Nice implementation. I've seen this puzzle before, does anyone know if it actually has an official name? Kind of reminds me of a variation on the tower of Hanoi.
Thanks! I’m not sure if there’s an official name for it — I originally saw the game a decade ago at my university and just recently decided to build my own take on it. It definitely has some Tower of Hanoi vibes. Let me know if you find an official name for it!
Not being able to repeat a successfully solved level sucks.
It would benefit from that model that is popular nowadays whereby higher levels are unlocked by solving lower levels, and you can choose to play any unlocked level whatsoever.
Glad you liked it! That last level twist was a fun addition — I wanted to keep things interesting. Your Python solver looks awesome, I’ll definitely check it out. Always cool to see different approaches to the same puzzle!
In some UK schools, this was the basis for a mathematics investigation for 11 year olds. Do the task for various numbers of frogs, records the results, and try to find a rule - then make a prediction for a certain number, and test it!
A great introduction to the Scientific Method, and a fun activity too.
As someone who is colorblind (Deuteranopia), the first level asks to swap the green and brown frog. With the hues used, I can’t tell the difference between the two frogs.
Good call! The facing direction was meant to help, but I can see how the colors could still be tricky. I’ll look into making it clearer — appreciate the feedback!
Hey everyone, I’m the author—really appreciate all the feedback here. If you want to follow along with what I’m working on, I’m over on Twitter: https://x.com/rv_labs. Thanks again!
I'm curious about the difficulty curve. As long as the puzzles take the form "group of left-facing frogs - empty pad - group of right-facing frogs", they'll all have exactly the same solution. If you can do level 2, you can do level 2222 without needing to come up with any new ideas.
This kind of blocks the notion of a "difficulty curve" - it's just a flat line.
Level 4 is different, but I'm guessing the reason people say it's easier is that the frogs on the left never need to interact in any way with the frogs on the right. You can just march your left-frogs onto the transformer, see them turn around, and march them back to their ultimate destination, then repeat the process on the right.
I don't think there's any solution that would allow a frog to jump over the transformer, so you're essentially required to do this - all frogs must make their way to the transformer, transform, and go back home - although you can make it look more or less complex. This essentially gives you two copies of the "level 3 and below" puzzle, one on the left of the transformer and the other on the right.
How so? This is an old puzzle with lots of versions online, I'm surprised it'd be struggling with it (except for the twist that I'll leave out for one of the levels).
It depends on how it's explained. The puzzle, depending on phrasing, is not too dissimilar from other movement puzzles like river crossing puzzles. What I'm curious about is how it was explained to the LLM and how an answer was expected in response. You can probably even find text descriptions of solutions online so if you use the right keywords it would likely give a correct solution even if your puzzle description was wonky.
I think it lasting 2 minutes is actually an asset. Even though a 2d version etc would be quite interesting, it is great to try such little nice things and then move on to one's day, not everything in the internet has to take that long. And unless one wants to show ads or sth, there is no incentive to actually make it longer anyway.
No? Sokoban involves a single agent pushing blocks around a grid. No block can ever move without being pushed, and nothing can ever cross anything else.
In this puzzle, the only way for anything to move is independently, and everything is free to cross anything else.
https://en.wikipedia.org/wiki/Tower_of_Hanoi
reply