i can't win yet, just started to find winning pattern today,

http://gabrielecirulli.github.io/2048/

any sure win tips?

I've got 4096 three times. With my technique I can reliably get 2048. Are you sure you want someone to ruin the game by giving you their technique?

=) wow..!!
wish to know, if you found out that technique by observing the pattern,
or you really do some math to get that technique.

No math. It just seemed the most natural way to keep the board uncluttered. The idea is to keep a pattern such that combinations can easily be made and you can keep clearing out the new 2s and 4s by combining them with what you already have.

I kept having all the larger numbers on the bottom left.

Well, if typedef's gonna say it explicitly, I guess I might as well do the same. I fill the top row, ordered highest to lowest left to right, then start the next row in reverse order, and (when the second row becomes full), the third in the same order as the top.

So you get


This allows easy chaining.

ya, tasted my first 2048, =) thanks,
i guess i got the idea already, =)

tasted my first 4096 today =)
so next stage is 8192.

now my highest score around 60k already, and it feels so good...

then i thought about a question.

1. is this game infinite loop forever until gamer crash the boxes?

2. what skill is actually needed to solve this 2048 puzzle? common sense? luck? and what kind of concept did i learn by playing this game?

3. could a computer A.I never crash boxes and could run the puzzle and keep on combining forever?

4. is there a mathematical proof, this game could run FOREVER?

i guess, someone should really do a PHD based on above question, dig deep into it and see what treasures inside.

I don't think so. If you're using the best of algos, you're still limited to RAM. If you want to store bigger numbers (something like 2^(1024)) you need to use RAM. (provided no processor extensions are assumed).
The game will crash after it has used all the system memory.
If my math goes correct the game will crash on a system with 4GB of RAM while storing a number huge as 2^(1073741824) i.e. a 1073741824 bit large number (provided one word on the system = 4bytes and there is nothing else in the memory except the number! ) <--- Someone needs to confirm that.

and it feels so good.. =)

I've seen posted 131072 screenshots. So keep trying.

omg how long would that take what a utter waste of time

131072, really, how that guy play to that number? i kinda doubt it.
maybe they should make some changes to this game,

after 2,4,8,16,32,64,128,256,512,1024,2048,4096
add a new row at the bottom,

I agree with that.

actually, i think game could build some sort of personal confidence, and winning or breaking record in a positive way,

but i am still struggle to understand, what skill did i learn by playing this puzzle,

my highscore is 80k something now, =)

btw,, here a new square game, nice, easier than 2048 =) stages by stages, not rank by difficulty imo,

http://gameaboutsquares.com/

it was featured in hackernews, ma in level 31 now =)



for level 29, you might want to hit undo after first step, surprise. =)

Stupid post removed.

glad you like it, HaHaAnonymous,

btw, i couldn't break level 31, still thinking how to solve it,

Revolution, does a screenshot with a previous highscore (the number in the bottom right) count? I got 133k, but forgot to get the screenshot.

wow, thats is awesome score, Tyler!
so far, my highest score still in 80k+