Started www.24theory.com a year ago. First, just the solver, then the single-player game, then the challenger game, twitter/facebook daily game.
Now it’s one of the top 1 million sites on the internet according to Alexa. I know it’s a small step, but still. Quite happy about what has been achieved. Alexa top 100,000, here we come!
Check out the stats: http://www.alexa.com/siteinfo/24theory.com
Now you can challenge your friends and see who’s faster in solving math 24 puzzles. Or if you are a math teacher and would like to assign your students some puzzles after school, you might find this useful.
Have fun!
The old game dies on (3,3,8,8). Someone sent me this screen shot, a dead browser:
Apologies, there was a stupid bug in the javascript code. Now it’s fixed. Game on.
I apologize again.
After several weekends’ of work. Now you can play 24 the math game here.
Have fun and make it to the record books!
Hopefully I will collect enough data to rank the difficulties of the 1362 puzzles.
now go play!
I recently got an email titled “Great Site !!! ” Emails like this make me want to put more effort on math 24 game and put more content on this site. With Manuel’s content, I post the edited email below:
Hello, my name is Manuel, i am from Barcelona, Spain.
First of all would like to thank you for all the information i got from your site.
24 is a great game a friend of mine teach me two weeks ago.
I am just learning Java and coded the game. Brute force algorithm building all permutations given for numbers, adding all permutations for signs and finally adding all possibles parentheses. Then evaluating each possible solvable. Founding all solutions.
In the other hand in Rosettacode page i looked carefully all codes, (did not found Java which really surprise me) but what surprise me the most was that all codes there, work like mine, brute force.
I will continue in summer, after my final exams, trying to do my best with the algorithm.
Keep this nice site up.
Cheers !!
Given two quadruples, say (1,2,3,4) and (7,8,9,10), which one is easier for a human to get a solution?
Looks to me 1,2,3,4 is fairly easy, 4×3×2×1, meanwhile, 7,8,9,10 is solvable with 9×8/(10-7) , but slightly more difficult in my opinion.
Turns out (7,8,9,10) has only one solution shown above, while (1,2,3,4) has three distinct solutions (3+2+1)×4, 4×3×2×1 and (4+2)×(3+1).
It seems to me there are three factors that decide the difficulty level of a solvable quadruple:
1) number of solutions, the more the easier.
2) type of operations, I cannot say for sure that the order from easiest to most difficult is + – × /, but I know 13/7 is as hard as it comes.
3) size of the numbers, the smaller the easier (it’s easier to do 4 + 3 than 13 + 7, slightly)
Obviously the best way to really figure it out is to develop a game, let millions of people play and collect their reaction time to each quadruple. That’ll be awesome and probably my ultimate goal.
Need to turn on FTP at both levels : https://members.nearlyfreespeech.net/24theory/support/faq?q=FTPDoubleTrouble#FTPDoubleTrouble
Now I can finally rest in peace.
It’s uber fast as I pre-computed all the solutions and store them in a long string. So instead of let a CGI do it in real time. I let a piece of java-script code search for the solutions in the said long string.
马上提供中文简体版 www.24theory.com/cn
Now I am finally done with the new solver. It gives no gimmick solutions. All solutions are entirely distinct from each other according the the rules on the theory page. And I believe my theory is pretty close to how humans judge distinctiveness among different solutions.
Now what? Now I am going to go read Anthony Bourdain’s “Kitchen confidential”.