Suppose you could wall off one of your pieces in the beginning. How many moves would the walled off piece have to have in reserve to make it worth it?

I would have to say it depends on how you play. The thing is, if you're good att keeping you other Ama's alive and eating away on the rest of the board, 3 or 4 squares might be enough. Cause your opponent would have to get at least that much. The bigger the better (of course), but its hard to say a minimum. The same goes for the reverse, how big a area can you 'give' your opponent. That depends on how good you are at getting an bigger area for yourself.

Well I read on some website that trapping a queen is worth around 10 points. So if you can trap a queen in an area less than 10 squares, it would be worth it.

I keep swinging back and forth between trying to trap/limit opponents queens and making big areas for my own. It's really a fascinating game. :-)

What was the web site? My google searches haven't found any strategy sites.

i haven't found anything about amazones strategy, too, thus i written something myself. discussion and additions welcome … :-)

amazone strategy

general

\* enclose your opponent in small boxes

\* enclose yourself in big boxes

\* keep your amazones as mobile as possible

\* reduce your opponent's mobility as much as possible

\* don't place your amazone on lines, she shares with an another amazone of her own color

opening game

\* move your amazones away from the board's border

\* move your amazones into your opponent's territory

\* spread your amazones over all areas of the board

\* don't create too many orthogonal borders

\* shoot arrows over long distances, else you limit your own mobility, too

mid game

\* create diagonal borders; these still enable your amazone

-- starting from particular fields only – to move to another area of the board;

in addition she can move back to her previous area or close the border immediately.

\* don't create too many orthogonal borders

\* if your opponent's amazones are spread over all areas of the board:

\\ then stay nearby your opponent (to prevent him from enclosing himself in a big box)

\* if your opponent's amazones are concentrated in a certain area of the board:

\\ then try to limit the mobility of more than one of your opponent's amazones with one move

\\ then consider to prepare building one big box for one of your own amazone

\* don't stick to your original plan, but apply a flexible strategy

end game

\* close borders

\* move your amazones to border fields of boxes

\* approach your opponent's amazone that way, that her access to the bigger area is blocked by your amazone

\* count score before you resign, some areas are not that big or small they look ;-)

A very short chapter on stategy is at file .\src\gamazons.html after (download and) unzip of http://www.yorgalily.org/gamazons/src/gamazons-0.83.tar.gz

quote:

One way to judge a good move or position is by determining the number of squares you can move to vs the number your opponent can move to. You can exert good control over your opponent by limiting his options. The bad news is, a computer opponent can calculate this a lot faster and more accurately than you can. If this is your only strategy, you're going to get wupped.

we 've had good success by starting out with a full out attack on one amazon and then working on positioning the rest in good locations. Try not to put your amazons too close to each other, try and move them to separate places so they own different squares.

Oh, and don't ever let this happen:

http://www.littlegolem.net/jsp/game/game.jsp?gid=25371&nmove=41

I had a huge territory under control, but I allowed my opponent to seal it off as no-mans land! I had previously noticed he could seal off one side, but I had an back entrance with my other amazon. On this move I forgot and sealed the back entrance, and my opponent sealed the front, and I was dead! LOL!

A web site that mentions a possible value of trapping an amazon is Jens Lieberum's paper on Amazong's evaluation function.

Updated link

Suppose you could wall off one of your pieces in the beginning. How many moves would the walled off piece have to have in reserve to make it worth it?

As a rule of thumb: It's worth to wall off the opponent's Amazon if the remaing territory is smaller than ( Board size - number of Amazons - number of half moves already played ) / number of Amazons.

For example, when each player has already submitted 10 moves: ( 10 * 10 board size - 8 Amazons - 2 * 10 half moves ) / 8 Amazons = ( 92 - 20 ) / 8 = 9 points of territory. The player who manages to secure an average of 9 remaing moves for each of his Amazons will win the game.

I think Frank's formula gives a number that's too low. Essentially, he's saying that there are x free squares, there are y amazons, so the average space available to each amazon is x/y. But that ignores the offensivs potential of each amazon. Once one of your amazons is walled off, you can be tactically outmanouvered so that your remaining amazons don't get their 'fair' share.

If you've got a piece that's got low mobility and/or is in danger of being walled off into a smaller region, then take the 'average' region. But otherwise, I think Frank provides more of a lower bound than a target.