Three Hundred :: Mechanic #172
  Squidi.net Sites:     Webcomics   |   Three Hundred Mechanics   |   Three Hundred Prototypes
 Three Hundred
   - Index Page
   - About...
   - By Year

 Collections
   - Comp-Grid
   - Procedural
   - Tactics
   - Tiny Crawl
   - Misc


 
PreviousMechanic #172Next

  Mechanic #172 - Clairvoyance
Posted: Jan 07, 2014

See the future and change the present. Obfuscate the future to confuse your opponent.


  Clairvoyance

 

[future1.png]

Fig 172.1 - Outcome seems likely.

This is a game of Chess played with the ability to see the future. Next to the Chess board, the next 10 turns played out in an optimal manner (as decided by the Chess AI) are displayed and can be cycled through, starting with the current turn. Also mentioned, is who would eventually win if the game played out in this exact manner.

Say you are playing White and you look over to the clairvoyance board. You can see that the optimal move for this particular turn is to move the rook up a few squares. However, it also says that if the game continues in this exact manner, Black will ultimately win the game. You can scan through the next ten potential turns to see if there will be a future opening that is better for you. There is, but you aren't willing to sacrifice your Queen in four moves to do it, so you decide to move your Bishop instead. Those nine turns are now updated with the new starting place, and a tenth turn is added to the end. The result is the now, White will win.

During Black's turn, the player sees that White is going to win ultimately. However, as long as this is the case, the White player becomes more predictable, as he will be more likely to follow the predicted moves to an assured victory. Black decides to follow the game plan for a few turns, before springing a Kirk-worthy move of defying illogic.

Having the next ten turns played out in theory, along with the potential winner made explicit, changes every decision a player makes in the game. No longer is it simply a game of Chess between two players. It is now a game of Chess between two players and fate itself.


  Obfuscation

It wouldn't be the Three Hundred if I didn't take it just a step further. Each player also has a limited number of ability to obfuscate the future in various, predictable ways. I'm thinking maybe a handful of cards that can be played on their side of the clairvoyance board.

You play an obfuscation card on a particular future turn (one per turn), and it will be active for a set number of turns after that. So you can play card A three turns from now, and it will last three turns (turns 3-5). You can also play card B four turns in the future, so on turn 4, you have both card A and card B active.

You can see what card will be played on which future turn, but you can not see what cards are currently active for your opponent (i.e. you can tell that card A is active for the next two turns, but your opponent can not. Both of you can see which cards will become active next turn).

The purpose of the obfuscation cards is to change the prediction of the game in very specific ways. It could be something as simple as hiding the Black player's turns from the list, or hiding turns 1-3 and showing only moves from the more distant future. But it can also get more interesting.

 

[future2.png]

Fig 172.2 - Things are hazy. Ask again later.

Here, the clairvoyance board has a Hazy Future card played on it by the White player. What it does is see multiple futures simultaneously. In the first move, the White player could move any number of pieces. The computer picks two of these moves - the optimal move and a suboptimal move - and displays them simultaneously. The difference in the board is shown through transparent pieces. The pieces that are not transparent are the same between both moves, while those that are have been uniquely moved in their version of the future. This happens for each of the 5 White player moves on the clairvoyance board, causing future board settings to be even fuzzier. It's sort of like if you were to display every possibility of [#009 - Time Shock] simultaneously.

Some other obfuscation events that could have potential:

- Goal Cards - Will change the predictive AI to create optimal moves towards a goal other than a general victory. For instance, you can play a goal card to gain a queen, and the AI will prioritize getting pawns to the opposite end of the board, or a kill the queen goal, where the AI will prioritize taking the enemy queen. Use goal cards to give the predictive AI more personality.

- Hide Turns - Will black out individual turns or even portions of the clairvoyance board (like the center 4x4 squares).

- Horsies Only - For the next move, the computer will only move the player's knights when predicting moves, regardless of whether it is the optimal move.

- Opposite Day - For the next random number of turns, the prediction of who would win the clairvoyance game is the opposite of what it should be.

- Suboptimal Turns - At some random turn after this card becomes active, the optimal move is NOT displayed, and is instead replaced by a randomly available move that does not result in the loss of the game or the loss of a playing piece.

- Illegal Move - At some random turn after this card becomes active, one of the computer's predictions will be replaced by an illegal move (pawns can't move backwards). Mostly used to make sure your opponent is paying attention and not looking too far in advance without considering what happened

- Winning Move - Rather that picking the optimal move for the current board (which will result in your opponent winning down the line), the computer will display the most optimal move that will change you into the fated winner. Be careful that this card doesn't activate while Opposite Day is active.

- Pawnfrontation - Why not? All the pieces on the clairvoyance board are transformed into pawns, making the clairvoyance board completely useless.


  Conceptual Notes

I'm starting a new section where I discuss the thought process I went through when coming up with an idea.


- The idea was originally formulated with the concept of the AI trying to guess how each player would play. The AI would watch the moves the player performed and take note of certain things, like goals attempted or favored piece usage. Then I realized that the very nature of seeing the future would change it. Not only would the AI be terrible at guessing what moves a player would make, it would be impossible because the player's moves would change based on what the AI guessed. By predicting the future, the AI would be changing it.


- The idea was also not originally Chess. Instead it was a typical SRPG-style tactics game, with the special abilities of the units being what ultimately became the obfuscation cards. So, you'd have a wizard that could cast a spell that would make the future predictions hazy.

I realized, however, that the idea worked best with a very specific set of behaviors. The first is that actions needed to succeed or fail, nothing in between. When a unit can do between 4-6 damage, you end up with multiple futures: the unit misses, the unit harms the enemy but not enough to defeat it, and the unit defeats the enemy. Those possibilities make predicting the future more random, and make reading future moves more confusing (I did what the computer suggested, but I got a different outcome!).

The other thing is that the idea works best when there are a limited number of moves each turn. If you can move every friendly unit each turn, then the difference between each turn would be impossible to read and predict. If only one unit moved per turn, there's a much clearer line of behavior from A to B.

I realized that Chess met these conditions perfectly and thought to myself, "great... now I have to design a Chess-like game". Then the little voice in my head said, "If Chess is perfect, why don't you just use Chess?" That was a duh moment for me.


- I had trouble drawing the Chess pieces. I've used the pawns before, but now I needed a full Chess set. It took several tries to do the knights, and I couldn't get the bishops to look right at all. I got something which was good enough. I wasn't happy with it, but it wasn't awful. Then I put them on the board and the pieces looked HORRIBLE when put together. Their black outlines clashed something fierce. I shrunk the pawns, widened the rooks, and mess with proportions until the black pieces at least looked decent. The white team still clashes a bit, but it isn't nearly as bad.

 

 





Copyright 2007-2014 Sean Howard. All rights reserved.