August 2025: Dogs Playing Poker

I wrote in June that “there’s something that’s so satisfying about extracting information from an initially indecipherable collection of objects and shapes”, and it’s for this reason that this month’s puzzle was incredibly fun and satisfying. However, extracting information from this puzzle felt considerably harder than the June puzzle, and there was an abundance of red herrings which made it easy to get caught up trying to process irrelevant information. This is reflected not only in my difficulty rating for this month’s puzzle, but also in my enjoyability rating: as satisfying as it was, this one brought a certain amount of frustration, costing it the 10/10 rating.

I once again collaborated on this puzzle with Fred Vu, since it’s the exact type of puzzle which is fun to solve in a team, and we usually feel we can place highly on these kinds of puzzles with a bit of combined creativity.

Stats:

Difficulty: 8/10

Enjoyability: 8/10

Leaderboard Placing: 1/245

The approach:

One immediately notices the dogs’ distinctive facial expressions, and the importance of this detail is confirmed by the puzzle’s wording — specifically the line about the dogs not being able to hide their emotions. A natural progression is to match each dog with an emoji, and the dog wearing the cowboy hat, for example, is a dead giveaway that this is indeed the correct approach. This led to our first hypothesis: the cards and chips in front of each dog somehow encode the corresponding emoji, and by using the eight given examples, we can deduce this encoding and therefore apply its inverse to the remaining dog to decode its facial expression into a unique pair of cards. We spent a considerable amount of time trying to make this work, with the unicode code points of the given emojis being a focal point of our investigation. This turned out to be one of a few wrong paths we went down — the others including attempting to assign meanings to the card suits and attempting to extract information from the dog breeds and paw placements, all of which were ultimately insignificant — but by dealing so closely with the emojis, we became familiar not only with their code points but also with their CLDR short names, which turns out to be what’s ultimately important. Reading clockwise around the table, the names of the emojis are as follows:

  • Flushed face
  • Drooling face
  • Cowboy hat face
  • Woozy face
  • Anguished face
  • Pouting cat face
  • Confounded face
  • Squinting face with tongue

Writing these names out in full turns out to be key, as is hinted by the sentence “what they’re feeling is practically spelled out for everyone to see”. From here, one must index into each emoji name by the values of the cards in front of the corresponding dog, read clockwise around the table (with ace low); crucially, the indexing starts at 1 and spaces are ignored. Doing so reveals the sentence “shift by chip count”, which was our first big breakthrough. This was quite satisfying, as it was me who had the idea of using the cards to index, and it was Fred who worked out how we should apply the indexing. From here, we knew a solution wouldn’t be too far away.

The next step is to do exactly what the previous clue instructs: each character in the phrase “shift by chip count” was contributed by a certain card which had a certain number of chips sitting by it; translating each character forward through the alphabet by the associated chip count gives the phrase “the canine of clubs”. It was then that Fred immediately deduced the answer: phonetically, “the canine of clubs” sounds like “the K,9 of clubs” — the dog at the end of the table was holding the K and 9 of clubs!

When the solution’s written out like this, it makes this puzzle feel quite easy, but there’s so much information available and infinitely many ways in which it could be used, so reaching the solution felt much harder than it may seem like it should have. The unicode rabbit hole in particular was quite hard to escape from, and it seems likely that other teams would get caught up on similar diversions. It’s for these reasons that I feel this puzzle is worthy of an 8/10 difficulty rating.