99 comments

[ 6.5 ms ] story [ 154 ms ] thread
(comment deleted)
Would have been nice if they'd tested it on Safari as well. It's playable, but frustrating enough that you'd switch to Chrome.
The real puzzle is to figure out how to play it in Safari.
On puzzle 5, why is the obvious short solution wrong? The marble reaches the finish line, and then the page says "Something's not right, keep trying!"
One unstated rule is that you have to use all pieces. Number three can already be solved with two pieces and that also doesn't count. I strongly dislike this kind of puzzle.
(comment deleted)
Thanks for sharing. Where does it even say you need to use every piece?!
It doesn't say it. It probably should.
An unstated rule: not said/stated.
Hmm, I got the solution it wanted after a lot of trial and error, but I still don't understand why that one was correct
That also happens on an earlier level. The answer appears to be that there is an undocumented requirement to use all of the tiles. I'm not sure why Google appears to believe that documentation is weakness.
There's even documentation. Clicking on the (i) icon says "Place tiles on the grid to create a path for your marble". It would be so easy to just change it to "Place all tiles..."
Because at Google, you're expected to spend your entire budget
There's a "help" popup indicating the existence of keyboard controls. Then, in level two, you have to rotate tiles, for which no control is documented. What gives?
Oh man I was so confused when I couldn't solve level 2. I even double checked the instructions that there was no rotate key. I'm on a laptop with a nub for a mouse and I'm spamming various keys to try to figure out how to get a piece to rotate and I haven't been successful.

I can also select a piece and hit space and it grows in size and goes outside of its box for some reason? Is that relevant to solving puzzles? Or a bug? I guess I won't find out since I can't rotate pieces to progress further ....

The presentation is nice but the content of the puzzle put me off. I think if you suggest a physical puzzle by presenting it as a rolling ball then you should honor correct physical intuition such that a ball isn't going to turn left by itself as it does in puzzle three. I'm interested in testing my wit, and I am fine with losing, but I am not interested in just finding the correct way to clap like a seal for the puzzle designer.
Also the ball starts on a flat surface and then goes uphill(??) in the finish tile.
That confused me too. Isn't that literally the wrong way around to how it should be?!
I guess we were expected to be astonished that a simple 2D game about connecting lines became 3D in a browser or something.
+1 for 3rd level.. the simple solution (as per me) using two tiles does not work, even though the ball reaches the finish podium.
Seems like you want your wit to be rewarded more than tested?
No, I don't think it wouldn't seem like that to any reasonable person. What makes you say otherwise?
There’s nothing wrong with it. That’s part of the fun of games – feeling smart and accomplished.

But it’s pretty clear and obvious (to a reasonable person) how the game works and you seem to want to just brag about how you’re clearly a better thinker and smarter than the game designer.

It's not clear at all and you can read comments here to find that out.

I don't know the game designer especially but what I actually think is that the system (ie org+people) that produced this puzzle is much smarter than I am. I just don't think the puzzle design is appealing and I gave an impersonal argument to that effect. You have a duty on this side to take comments in good faith. If I give you a factual argument about why I dislike a puzzle you don't just get to accuse me that it's really about intellectual girth.

I think most people who commented here did so to dunk on the puzzle maker because they felt smarter than them and wanted to show it.
What do they (we?) have to say that you'll believe that we just don't like the puzzle for the non-intelligence related reason we stated?
I thought the people who didn’t complain (or who didn’t comment at all) were much smarter.
> such that a ball isn't going to turn left by itself as it does in puzzle three.

I'm willing to deal with some trial and error with stuff like that, as long as it's predictable and cause-and-effect is consistent. What I'm not willing to deal with is what I'm seeing in puzzle seven. I have a crossroads with a U-turn to the north and a curve to the south. The ball enters the crossroads from the east and goes south. If I remove the northern U-turn (which the ball hadn't visited and was therefore useless), the ball now goes north (and into the void) instead of south to safety.

Then you have the obnoxiously loud music that can't be turned off separately from the sound effects (so you either have annoying music or no audio feedback at all), a condescending "keep trying" popup that treats every test as a failed attempt, slow animations that make for an annoyingly slow feedback loop on trying new things...

It's a really cool concept, fun presentation, but execution is all sorts of terrible. You could easily run a game design masterclass centered on fixing this thing.

The secret rules appear to be:

- You have to use all the tiles.

- The ball must visit every part of every tile.

- The finished layout must look elegant.

Just assume that there is no goal other than closing all the paths. A tile that connects to empty space means you're wrong. A tile that connects to another tile means you're right. Any and all other "rules" will be changed as necessary to ensure that, if a solution is pretty, it's also correct.

That's not a puzzle, though. It's just testing whether we agree on what "elegant" means.
Counterexample (working solution): https://imgur.com/a/VxDJc6M, level 9

* all the tiles are used: yes

* every part of every tile visited: no. bottom right tile is not visited at all; the dangling path of the three-way tile is not visited

* elegant: no. Dangling paths.

(comment deleted)
Can someone just provide me the solution to #3? I literally cannot figure it out.
Start, curve west to south, "Y" piece, curve north to east, curve west to north, reverse, back through the last two curves and the "Y" piece (unintuitively the ball will take the left to the finish).
> unintuitively the ball will take the left to the finish

really that makes no sense, physically or in the universe of the game :)

I scratched my head for a while trying to find a configuration where the ball traversed that Y piece intuitively (entered the curve from the west, was reversed, and then exited the straight to the north) before giving up and trying the same "impossible" solution.

Maybe it's a tie-in to their presentation - perhaps they have a new model like Sora which has a poor understanding of physics. (This would explain why the ball starts flat and ends going uphill as well.)

