Suppose there's a bag-emptying situation, and you're simming two plays: a two-tile play and a five-tile play. Then suppose that the vast majority of iterations in a many-plies sim go like this.
us: TwoTilePlay / opp: Play / us: Play / opp: Outplay + PointsFromOurRack us: FiveTilePlay / opp: Play / us: OutPlay + PointsFromOppRack In a two ply sim, we would only be adding TwoTilePlay - Play + Play, which could be 40 points off in some situations. John On 2/27/07, Winter <[EMAIL PROTECTED]> wrote: > --- In [email protected], "John O'Laughlin" <[EMAIL PROTECTED]> > wrote: > > > Use many plies when the game is far enough along that a 2 ply sim's > > iterations would mix pre-endgame and endgame or endgame and game-over > > situations. If there's a chance that the last move in an iteration > > will empty the bag, use many plies. > > Still not quite clear on this. > > So let me ask this. > > While simming my Saratoga games I tried to use that empty-the-bag > heuristic to determine when to switch to many plys, and what I found > was that sometimes the results did not pass the sanity check. For > example, my play would be shown to give up 40 equity points and then > when I switched back to 2-ply it would only give up five. > > What accounts for such a great disparity in valuation?
