2023-09-28 17:34:22 +00:00
|
|
|
# Oxofocus
|
|
|
|
|
|
|
|
A Naughts and Crosses game that learns as it goes
|
|
|
|
|
2023-09-28 18:10:23 +00:00
|
|
|

|
|
|
|

|
|
|
|

|
|
|
|
|
2023-09-28 17:34:22 +00:00
|
|
|
To start with the computer will play random moves and slowly increase
|
|
|
|
its skill based on a set of logic rules. During your first few games
|
|
|
|
the playing logic will apply a new logic rule after each time you
|
|
|
|
win. Once you reach eight wins the banner area will turn green
|
|
|
|
indicating that the computer is now playing at maximum strength.
|
|
|
|
However it is not as easy as you think and if you make a mistake and
|
|
|
|
loose a game your score goes back to zero. The more you play against
|
|
|
|
a weaker algorithm the more likely you are to loose concentration.
|
|
|
|
|
|
|
|
Have you got the focus and concentration to get to maximum playing
|
|
|
|
strength. Do you know all the winning moves to out fox the algorithm
|
|
|
|
?
|
|
|
|
|
|
|
|
Written by: [Hugh Barney](https://github.com/hughbarney) For support
|
|
|
|
and discussion please post in the [Bangle JS
|
|
|
|
Forum](http://forum.espruino.com/microcosms/1424/)
|
|
|
|
|
|
|
|
Credit to `MissionMake` for
|
|
|
|
[tictactoe](https://banglejs.com/apps/?id=tictactoe) where I have
|
|
|
|
borrowed the grid drawing code
|