Clausewitzian Chess experiment introduction by Rob Seater.
Leveraging simple, well-known games can help us to model complex cognitive concepts through designing modified versions of the game and its rules. This is what a team at Lincoln Laboratory did with Clausewitzian Chess, which is based on the game of chess, and you will do the same in this exercise.
You might already be familiar with how chess works but even if you are not, we will provide all the relevant background that you will need in this exercise. Consider that there are several kinds of chess pieces (bishop, knight, rook, etc.), and each of them moves in specific ways. For example, rooks only move in straight lines. In traditional chess, both players have the same pieces and can see what the other person is doing.
Sounds like a given, right? What if it wasn’t? What if you can only see up to a certain number of positions near you?
The idea behind Clausewitzian Chess is that you don’t always know everything about a given situation. In other words, both sides aren’t on exactly the same footing (in this case, starting with the same number and type of pieces) and you can’t see the full picture (in this case, the full board). Think of the Battleship game where you slowly learn more about your enemy’s location based on your guesses.
Consider a military officer in action. They have to make decisions but don’t have all the information about where things are and what the enemy is doing. Even in these uncertain conditions, they have to make decisions and we would like to understand the fundamentals of this decision making. Clausewitzian Chess, designed at Lincoln Laboratory, allows us to explore exactly that. The feedback that the officers received from having warfighters work through various chess scenarios using this type of chess allowed them to more effectively design training techniques about how to manage environments with fog, friction, and chance than without. For the following experiment, Clausewitzian Chess can also be used to work through our own reactions to situations where we have different levels of incomplete information.
As you progress in this experiment, you can choose the types of pieces each side has, how many, and how much “fog of war” there is. After setting these criteria, you can then run your experiments.
Let’s define a couple of key terms for some of the criteria you can work with.
- Asymmetry – The difference in resources and ability that each side has. In our case, we can vary the type of pieces, the number of each piece, and the visibility of the pieces.
- Composition – The kinds of warfighters, their available skill sets, and their available resources. For this experiment, this means that you can change up the numbers and types of available pieces that you will work with to solve your scenario.
- Fog of war – The unknowns, about your situation or your enemy’s situation, that unfold as you run through the scenario. This is represented by how much you can see of the enemy’s pieces on the chess board.
- Friction – The possibility that your desired actions are not followed, either because the actions are not possible or because the warfighter is unwilling to follow. For example, if the means of communication is down and the orders could not be received. In our case, this means that all pieces of a certain type will not move as you want for that given turn.
Friction is not required for the first two scenarios, but you can try it out in scenarios of your own making.
You will set up three different scenarios to experiment with. But before we start, let’s learn some chess basics relevant to this experimentation.
Test Your Knowledge on Chess
- How much does strategy in chess rely on having a standard chess board with traditional pieces?
- If traditional chess tactics rely on having certain pieces start in set squares and move in fixed ways, what would happen if this was not necessary? How might your strategy change?
- How does having only partial information change the way you act?
