New Game: Levee
marks...@gmail.com
black stones.
Mike Terry
news:1180800020.9...@a26g2000pre.googlegroups.com...
It seems to me that the first player has an easy win by simply playing all
moves on her opponents home intersections.
In theory, it might be possible for the second player to force a win before
this happens. In practice, however, it seems quite easy for the first
player to see this coming and concentrate her moves on the right area of her
opponents home intersections to foil any such plan.
I've only played a couple of 9*9 games against myself, though, so maybe I'm
missing some key strategic concepts! (Or maybe the character of the game
changes on larger boards, although I would think the above strategy would
work even better as the board gets larger...?)
Regards,
Mike.
Ed Murphy
While you're working on the 14 squares marked X:
.........
X.......X
X.......X
X.......X
X...*...X
X.......X
X.......X
X.......X
.........
I'll work on the 9 squares marked Y:
.........
.........
.........
....YYYY.
...Y*....
....YYYY.
.........
.........
.........
Disrupting this is surely possible, but I suspect it isn't as
easy as you make it out to be.
marks...@gmail.com
<news.dead.person.sto...@darjeeling.plus.com> wrote:
>
> It seems to me that the first player has an easy win by simply playing all
> moves on her opponents home intersections.
>
game with such an obvious flaw.
Please see:
http://www.marksteeregames.com/Levee_Edge_Fill.pdf
Mark
marks...@gmail.com
> >Mike Terry wrote:
> >
> > It seems to me that the first player has an easy win by simply playing all
> > moves on her opponents home intersections.
>
>
> .........
> X.......X
> X.......X
> X.......X
> X...*...X
> X.......X
> X.......X
> X.......X
> .........
>
> I'll work on the 9 squares marked Y:
>
> .........
> .........
> .........
> ....YYYY.
> ...Y*....
> ....YYYY.
> .........
> .........
> .........
>
> Disrupting this is surely possible, but I suspect it isn't as
> easy as you make it out to be.
What he said.
Mike Terry
news:1180829459....@o11g2000prd.googlegroups.com...
Right - this is the key question: how easy is it to disrupt the above
counter-strategy?
I suggested it seemed quite easy, and you're saying it isn't. Your only
example in your .PDF file is like the one Ed gave above, and it is *so*
trivial to disrupt this case that I wonder if you have seriously tried the
strategy? The above example (and variations) was the first thing I tried to
test my strategy before I posted, and it seemed to me I could easily disrupt
such attempts.
We could settle by trying say 10 games of 9x9 against each other, and if my
edge strategy wins each one with little thought that would show something?
(I'm not convinced my strategy will win, as I've not played much, but I'm
not sure how much you've actually played either! And if it turns out that
the game is indeed more complicated than I first thought, I'll be delighted.
:-)
Do you have a site where we could play online? Or maybe 10 simultaneous
games in an email exchange or something would be workable...
Regards,
Mike.
marks...@gmail.com
<news.dead.person.sto...@darjeeling.plus.com> wrote:
> > > > It seems to me that the first player has an easy win by simply playing all
> > > > moves on her opponents home intersections.
>
>and it is *so* trivial to disrupt this case that I wonder if you have seriously tried the
> strategy?
I was trying to demonstrate two things in my PDF file. First of all,
I'm answering the question, "Is it possible for Player 1 to win simply
by filling in all of Player 2's edge points?" I think we can all
agree that the abundantly obvious answer to that is no.
The second thing I was trying to demonstrate is that the obvious
counter-strategy to the edge-fill strategy is only slightly more
sophisticated than the edge-fill strategy. I think it's safe to
assume that a strategy to foil the strategy that foils the edge-fill
strategy would be more complex yet. Simply because the counter-
counter-counters start piling up, it starts to get a little
complicated. More of a guess than a fact though. There could be a
kink in my theory, but only time will tell.
>
> We could settle by trying say 10 games of 9x9 against each other, and if my
> edge strategy wins each one with little thought that would show something?
The only thing that would demonstrate is that I'm a lousy game
player. I'll concede that up front. Now on a 9x9 there's probably
going to be a first move advantage, even among beginners. I've only
played one complete game: a 9x9 against myself. But I've started a
couple of 19x19 games on superdupergames.org, one of which I'm quickly
losing, as anticipated.
> (I'm not convinced my strategy will win, as I've not played much, but I'm
> not sure how much you've actually played either! And if it turns out that
> the game is indeed more complicated than I first thought, I'll be delighted.
