0h n0 – A new game from the makers of 0h h1 | Hacker News Reader

0h n0 – A new game from the makers of 0h h1

0hn0.com

345 pointsm0rph3v591 comments

Staring at the circles had a very odd mental effect for me. After playing a few rounds other websites I look at seems very square, straight, and ridged.

This is the second time I've ever felt this effect. The first time was with the 2048 variant Hexagon 16384 [0]. I'm starting to suspect that it's something having to do with spending a long time staring at a set of regular circles.

I wish I knew a neuroscientist so I could ask them about this effect.

[0] http://rudradevbasak.github.io/16384_hex/ with HN discussion at https://news.ycombinator.com/item?id=7437009 and my comment at https://news.ycombinator.com/item?id=7443465

Wow, thanks for pointing this out to me. I thought my phone had updated or something because everything looked so blocky... I had my friend check out my phone and he didn't see anything different. I thought I was going insane :(

Echoing this sentiment. Closing out of the game and returning to my browser had me wondering if Chrome had updated to a new theme with harder edges

I immediately though "Man, has hacker news always been so angular? Wonder if styling changed... my task bar seems harder too... wat"

Really interesting effect

Not just chrome, my whole Yosemite environment seems so sharp.

I imagine it's along the same lines as when everything looks blue after you take off pink ski goggles. I read a study where people wearing glasses that flipped everything upside down would eventually start to see normally, and then after taking the goggles off had upside down vision for a while.

This seems to suggest that your brain has an intrinsic expectation of what the world should look like, either built in by evolution or from year of experiencing the world a certain way. I wonder what sorts of things are filtered from our perception because they don't correspond with the brain's expectation of reality.

The game trains you to see circles as squares. You're looking at circles but they really act like squares so when you go back to reality it's like you forgot squares existed because the whole concept of square is made obsolete in the game's world.

I'm really surprised this works so well, it really shows how plastic and manipulable our brains are. This could become a reference example on perception or an analogy for higher level perception-shaping.

I could be wrong, I can't remember fully. But if I recall correctly they don't start to see normally, i.e. the right rotation, they simply adjusted well to the upside down view.

Wow, that was exactly what happened to me... Everything is so sharp and square here.

Didn't even realize that this might be the issue. I thought something got updated.

I also had this effect, my macbook keyboard don't look like it used to..

Recommended if you like this: Light Up from Simon Tatham's Portable Puzzle Collection[1]. The UI is not as polished, but the collection includes a great number of other puzzles, and has iOS[2] and Android[3] ports.

[1] http://www.chiark.greenend.org.uk/~sgtatham/puzzles/js/light...

[2] http://hewgill.com/puzzles/

[3] https://chris.boyle.name/projects/android-puzzles/

This actually reminds me more of Simon Tatham's Range[1].

[1] http://www.chiark.greenend.org.uk/~sgtatham/puzzles/js/range...

Oh, I haven't tried that puzzle yet.

Oh my goodness. I didn't even think to search for an iOS version. "Net" has for some reason become one of my desert island games. (Thanks a bunch)

It would be useful to have some explanation such as a tooltip for the icons at the bottom of the board. Their functions are not necessarily clear.

I appreciate a game like this: simple, but challenging and entertaining.

The X in particular. It seemed like a bad idea to click, and it was.

A big improvement over 0h h1. I never felt like I had to think when playing 0h h1. Every puzzle was solvable by searching for a series of patterns that had an obvious solution.

0h n0 seems to be just complex enough that sometimes you actually need to think about the problem. At least that is the case so far.

Actually, not being so complex that I actually need to think is what I really like about 0hh1.com.

It requires a non-zero but pretty low level of "executive function", but significantly less than most logic puzzles.

In particular, a few times per game, I have to apply a little bit of "frontal-lobe" or higher-level skill to prevent myself from losing my patience for constantly applying the lower-level or "perceptual" skills required by 0hh1. In particular, I have to apply frontal-lobe skills to resist the temptation to engage in the "wishful thinking" of falsely believing I have perceived a correct next move to make.

(I engage in 0hh1.com mostly to relax or unwind, not to hone any higher-level skill.)

For me it was somewhat (but not exactly) the opposite: I encountered 0h n0 first, and was quickly able to finish each level without the computer hints, while I currently still need the hints sometimes for 0h h1.

It seems 0h h1 has a slightly harder learning curve before the pattern recognition kicks in, while 0h n0 doesn't rely as much on pattern recognition, but is easier to learn.

Is every level solvable without guessing? Or are there multiple solutions for some levels? I feel like I had to guess sometimes, but maybe I haven't thought about it enough.

I have yet to come across a level that requires guessing. They definitely exist in theory though(uncertain whether the level generation algorithm is smart enough to eliminate them).

There are no multiple solutions due to an unstated constraint that resolves ambiguities. I believe it is that a blue cannot be surrounded completely by reds.

