Could you explain Numberwang?
Here’s a video to help.
The Doctor Who version is pretty wonderful.
I like how it’s different; it makes it less of a clone.
I’ve probably spent north of 30 hours playing the original 2048, but the fact I can play the Doge version quite effectively right off the bat suggests to me that I’m not really using any mathematical parts of my brain in doing so. Interesting. I still haven’t really figured out how am I solving it.
Nonsense. There are 13 digits. In the original game, these are labeled two, four, eight, sixteen, thirty two, etc, but that’s just a scheme to differentiate them, and order them by value. The only arithmetic operation involved is (n+n)=n+1
Granted, the numeric values assigned do a great job of differentiating the high level tiles, but you must be good at ordering the doge.
But even without numeric values, this game exercises the same skills used in a variety of other maths. Discrete math, for instance.
2048 is a ripoff of a game called 1024, which is a ripoff of a game called Threes.
Threes is a beautifully designed game for your iPhone and iPad, and there’s even an Android version now. It’s well worth a couple of bucks: http://asherv.com/threes/
Checkout out this great android app for the game.
I’m surprised nobody’s mentioned 9007199254740992 yet
Have you tried just typing up, left, down, right, over and over in sequence? I find it tends to score me roughly 2500-3000 80% of the time.
I mean this in the nicest, nicest way possible, @jlw: fcu.k y.ou. I have now lost 3 hours of my life and it looks like it’s going to turn into 300. Thank you SO much.
There’s a very effective strategy for 2048:
- always keep the largest tile in a corner (I prefer lower left)
- keep the next-largest tiles, sorted, in the adjacent row (bottom)
- never press “up” except when no other move is possible
- try to always keep the bottom row full so that new tiles wouldn’t push the largest tile out of the corner when moving sideways
- treat the remaining three rows as the assembly line that works on the tile to be merged into the lower right corner
It may be counterproductive to play strictly by these rules at the beginning (until you get tile 256 or so), but it really works quite well. There are of course some nuances you have to figure out on a case-by-case basis (e.g.: what to do if the largest tile is pushed away from the corner, or how to deal with the cases when you do need to go up).
The extra space makes it very forgiving and easy to automate. I set up an infinite loop that simulated keypresses “left, down, right” (originally just “left, down” with manual presses of “right” when it got stuck, but I got tired of that) and this mindless bot got a score of 4,468,524 and largest tile 131,072:
For Linux users who want to try it out, the Bash command is:
while true; do xvkbd -text '\[Left]\[Down]\[Right]'; done
Beware that the keypresses are X-server-wide, so you probably won’t be able to do much else while it is running.
I was doing fine until I accidentally hit the spacebar
Brilliant, or rather, THANK YOU for releasing me from algorithmic hell. Because you don’t just play a game. You have to game the game.
Excellent. But, let me suggest an improvement:
while true; do xvkbd -text '\[Down]\[Left]\[Down]\[Right]'; done
This seems to be a bit more effective at keeping the board relatively clean.
The direction codes for GM.move() are: 0 - up, 1 - right, 2 - down, 3 - left. The number in setTimeout is the delay (in milliseconds) between subsequent moves. I’ve been running this for several hours now, it got to tile 2,097,152 and score of over 62 million so far and keeps going.
“addicting” is not an adjective.
This topic was automatically closed after 5 days. New replies are no longer allowed.