++ed by:

1 PAUSE user

Klaus Wittrock
and 1 contributors


Games::Sudoku::Trainer::Obstacle - get over an obstacle while solving a Sudoku puzzle


This part of the documentation for SudokuTrainer aims at people that want to use SudokuTrainer to get over an obstacle while solving a Sudoku puzzle by the use of Sudoku solution strategies. If this isn't what you expected, please inspect section "GUIDE TO DOCUMENTATION" in Games::Sudoku::Trainer::General_info.


    perldoc Games::Sudoku::Trainer::Obstacle


You are stuck in the middle of solving a Sudoku puzzle, and you don't know why. In the following sections, steps 1 to 3 describe the normal procedure to find a continuation. Steps 1a and 1b give suggestions how to deal with special situations that may occur in step 1.

                  -->| Step 1a |-->
                 /   +---------+   \
                 |                  |
   +--------+    /                   \     +--------+    +--------+
   | Step 1 |---+---------------------+--->| Step 2 |--->| Step 3 |-->\
   +---+----+    \                   /     +--------+    +--------+    |
                 |                  |                        A         |
                 \   +---------+    |                        |         V
                  -->| Step 1b |-->/                         \--------/
                     +---------+                         (repeat train)

Step 1: See how far the trainer could solve the puzzle

You feed the puzzle to the Sudoku trainer to see whether it can help you. The first question is: Can the Sudoku trainer solve the puzzle, and which strategies does it use? You select pause mode non-stop and click the Run button. You see that the puzzle got solved. You select History | Summary and look at the strategies used.

Step 1a: Change the priority list

In step 1 you see that the trainer solved the puzzle, but used a strategy that you don't know. Next find out how far it can go without this strategy. Restart the puzzle, select Strategies | Change, move this strategy to the end of the priority list, then select Pause mode | Strategy | <this strategy> and click the Run button. There are several possible outcomes:

  • The trainer solved the puzzle without a pause. So now it got solved without this strategy. Save the modified priority list in a file, then proceed to step 2.

  • The trainer pauses, and there are values on the board that you didn't find. So this strategy is needed to solve the puzzle, but you can train the strategies that are required to find all the values that you didn't find. Save the modified priority list in a file, then proceed to step 2.

  • The trainer pauses, and all values on the board have also been found by you. So you have solved the puzzle as far as possible without this strategy. There is no help, you have to learn it.

Step 1b: Work with a partial solution

The trainer couldn't solve the puzzle in step 1. Nevertheless you may still use it to train a strategy. Compare your partial solution with that of the trainer. You found all values that the trainer did find? Sorry, then it cannot help you. Otherwise proceed to step 2. Your puzzle will probably remain unsolved finally, but you got nearer to the solution and you trained at least one strategy.

Step 2: Find the key point

You know all stategies that the trainer used to solve the puzzle. So you want to watch the trainer while it finds the first value that you didn't find. To do so, you have to pause the trainer at the key point, i. e. at that state where all found values had also been found by you, but the next value will be a new one. This sounds trivial, but it isn't, since the trainer proceeds in a sequence different from yours, and you hardly can guess which value it will find next.

Here are three alternative procedures for finding the key point:

  • Start from your partial solution

    Enter your partial solution as the initial puzzle. This is often the quickest way: you are immediately at the key point. However, there is one risk: maybe there is an error in your solution. Then you are on the wrong track.

  • Look at the history

    After step 1, select History | Details and search for the key point in the list.

  • Step through Value found

    Select Pause mode | Value found | anywhere and click the Run button repeatedly until the new value displayed (marked by a black rectangle around it) is one that you didn't find. Remember that the trainer is one value ahead, so the key point is the predecessor of this value (marked by a gray rectangle around it).

Once you found the key point, write it down; you will need it in the next step.

Step 3: Train the next strategies, beginning at the key point

Obviously you need some training on the strategies that lead from the key point to the next value, otherwise you would have found the next value by yourself.

Restart your puzzle, select Pause mode | Value found | at a cell, enter the name of the key point, then click the Run button. The trainer will pause at the key point. Now select Pause mode | single step and click the Run button. The trainer pauses at the next successful strategy. Try to find it with minimum help by the Details button, as is explained in section "Train the strategy" in Games::Sudoku::Trainer::Training. Then click again the Run button and try on the next strategy, and repeat this until the next value is found.

Repeat this whole step several times, trying to succeed with less help, until you are confident that you can find the first new value by yourself. Maybe you prefer to solve your puzzle first, but don't forget to return to the trainer later.