A Pencil And Paper Algorithm For Solving Sudoku Puzzles: Fill & Download for Free

Open the paper and examine the puzzle. Ues only pencil but put some numbers in the proper place. Start by looking for the same mumber in the other associated boxes. When it becomes obvious that you are in, over your head, fold the paper with the numbers facing outside, if possible with the highest numbers on top.Hand it to my wife and tell her, I have solved the puzzle, mentally and that there is a mistake made at the originator, so it cannot be solved properly because of that mistake.Lay the puzzle down, and tell her you will show her where the problem is, in a few minutes after you finish showering (or what ever excuse you can come up with) and go about your business.She will soon tell you how wrong you are, laughing at your failure! She will go on to show you that she has solved it.She will have a big grin on her face but the problem will be solved!You will know that you can use all assets at your disposal!

Because:It has the qualities of science and the qualities of craft, meaning that in some areas you can learn, practise, and master it, while in others, learn, understand, extend, or disprove it. It binds the abstract and the concrete. It (mostly) operates on a solid and sound theoretical basis (often rooted in Math) but does also acknowledge the 'softer parts' via short cuts, heuristics, etc. If you are more into the 'hammer and nail world', computer engineering may be for you.It has a low entry barrier, meaning that nowadays computers are cheap, OSes, compilers, and other software is often free. If that is not convincing, algorithms, flow charts, logical truth tables, UI mock ups, etc. can be created using imagination, a sheet of paper and a pencil.Learning about it has never been easier (think MIT OCW, Coursera, Udacity, CodeChef, TopCoder). Seek out like-minded in the world via virtual communities, such as Quora, StackX, etc. No need to apply for a time share on a computer as big as a house. Run your code locally, on your university's or school's infrastructure or somewhere else (you know 'the cloud').Put together low entry barrier, and a Silicon Valley mind, and you have yourself an entrepreneurial spirit, potentially working on the next big thing (ask fellow art, music, etc. students about what they think about their future job prospects). That is, you can work from almost anywhere (power supply, internet, supply assumed), on anything virtual.The field is relatively young but draws from a rich science history. While there are established areas of C.S. others are very open and vibrant, drawing the young, middle-aged, and old curious minds into their centre.It may teach you one thing or two about the (yet) most powerful computer there is - the human brain.Its employment opportunities are diverse and relatively good.It is sexy, if one believes popular media stereotypes about being a nerd, a geek or whatever you want to call it.A C.S. problem is part of the millenium problems (http://www.claymath.org/millennium/P_vs_NP/), if you solved it, you would be awarded 'some cash'.If intellect is sexy, C.S. has some very sexy stars (think Knuth, v. Neumann, etc.) And who would not want to be around the sexy and the glamorous?!But the best for me, is its interdisciplinary nature and the reason for why I will never get bored. That is, you can:work on real-life applications (cook book, shopping cart, etc.),work on games (board games, social games, etc.),create physical, chemical, biological, metereological, or astronomical simulationsprogram virtual synthesizers to create sound, music and speechcomputationally make sense of the human languagesmake your computer 'think' using artificial intelligencemake your computer compute (in the mathematical sense) or solve your sudoku puzzlediscover the unknown via theorem proovers, etc.help to drive cars, trains, planes, ships, space shuttles, rocketshelp to operate cities (power plant, water supply, traffic control, etc.)help people communicate via phones, electronic mail, instant messaging, etc.help to perform surgery by devising surgery robot control logic

Edit: Previous to the merger of questions, this was an answer to: "Is there a trick to solve evil or super-genius levels of sodku?" I would not recommend this method in general, but for the worst cases, this is the only option.Yes. The point is the evil typically does not admit to a solution by the usual techniques. So you have to use guessing and backtracking (often called brute-force). On the other hand, picking what to guess at any stage is what requires judgement, so it is not really brute force.Go as far as you can until you get stuck. Now take a guess. There are many computer versions that make it easy to keep track of guesses, by allowing coloring of choices. If you are doing paper/pencil, you could use colored pens. Typically you want to make a two-way choice (this cell is either 4 or 7), so you write 4 in red and 7 in blue. Pick a choice where you can see that either choice will resolve many other cells quickly.If you ever find a cell with both the red and the blue answer being the same number, then you can enter that in black, because that is a conclusion that is true in all cases. Eventually you will find that one of the choices leads to a contradiction, which means that the other choice was correct.A good choice might just solve the entire soduku. If you make a bad choice, you can still perhaps get a few black pieces of information from it, after which you may decide to try again, with a different choice. I find that using this I can almost always do the hardest, most unreasonable levels, but it might take a half an hour or rarely more.Edit: Many people are mentioning computer algorithms like depth-first-search. For an algorithm, the fastest is the dancing links algorithm by Donald Knuth, which is just a depth first search, but it makes an intelligent choice about which fork to take, whenever it forks. You might ask why worry about speed when it only takes less than a second. Because if you want to generate a puzzle using a program, you might want to generate millions, and then select the hardest, or the one having particular features. Each generated puzzle would have to be checked for unique solutions, solvability, etc.