For a quick, breezy puzzle that you should be able to do within the space of a coffee break, this feels like it moves way too slowly. Takes a bit for the ball to start rolling and the fancy zoom-in animation gets tedious fast.
(comment deleted)
> the fancy zoom-in animation gets tedious fast

Broadly, this criticism applies to nearly all UX.

Impressed that it actually got my M1 Macbook Pro's fans going, lol (Firefox)
(comment deleted)
(comment deleted)
[flagged]
(comment deleted)
What a frustrating, poorly designed puzzle.
(comment deleted)
[flagged]
Chances are Google didn’t make it. They hire a ton of design studios and vendors for stuff like this.
You are indeed correct, although a few Googlers typically will help as well with the launch. If I remember correctly the front end JS devs handle most of the site and the puzzle itself is done separate. However, perhaps they changed it up this year.

Still a waste of time and money, though.

There is a famous Marissa Meyers party at Yahoo! that this reminds me of. I thought with all the layoffs, Google was getting disciplined and everyone in line. Maybe that is not the case. When Microsoft brought in Nadella, they did some belt tightening as I recall. My understanding is that Meta is using the whip to get everyone focused. So maybe they have a better future? Or maybe G doesn't think they have an existential threat.
Seems like this crowd is upset over a children's puzzle game. Thought it was alright, has decent graphics that shows what can be done procedurally without the need of carefully hand crafted assets.
See if being a children's game has any impact on your feelings about Winnie the Pooh's Home Run Derby then :p.

Small Google games are often actually really decent, especially in comparison to this, so I think that's where some of the disappointment comes from. The Doodle Champion Island one was my favorite.

There are frustrating implimentation details like puzzles not being solved if you do something that looks obviously right but some subtle unwritten rule means it's wrong. In game design (and all UX really) when something looks right then it should be right. The user shouldn't ever have to think very hard to understand why something is wrong. This game fails in that regard.
I don't think it was intended to be a children's game; they're not the target market for Google I/O.

But also, in this be situation there isn't much positive to be said about the game. Everything about it is badly done. I'm fairly sure the criticisms (vague and badly communicated rules, clunky UI, graphics style that doesn't fit the puzzle, announcing music without a separate music control, way too many slow animations in the wrong places, browser compatibility issues) would be the same if this was a Show HN.

It's just that none of it is being given as constructive feedback. But even that makes sense: unlike for a Show HN nobody will be applying the feedback to fix the game.

(What I don't get is how this is so high on the frontpage, unless it's just for the schadenfreude.)

> But also, in this be situation there isn't much positive to be said about the game. Everything about it is badly done.

Interestingly, this appears to be completely intentional; the 2023 puzzles have the same problems.

(comment deleted)
Cool, solving the puzzles gives you pleasing designs.
That lack of antialiasing is horrible. I'm pretty sure it's not my browser settings because the three.js examples look ok.
(comment deleted)
Is one of the rules that you have to use all the puzzle pieces? Because puzzle #3 is easily solved with only two pieces yet that is not the "correct" solution.
Unfortunately yes, you have to use every single piece.
And that's not explicitly stated.

Had the same "problem" too. I would expect better from an adtech/software firm.

Kind of a good analogy for Google in the 2020s, isn't it?
Getting WASD and Space instructions on mobile too...
maybe a gamer recognizes game logic but I can't understand even level 1. I guess I should start playing some games.
Slightly unrelated but when opening this in private browsing mode, the page displays a warning “you have opened this page in incognito mode …”

This is the first time I’ve ever seen a page fingerprint me using private mode and tell me about it

Just a small notice that Big Brother is disappointed in you :)

I wonder how do they detect it, maybe there is no LocalStorage on incognito mode, but TBH it should not be visible to a webpage

"Chrome's FileSystem API is disabled in Incognito Mode to avoid leaving traces of activity on someone's device. Sites can check for the availability of the FileSystem API and, if they receive an error message, determine that a private session is occurring and give the user a different experience."
That’s one way, but they would have had to have used multiple solutions depending on the browser.
the interface on mobile leaves a lot to be desired
Nice little game.

Unfortunate my company already declined my request for Google I/O :(

Things started getting tricky at level 7. I can see that level 8 is solvable, but I'm not going to have time to solve it before leaving for work. Neat little puzzle.
Clicking start leads to "Application error: a client-side exception has occurred (see the browser console for more information)." apart from Safari, which serves me up a frozen game. Console errors complain about

"THREE.WebGLRenderer: A WebGL context could not be created. Reason: WebGL creation failed: * tryNativeGL (FEATURE_FAILURE_CGL_FBO) * Exhausted GL driver options. (FEATURE_FAILURE_WEBGL_EXHAUSTED_DRIVERS)"

Any tips? Are there magical flags that should be set somewhere?

I'm on Firefox 123.0.1 on Linux 6.6.19 using Nvidia driver version 550.54.14 and it's working without issues for me.

The later levels are really tough!

are they training Gemini with this puzzle ?
Level 12 seems to explode with possibilities. Any hints to reduce the search space?
I finally solved it after many rearrangements. As a hint, try making a roundabout in the bottom left...
Focus on laying out all the pieces, such that there are no dead ends. There's very few ways to do that, afterwards you can quickly brute-force the rest.

The winning move might be to not play, tbh. It was by far my least favorite level.

Spoiler alert: in puzzle #3 what properties does the "track switch" piece have that makes the sphere take the turn instead of going straight on its way back from the bouncer? Doesn't look very physically correct.
the UI is really a mess, I didn't find there are 5 pieces in mobile...