Skip to main content Skip to navigation

Workshop 3


Workshop 2 will give you some idea of the modelling techniques and components that are involved in building the Web-Eden Sudoku model. This is what computer scientists call a "bottom-up" approach to understanding a complex object. In this workshop, you will be exploring the model "top-down", investigating the relationship between the entire model and the features that are of interest when thinking about solving a Sudoku puzzle. These features exemplify another of the key concepts of Empirical Modelling: the observable. An observable refers to anything to which a status or value can be ascribed - as happens when the term is used in science, it does not necessarily refer to something whose status or value is directly visible

Task 1: Observation and Agency in Sudoku puzzle solution

This task will introduce you to some of the key observables and mechanisms within the Sudoku model that can be used to assist solving. In this task, you will be expected to understand how several different observables associated with the Sudoku model are linked through a complex network of dependencies in such a way as to enable you to represent partial knowledge about the contents of cells and the constraints to which they are subject. You will also get some impression of the rich interplay between observables in the external world and observables in the computer model that is characteristic of software engineering. Now available at Workshop 3A.

Task 2: Applying the Sudoku model in puzzle solution

In this task, you will be building on the knowledge of key observables and mechanisms gained in Task 1 to explore ways in which you can apply the Sudoku model in puzzle solution. An important feature of the model is that it will enable you to express knowledge about a puzzle that does not directly determine the digit in any particular cell. By exploiting the model within the Web Eden environment, such knowledge can be recorded and communicated to other human solvers. Now available at Workshop 3B.

Task 3: Investigating strategies for collaborative solving

In the final part of this workshop you will be working as a researcher with a view to investigating how sudoku might be solved collaboratively. Such 'collaborative solving' means that your sudoku grid is shared with others and you can all contribute to solving a puzzle. The link to Workshop 3C opens a Web Eden environment where the sudoku grid is shared. In this environment, you can load 9 new puzzles that you can all work on together. Some puzzles are more difficult than others. The idea is that you might solve a small part of the puzzle and then someone will come along later to continue solving. Another possibility is that if two or more people are connected to the same puzzle at once then you can solve it together at the same time.

In order to ensure that you can find other people to help you collaboratively solve a puzzle, we are recommending some specific times to look at particular puzzles:

Puzzle No.    Time
1        Friday 5pm
2        Friday 6pm
3        Friday 7pm
4        Friday 8pm
5        Saturday 10am
6        Saturday 11am
7        Saturday 12pm
8        Saturday 1pm
9        Saturday 2pm

You can try to solve these puzzles at the above times, or you can try solving any puzzle at any time. Remember that at any time you might be able to help someone else complete a puzzle that is already started, or you might be able to start a puzzle that someone else can complete later.

This should give you an idea of what it is like to solve sudoku collaboratively. For the task, you should consider different ways that collaboration could be organised. Rather than letting each player solve in their own personal way, it might be a good idea to assign each player a particular task. For example, one group of players could use colour, and another group of players could use the original black and white version. Or different players could look at different digits. Use your knowledge from the previous tasks in this workshop to think of new collaborative strategies for solving sudoku. The forum would be a good place to discuss these strategies!

Creating this first prototype of the collaborative solving environment has  given us a potentially very rich body of ideas about what different modes of communication might be appropriate, and how EM principles might be best exploited in that connection. If you study the second part of Workshop 3A, you will see how it is possible to use the EM Sudoku model to store information about the solution of a puzzle that isn't explicit in the grid. As of now, we have not incorporated information of this kind in our collaboration environment. One reason for this is that it is hard to make it explicitly visible to participants, and this could be very confusing. If there is sufficient interest over the next days (I know many of you will be glad to be on holiday, and perhaps not keen to sit indoors solving sudoku collaboratively in such glorious weather as we are sure to have), we shall be adding features to the collaborative environment as and when they become available. This is entirely in the spirit of Empirical Modelling. What is quite as important to us is your feedback though, because there may be the chance to adapt what we are doing to take account of ideas you obtain through taking part in the experiment.  And, once again, please use the forum to share and discuss your ideas with others.