does not see a simple winning move

Issue #12 on hold

dsmic created an issue 2012-02-17

move 277

Comments (6)

dsmic reporter
winning move is j14 by the way:)
- 2012-02-17T21:55:04+00:00
dsmic reporter
If I turn of progressive widening it finds the move:(

Another example where the progressive widening is killing the move. I think I have to understand this concept in more detail, so I will read the papers first. Can you tell me the one you used for implementing?
- 2012-02-18T07:45:26+00:00
dsmic reporter
OK, I had a look into Remi's paper and into the implementation t(n+1)=t(n)+(a*b^n)*(1-c*p)

If I calculate this, is it correct, that we only look at the first 17 moves suggested by the playout heuristic, if I have 10000 simulations?!

Just to be sure I am understanding correctly...
- 2012-02-18T10:15:58+00:00
Francois van Niekerk repo owner
- changed status to on hold
Your comment sounds about correct. This is a side-effect of using progressive unpruning/widening. I don't think there is currently an easy way to fix it.
- 2012-07-09T14:02:41+00:00
Francois van Niekerk repo owner
- marked as enhancement
- 2012-07-16T12:45:36+00:00
Francois van Niekerk repo owner
- removed responsible
- 2016-08-22T12:33:35+00:00
Log in to comment