• Welcome to the Speedsolving.com, home of the web's largest puzzle community!
    You are currently viewing our forum as a guest which gives you limited access to join discussions and access our other features.

    Registration is fast, simple and absolutely free so please, join our community of 40,000+ people from around the world today!

    If you are already a member, simply login to hide this message and begin participating in the community!

2048 puzzle

Kit Clement

Premium Member
Joined
Aug 25, 2008
Messages
1,631
Location
Aurora, IL
WCA
2008CLEM01
YouTube
Visit Channel
Lackluster effort after the 2048, but hell, after getting 1024+512+256 three straight times and losing each time, I'll take it.

6uPlLZz.png
 

JTW2007

BattsMan
Joined
Nov 29, 2008
Messages
1,591
Location
WCA
2008WARL01
At last I can go do other things.

VFRZIk4.png

Tip: Sometimes it's possible to get into a spiral pattern where your highest tile is in top left and the top row decreases in value from left to right. You build up the (4, 2) tile and then pair with (4, 1), and then the whole thing cascades leftward as long as you have consecutive powers of two in the top row. This has an added benefit which is that since you don't normally need the whole rest of the board to build, biggish powers tend to accumulate in the middle left side of the board, allowing you to sometimes pair upward into the top row from there.
 
Last edited:

JTW2007

BattsMan
Joined
Nov 29, 2008
Messages
1,591
Location
WCA
2008WARL01
I messed up, but I was well on the way to 4096.
EYZgamw.png

I think I figured out a method that works fairly well:

Notation: Up/left/right/down arrows are U/L/R/D. Squares are indexed like pixels, in pairs of (x, y) starting at (1, 1) in the top left.

Use U and L to build as large a block as possible in (1, 1). Avoid shifting this block to the right at all costs. If necessary, burn off turns by mashing U until the top row fills. When the top row is static (no horizontal combinations or open spaces) you can make R moves without fear of displacing your large block from (1, 1). You should have inadvertently built up some large numbers toward the left of the middle two rows. Now that you can use R moves, move these over and combine them into the top when possible, and then collapse the top row as far as possible to the left. This should ensure your largest term always stays in (1, 1). Your goal is to have the top row occupied by consecutive powers decreasing rightward.

Eventually you'll end up with a pretty big number in (4, 1). Now you begin a new phase where the goal is just to grow (1, 4) so you can collapse left again. Specifically, you want to increase the value of (4, 2) as quickly as possible so you can combine up into (1, 4) (I'll try my best to describe how I do this but it's a little unclear to me still). You can separate out the new spawns in rows 3 and 4 using U and R, so that you're sending larger numbers toward (4, 2) from both the left and the bottom at the same time. You should be able to get to a point where no matter what the new block is, you can combine it in one direction or the other. It is very likely that one of these chains will make it all the way to your build square. Occasionally, the zugzwang will force you to do an L. If the new block does not spawn in (4, 2), go back immediately and continue. If the new block does spawn where you were building, you'll have to start over with the building-up process and look for an opportunity to combine your old (4, 2) tile up into the top row.

Eventually, new spawns in your build square will have forced you to move a lot of high-value tiles to the upper left. When this happens and you can't combine them with R/L/U, you'll run out of space to build and need to drop down a row. Now you can build in (4, 3) using the bottom row and left two columns to maneuver in the same way you did when you were building (4, 1).

One thing I noticed in this method is that it's very hard to get stuck with an uncombinable rectangle on your turn. Since you're using the 2s and 4s to build both horizontally and vertically, you'll almost always have an asymmetrical board, and when you do have a rectangular board, you'll be almost guaranteed to have either U or R/L as a legal move.

I'm still trying to figure out a way to protect the top-row block order from being disrupted. I sometimes get forced to make R moves when I don't want to, and if the new block spawns in (1, 1), it's pretty much gg. Anyone have any ideas for improvements to this system?
 

Kit Clement

Premium Member
Joined
Aug 25, 2008
Messages
1,631
Location
Aurora, IL
WCA
2008CLEM01
YouTube
Visit Channel
It's nothing special, really. It's mostly what JTW does, but I avoid building in (1, 4) and instead build my 4th largest number in (2, 3). Corners are just too hard to build with. The next two largest I try to have as (2, 2) and (2, 1), but given the added flexibility of working with the middle of the board, this can change depending on the situation. I use (1, 4) and (2, 4) as filler, and build them up so I have extra pieces to both block off my snake pattern and have some pieces that can often bail me out of trouble.
 

Jaycee

Member
Joined
Jul 17, 2011
Messages
1,843
Location
Crestwood, Illinois
YouTube
Visit Channel
^My friend says he likes the doge one because "it requires less thinking". >__>

Just beat 2048 for the second time! I'll edit with score when I finish the game. :3 EDIT: Lost shortly after; I got to a point where I had to move my 2048 tile from the corner and it went downhill from there. The biggest thing I got afterwards was a 256. Final score was in the 22000s, doesn't beat the score from my other win which was 26k.

I think I caused this game to spread like a disease through the honors and AP population of my school. A classmate saw me playing while we were watching a movie clip in English, he started playing and showed it to a friend of his, and that friend showed it to more people and it spread from there. My AP Government had a 5 minute conversation about the game when we had some free time, and half the class (15 kids/30) were discussing the game. Yay me! Lol.
 
Last edited:

TDM

Member
Joined
Mar 7, 2013
Messages
7,006
Location
Oxfordshire, UK
WCA
2013MEND03
YouTube
Visit Channel
We could also have little competitions on the lowest number of moves to get to certain numbers. Such as 56 moves to get to 512.
Did you really get to 512 in 56 moves? Is that even possible?
20 for 32
37 for 64
82 for 128
155 for 256
196 for 256+128
230 for 256+128+64+32+16+8+4+2
270 for 256+128+128
311 for 512
 
Top