Assignment 2 more preview

- basic step: solver similar to TicTacToe example
- deal with three outcomes, win, loss, draw
- many improvements possible
- move ordering: look at (probably) good moves first
	- find moves that help in more than one direction
   .
.oo..    intersection of two promising
   o     patterns
   o
   .

   .
.ooo.    playing here gives two good
   o     threats
   o
   .

- evaluation function: how good or bad are states?

Example: 
- how many possible 5-in-a-row
are there on the board? 
	- For us? For opponent?
	- How close are they to complete?

- move pruning: recognize threats
- In these situations we can greatly restrict the set of moves to look at

1. immediate win:

xxxx.

xxx.x

xx.xx

xxx.xxx
etc.

2. block opponent's immediate win:

xoooo.

ooo.o

oo.oo

2b. Give up when opponent has double threats
    .
xoooo.
....o
....o
....o
....x

3. win in 2 moves

.xxx..

.x.xx.

- optimization:
	- where does your program spend time?
	- can you find smarter ways, 
	  save computation?
- important goals: keep soundness and 
  completeness
	- you program should never give 
	  a wrong result
	- your program should never cut 
	  all the winning moves
	- some rules are only heuristics, not 
	  exact. Can still use them for move ordering, but not as absolute truth