the Unix time of completion for your Wordle each day.

I was originally planning on making a heatmap to show people’s completion time patterns, but since I was tracking my own times, I found that the NYT’s data on my completion times was erratic and occasionally incorrect (within the same day but off by hours) up until like 3 months ago? So it’s kind of funny to think that we can track bug fixes through simple data comparisons.

kylecazar

5 months ago

Great to see that others struggled with CORER in October as well.

Word still upsets me.

crtez

5 months ago

Funny that you mention that--because CORER may have just survived an attempt at the throne. ROWER (01/19/25) was within inches, according to WordleBot. On the chart, it doesn't look that close because, for some reason, the summary.json that contains the NYT average didn't match up with the average shown on WordleBot that day. Regardless, it seems like CORER seems to take the edge.

Reference image: https://github.com/crtez/is_wordle_harder_or_am_i_just_stupi...

kylecazar

5 months ago

Interesting! No surprise, I hated ROWER too.

I don't know why, but words with an agentic suffix just seem.... Cheaper to me. It's like I want Wordle to be better than that so I don't guess them. Maybe it's the same for others.