A puzzle-solving contest.

[ErikH2000]

I run a contest on my website each month. Usually, they are related directly to Caravel games, but this one is for anybody that likes solving puzzles. You don't have to know anything about DROD or any other games to enjoy it. If you've played Aargon or Black Box (Twilight Games) the mechanics will be familiar.

Have a quick look! I warn you that the regular crowd that participates in contests is freakishly smart, so if you play, you will have some real competition.

http://forum.caravelgames.com/viewtopic.php?TopicID=9070

-Erik

[svero]

If you've played Aargon or Black Box (Twilight Games) the mechanics will be familiar.

Have a quick look! I warn you that the regular crowd that participates in contests is freakishly smart

Pfft.. Maybe so , but did any of *them* actually write Aargon? I didnt think so... I'll just stay out of it so it's fair to those guys :-)

[ErikH2000]

Heh heh. You don't know these guys, Steve. They would crush you.

-Erik

[TimS]

So very tempting... You people and your challenges-that-are-tempting-despite-having-small-prizes. First the comedy game competition, now this random solve-the-puzzle thing. Why must you suck valuable time away from my game development? WHY?

[Kite65]

Despite being against the spirit (I read your comment in the forum) of the competition, It would be interesting problem to build a program to solve.

Each square can be empty, forward deflector or back deflector,
there are 64 squares so approx 3 to the 64th power different configurations
[roughly 3,433,683,820,292,512,484,657,849,089,281] :) )

So I don't think a brute force program is practical. Perhaps removing squares that don't have deflectors would reduce the search space to managable numbers. There may be other short cuts. The interesting part, for me, is the optimisation of the program rather than finding the actual solution.

Though solving one with paper and pen would be faster than building the program.

Mind you, I am supposed to be completing my games rather than being distracted by interesting computer science problems :(

[ggambett]

I agree. It would be very fun to write a program to solve the puzzles, but it's against the rules. A different category perhaps?

[ErikH2000]

Despite being against the spirit (I read your comment in the forum) of the competition, It would be interesting problem to build a program to solve.

Each square can be empty, forward deflector or back deflector,
there are 64 squares so approx 3 to the 64th power different configurations
[roughly 3,433,683,820,292,512,484,657,849,089,281] :) )

So I don't think a brute force program is practical. Perhaps removing squares that don't have deflectors would reduce the search space to managable numbers. There may be other short cuts. The interesting part, for me, is the optimisation of the program rather than finding the actual solution.
Yeah, I definitely understand that appeal. With these node-searching problems there is almost always some clever way to optimize the search. Like the Eight Queens problem can be solved orders of magnitude faster than simple node-searching with some simple optimizations.

One way to look at the challenge is that you want to come up with an efficient way for a very slow computer (i.e. your brain) to find solutions. With some practice, there are ideas that will come up. Like one person says that generally, you should start placing deflectors on the edges then work your way in.

Cheap, imperfect algorithms made for human brains can be more interesting than algorithms that perform perfectly but require computers to be practical. Like I have a simple algorithm for deciding how much I'll bid for a hand of Spades that works pretty well and lets me make good bids without holding up the table.

Mind you, I am supposed to be completing my games rather than being distracted by interesting computer science problems :(
Awwwww, all work and no play make Kite a dull boy. Also, a lot of useful ideas for games or other endeavors can come out of screwing around a bit. And it only takes maybe an hour or two to play. Come on. ;)

-Erik

[ErikH2000]

I agree. It would be very fun to write a program to solve the puzzles, but it's against the rules. A different category perhaps?

Hmm. It seems a difficult thing to judge properly because of the different languages and platforms people use and my reluctance to run binaries from strangers on my computer. Posting source code for people to look at (here or there) might draw interest and discussion, but I can't see a contest working for me.

-Erik

[Kite65]

One way to look at the challenge is that you want to come up with an efficient way for a very slow computer (i.e. your brain) to find solutions. Cheap, imperfect algorithms made for human brains can be more interesting than algorithms that perform perfectly but require computers to be practical.

Awwwww, all work and no play make Kite a dull boy. And it only takes maybe an hour or two to play. Come on. ;)
-Erik

OK, I tried the sample puzzle for 20mins, got pretty close, only the FastForward FastRewind beam wasn't connected at the end. But it's like the Rubiks cube, nearly solved is still NOT solved. Sometimes the last details takes 90% of the time to fix.

Quite interesting to play, though. Even If I didn't build a computer solver, an program to handle the display of the developing solution would be a great help. I did my attempt using paint, got a screen shot of the puzzle and then just draw in the deflectors, where i needed them.

I suppose, for me it's similar to Suduko. I can see the puzzle but it lacks the attraction to make me play it as intended. I keep thinking that I could build a program to help and that gets in the way of the puzzle. Perhaps because the puzzle takes quite a long time and there is no incentive to complete it quickly.

I have a puzzle game in the early stages of development, which has both timed and untimed modes. The untimed modes are a bit like the above puzzle, you do them primarily as a challenge, rather than the timed mode, which is simpler and more fun. So I think I have learnt a little about the development of untimed modes in my own game. Ta :)

Unfortunately, with the demands of real life, an spare hour is pretty hard to find. :(