Three Emojis is a daily word game designed for language learners. You get seven letters and a list of blanked-out words to find. When you discover shorter words, they automatically fill into longer ones—like a crossword—which turns out to be really useful for languages like German.
Each word also gets three emojis assigned to it as a clue, created by GPT-5 to try and capture the word’s meaning (this works surprisingly well, most of the time). If you get stuck, you can get text/audio hints as well.
It supports German and English, with new puzzles every day. You can flag missing words or suggest additions directly in the game. The word lists include slang, abbreviations, and chat-speak—because those are, in my opinion, a big part of real language learning too (just nothing vulgar, too obscure or obsolete).
Every word you find comes with its definition and pronunciation audio. If you want infinite hints or (coming soon) archive access, you can upgrade to Pro.
Feedback is very welcome, it's my first game and I'm certainly not a frontend guy. Happy spelling!
zaene•3h ago
Some feedback on the UI - at least on desktop Chrome, the title part of your scoreboard area is being cut off and the "How to Play" section has the text totally flush with the top, so you might want to look into your margins or padding there to give everything some breathing room. There's also a bug in your shuffle algorithm -- the letter in the top right cell never changes when you shuffle the letters. The cell buttons are also a little unresponsive. It seems like there are some dead zones where the hover animation doesn't get triggered and clicking the cell doesn't actually input the letter.
I like how you've added some features to make it easier for language learners to find words. Are you yourself studying German? And if so have you found it fun/useful?
knuckleheads•2h ago
I am still nailing down how many words to put in and therefore how long the game is. Right now somewhere between 30 and 75 feels good, I had it at 100 and friends complained that it was too long. To get a sense of your data, make a histogram of total number of words per letter set. This is different per language and does give you sense of what's up, as well as how good your word universe is. Conjugations/inflections help pad this out a lot as well.
Get the word frequencies for each word, wordfreq is helpful. Then, do a greedy algorithm, start accumulating the list of letter sets, one for each day. For first day, take the letter set that maximizes the sum of the squares of the zipf from wordfreq and has the number of words that you want for that day. For the next day, remove all the previous words from the words for your possible letter sets and repeat the sum of squares of zipfs. Then just keep running that, and it will maximize the most common words that haven't been seen yet.
Additionally, I tried to filter out a lot of words out front before running the game covering algorithm. No vulgar, nothing obscure or too obsolete. It's relatively cheap to run a very large word list through GPT-5 and ask a couple questions about each word. Do that once and you have filtered out a fair portion of the list. Build in a system for ad hoc blocking and it gets you most of the way there.
That is very helpful about the buttons and formatting. Thank you so much! I have that fixed up soon. Big lol on the shuffle algorithm, I hardly ever use it so I hadn't noticed. Thanks! I am in the middle of studying German, yes. What I've found is that it is very helpful for introducing me to new words for sure, that I am not sure I could have seen otherwise. It also helps me see the patterns within how the words are formed, just trying to puzzle out things. I have to take a lot of guesses and I feel like I am getting better at guessing which letters will ended up going where, even if I have never heard the word before.
zaene•1h ago
knuckleheads•1h ago