Here's a solver for the puzzle (uses the clingo ASP solver): http://www.takingthefun.com/2015/02/0hn0com-solver.html

Using the solver, I calculated that an empty 2x2 puzzle can be solved in 10 different ways, an empty 3x3 puzzle can be solved in 250 different ways, and an empty 4x4 puzzle can be solved in in 22946 different ways.

I can't seem to figure out how to play. I'm trying to make each blue circle count the number of other blue circles it has vertical and horizontal line of sight to. However, when I do this, there's no indication that I won. And some levels are very clearly impossible to satisfy. The hints are rather opaque. ("Looking further in one direction would exceed this number." What?)

I was able to figure out the rules by installing the android version, which explains that you need to fill all the dots to win. There can be no empty white dots. They must all be red or blue, in case anyone else is wondering.

When playing this kind of game, you are not actually playing, but training your brain to be like a cpu which is dumber in logic but stronger in following a procedure.

And your fun is over when you figure out the procedure, coz all you have to left is to get the input from whatever on the screen, feed it to the procedure and output the result to the mouse until u win.

Yes. You don't seem to need any backtracking to play these games correctly. This is also why the computer seems to always be able to give clear hints. (Can it really always?)

One question is whether this holds for any possible board (starting position), or if they need to be constructed in a certain way for this to be true.

I think a game like 2048 will remain interesting longer, because you need to look further ahead and take the random factor into account.

The source is on github: https://github.com/Q42/0hn0

Very cool, just cost me a bunch of time I've really should have spent on more productive things ;) And will cost me even more in the future.

Any reason why the Android App needs access to local storage? (0h h1 doesn't)

Thanks a lot for this! I've been playing to Oh h1 like crazy (current score on my laptop is 23,376, and 10k on my phone).

I've never quite liked Threes and 2048, or other games like it, as they either feelt too high paced or too intellectual to me.

Your two games on the other hand hit right at home with me, and I actually feel like I'm getting smarted. I'm sure it's placebo, but hey it feels like it helps so it's helping.

Anyway, thank you so much for the games!

Ojoo!

25638 phone. I'm slightly saddened that 0hh1 only includes some of the possible deductions. I generally play 8x8 and I've not found a case where a RBR----R type of row requires a deduction about where the R can not be because it would match an existing row. These come up where you have three unknowns in a row, but I think never for four.

Enjoying messing around with this. I ended up getting to a score of 100,000 on 0h h1 and realized I should probably stop playing it :P

Feature request: user-defined lock of selected circles, similar to how the existing red circles are locked.

(This is like the Minesweeper flag option)

Another feature request... if a tile is already illegal (too many committed blue in line of sight, or no way to commit more blue to satisfy requirement), flag it already. I think the biggest pain is the counting. Otherwise, a nice game!

That's what the hint button is for.

If you start adding any automated processing like that, then you're only a couple more processes away from the game just solving itself without you.

The elegance of this implementation is that you are doing all the solving. Unless you specifically ask for help from the hint button.

It is a balance between playability and automation. Even minesweeper does that (auto deduce that a large region can be opened up). What I described is only the first-order "convenience" checking to avoid the pain of going way back because of a miscounting.

Regarding the hint button. I don't want to use it precisely because I don't want to "win" by repeatedly getting hints.

I really like the realtime feedback idea, similarly you could always show how many additional dots are needed (e.g. the starting counts are correct and count down as you add visible blue dots in it's line of sight), 0 is blank, negatives could be hilighted as issues. Then you don't have to count existing circles, and errors are clear in real time.

Is it fair to rely on the fact that there is only a single solution to make deductions?

Yes. Any deduction that you could reach from that fact could also be reached from other constraints on the board.

Here's an example illustration, from the top-left corner of a board:

  y B R
  x R _
  _ _ _

If x is red, then y is indeterminate, either red or blue is valid there, no other constraint can "see" y to determine it. But if we know the puzzle has a single solution, then we can correctly say x must be blue, so that some other constraint south of x can see through to y. This could be worked out later from that other constraint. But the rules yield this emergent property of a shortcut, and any such emergent property is fair game for logical deduction.

If there's anything unfair about this approach, it's that the existence of a single solution is merely implied by puzzle conventions and not explicitly stated as a rule for 0h n0 here.

There's actually a constraint that no blue circle can be alone, which you can find by reaching an end board with a solitary blue that the game will prompt you about. So X red and Y blue would be well-determined.

However there does appear to be another constraint that is not mentioned explicitly, which is that there are no blue circles that are not in line-of-sight with a numbered circle.

That last isn't defined explicitly. It's an emergent property from the constraint that the solution must be unique.

Any blue circle (or any linear group of them) that is not in line-of-sight with a numbered circle would have no constraint preventing it/them from being red instead. Therefore the solution wouldn't be unique, so such blue circles can't exist.

Some starting conditions have multiple solutions. A trivial example would be a 2 x 2 game with two "2" at opposite corners.

I believe that such instances are not generated by this game, however.

