2048 AI • Hacker News Board

Re: 2048 AI

#21

Post by munificent » Tue, Mar 11, 2014, 4:14 PM UTC

brink wrote: post #16

syntern wrote: post #9

A new browser performance benchmark has born?

It runs based on random numbers, so it's hardly a reliable benchmark.

Just use a fixed seed and it's 100% deterministic. Benchmarks use random numbers all the time. (In fact, PRNGs themselves often make OK benchmarks.)

Re: 2048 AI

#22

Post by jheriko » Tue, Mar 11, 2014, 4:14 PM UTC

very cool.

now... for a who can make the optimal AI competition... :P

Re: 2048 AI

#23

Post by izzydata » Tue, Mar 11, 2014, 4:15 PM UTC

I've been playing this for awhile now and I think I have found that the best method is to only use 3 directions. This forces your highest number into a corner and only spawns 2's and 4's in the opposite corner. You build up numbers that cascade down to the corner. It almost never gets stuck, but if you do I guess you'd have to push the 4th direction you haven't been using.

Re: 2048 AI

#24

Post by ulam2 » Tue, Mar 11, 2014, 4:20 PM UTC

Fuxy wrote: post #17

rcthompson wrote: post #14

Is this game always guaranteed to be winnable, or is it like Solitaire?

I would go with Solitaire sine mine just lost. Did reach 1024 though.

Mine too got stuck at 1024.

Re: 2048 AI

#25

Post by pedrocr » Tue, Mar 11, 2014, 4:21 PM UTC

Interesting. It seems the solution I and a couple of people more came up with of keeping the top value in a corner isn't being used. It's amazing how it's able to keep the large numbers together throughout the game.

The gameplay is much more interesting but it seems that getting from 1024 to 2048 is much harder this way, probably because the randomness accumulates over time making the depth search unreliable. The more structured solution accumulates much less risk so has some margin to deal with this.

Re: 2048 AI

#26

Post by crusso » Tue, Mar 11, 2014, 4:22 PM UTC

While the game is difficult to finish, I'm kind of saddened at how far I can go just by randomly hitting up arrow, left arrow, down arrow, right arrow, etc.

Re: 2048 AI

#27

Post by eck » Tue, Mar 11, 2014, 4:23 PM UTC

What are the rules for new tile placement? Is it deterministic, random, or adversarial?

Re: 2048 AI

#28

Post by sritchie » Tue, Mar 11, 2014, 4:24 PM UTC

Boom, 13,068 points and 2048! I'm feeling some pride in my AI.

http://cl.ly/image/3G1a0e1T0Q2O

Re: 2048 AI

#29

Post by deletes » Tue, Mar 11, 2014, 4:25 PM UTC

How does it decide where the new tile will appear, or it just tries every possibility?

Re: 2048 AI

#30

Post by ovolve » Tue, Mar 11, 2014, 4:25 PM UTC

eck wrote: post #27

What are the rules for new tile placement? Is it deterministic, random, or adversarial?

Random location (uniform). 90% chance of a 2, 10% chance of a 4. This actually made it hard to model the computer move in the search.