I had an example with two possible solutions (3 empty spaces were left, and either BRB or RBR would have satisfied the conditions). I meant to take a screenshot after finishing but the final position fades out (and adds numbers everywhere) so I should have done it before the end.

I would give an unqualified yes to this. All rules and constraints of a puzzle are fair game for making deductions.

I don't understand what one of the hints means: "One specific dot is included in all solutions imaginable."

Took me a bit to understand it too, it means: Of all the possible permutations that the surrounding dots (of that dot) could be, there is at least one dot that is the same (so that while you can't 'solve' that dot, you can still start to fill out the surrounding).

Example, in the top-left corner of the board:

  2 _ _ R
  _ _ _ _
  R _ _ _

The location to the right of the 2 must be blue. If it's red, then there's no way for the 2 to see two blues. That's what it means, that that dot is included in every possible solution for the 2.

I think it has something to do with two adjacent "known" dots (for example, I got that on both a 6 and an 8 when both were right next to each other).

Nice game, quite polished.

Is the solution unique to each puzzle?

The style is similar to my game NoThree for Android (https://play.google.com/store/apps/details?id=com.ovaludi.no...).

It seems to be the case.

Playing this game requires using the mouse all the time, whereas playing 2048 only required the four arrow keys on the keyboard. I use the mouse only when I have to. Unfortunately, requiring the mouse makes the web interface of this game klunky.

Edit: 20 minutes later, and this is a big deal. I'm sure it's possible to enable the four arrow keys and enter key so we can use the arrow keys to navigate around the grid, and the enter key in lieu of the mouse left click. Other keys on the numeric pad could also be utilized, such as + for the "eye" hints.

You should PR it in! The source is on Github: http://github.com/q42/0hn0

The game is weirdly dystopian for something made from circles and two colors.

Could you elaborate on what you mean by weirdly dystopian? I don't feel like I got the same vibe.

I would like to play without the hints when the board is full.

Yeah this makes the end game too easy.

Even if this is a different game it somewhat reminds me of countless games of minesweeper 20+ years ago. Luckily (or unluckily?) I'm no more as addicted to these kind of games as I used to be or I would spend all the night playing this one :-)

This is a fun little game. Starts out hard, until you start learning some of the rules. 1's are your friend, as well as large numbers. It'll get easier as you learn more patterns, but there are some that are a good challenge.

Interesting take on minesweeper

Oh cool - another version of Logic Dots: https://itunes.apple.com/us/app/logic-dots/id908720413?mt=8

Here's the older one: https://news.ycombinator.com/item?id=8601774

How is that the same :/

EDIT: thanks! Now I don't have to wonder about the potential effects of undo. I just have to wonder why I didn't figure out which icon triggers it...

Perhaps time is a distracting concept in that icon?

among the five buttons at the bottom of the page, the one next to the leftmost button is undo.

I know this late comment is likely to be ignored, but: on Android, my battery drains dramatically fast while playing this game. Curious what they're doing that uses so much CPU...

The game somehow reminds me of go. Sometimes you have to "grow" the blue dots, and sometimes you have to stop its "growth". Like playing both sides at the same time.

I get this a lot, so many games seem to remind me of Go.

I think it's because Go is like the Lisp(or maybe just λ-calculus) of games - as my co-worker put it: remove any of the rules of go and you would no longer have a game.

Very addictive! Reminds me almost of a reversed minesweeper.

Nikoli has a lot of interesting puzzles.

[0] https://member.nikoli.com/index.html?lang=en

If you like this puzzle/game you will love nikoli.com

Reminds me a lot of nonograms - http://www.puzzle-nonograms.com/

It's more like http://www.chiark.greenend.org.uk/~sgtatham/puzzles/js/range...

This one especially: http://www.puzzle-nurikabe.com/

Like Minesweeper crossed with Sudoku (and with perhaps a dash of Reversi/Othello for good measure). Clever.

and a helping of picross http://picross.net/

though this implementation is a little gimped- rotating to the correct of the three possible states (on,off,nil) often requires two clicks, but a doubleclick is no good..

I don't enjoy playing these games but I might use it as practice and write a CSP solver on the weekend :)

Just a head's up to the creator if he reads this: When I tried to tweet my score the tweet was too long.

I've found that the same sort of thinking that's useful for nonograms is useful for this game.

190 on size 8 and back to work :D

Can you make it bigger than 8x8?

So who will be first to make a 0h-n0/2048/flappy-bird mashup?

Perhaps it should be called 5h n1 if it "goes viral"...

The hints are fantastic

Everything on HN looks so square after playing that game...

Brings back fond memories of playing minesweeper.

Minesweeper was timed though, so the point was to go as fast as possible. But similarly to this, there was not much thinking involved after a bit of practice.

edit: just noticed you can enable timed game here, still liked minesweeper better :P

0h h1 seems more fun to me.

Nice Game

Isn't this just minesweeper?

Reply on news.ycombinator.com