> :-)
Let's keep our fingers crossed. I only develop games that seem
architecturally interesting to me. It rarely happens that they
develop flaws. Somewhat more commonly, though still infrequently,
they turn out to be not fun to play. I don't really have any control
over that. If they work, and they seem pretty cool, I release them.
I look forward to playing you Mike, although I'm sure you're going to
kill me. I keep a running challenge open as (you guessed it)
MarkSteere.
Mark
Mike Terry
news:1180836421.7...@d30g2000prg.googlegroups.com...
Well you could try out the edge strategy yourself, and maybe you'll start
winning! :)
Obviously you don't play randomly on the opponents edge - you just "squint"
at the opponents stones and think "Is he starting to bunch up his stones to
present a focussed threat to force a channel to a particular point (or
general area) on his edge?". If so, play your stone at or near that
point... It seems to take *way* more moves to build up a channel threat
like this, compared to responding to block it. On the other hand, if the
opponent plays unfocussed moves dotted around the board, simply reply on the
opponents edge roughly parallel with the played stone (on the same side of
the board). (I.e. maintaining an overall "balanced" position.)
I tried out a couple of 19x19 games against myself, but if anything the size
seemed to *help* the edge strategy. It seems that playing further away from
the centre actually helps with the blocking strategy, as you have more time
to see a threat coming!
One thing I did notice - the most troubling counter-strategy I found
involved playing moves on the diagonals - these moves are connected, and can
end up threatening to direct a super-wide channel to one of the opponents
entire edges. If you've been playing too many moves on the opposite edge,
you could be caught out by this, although it doesn't seem to be a problem if
you keep a balance between your moves and your opponents as I described
above...
E.g.
. . . . . . . . .
O . . . . . . # O
O . . . . # . . .
. . . . # . . . .
O . . a * c . . .
. . . . b # . . O
. . . . . . # . .
O . . . . . . # O
. . . . . . . . .
O has carelessly played too many moves on the left, given the number of
moves # has made on the right. If # can get to play a and b, O won't be
able to block the edge fast enough. (But even at this point O can easily
win by temporarily abandoning the edge strategy to block at c...)
> > (I'm not convinced my strategy will win, as I've not played much, but
I'm
> > not sure how much you've actually played either! And if it turns out
that
> > the game is indeed more complicated than I first thought, I'll be
delighted.
> > :-)
>
> Let's keep our fingers crossed. I only develop games that seem
> architecturally interesting to me. It rarely happens that they
> develop flaws. Somewhat more commonly, though still infrequently,
> they turn out to be not fun to play. I don't really have any control
> over that. If they work, and they seem pretty cool, I release them.
Well, most of these sorts of games I never look into at all, so I fully
agree with you this one sounds interesting. It would be a shame if it turns
out to be unbalanced for some reason.
Hmmm, superdupergames.org? I've not come across that before - I'll have to
investigate it...!
Mike.
Mike Terry
news:1180800020.9...@a26g2000pre.googlegroups.com...
> Levee is a *very* unusual game played with a Go board and a set of
> black stones.
>
> http://www.marksteeregames.com/Levee_rules.pdf
>
Mark - I see that in diagram 3 in your rules you are showing an "illegal
move for player 2". This isn't a good diagram, as player 2 has already lost
in this position! I.e. the game has already ended, so presumably *all*
further moves are illegal :-)
> Mark
> marksteeregames.com
>
marks...@gmail.com
<news.dead.person.sto...@darjeeling.plus.com> wrote:
> Mark - I see that in diagram 3 in your rules you are showing an "illegal
> move for player 2". This isn't a good diagram, as player 2 has already lost
> in this position! I.e. the game has already ended, so presumably *all*
> further moves are illegal :-)
>
You're absolutely right. Thanks for the tip. I will fix that right
now.
I never saw you at SDG. Still interested in playing Levee?
Mark
Mike Terry
news:1180906812.4...@o11g2000prd.googlegroups.com...
Sure - I've been busy today, but I've just been to the site and registered
an account (darjeeling). It's bed time now, and I'm working tomorrow, but
I'll keep my eye out for you when I've got some free time...
Mike.
>
> Mark
>
marks...@gmail.com
<news.dead.person.sto...@darjeeling.plus.com> wrote:
> > > Mark - I see that in diagram 3 in your rules you are showing an "illegal
> > > move for player 2". This isn't a good diagram, as player 2 has already
> > >lost
I decided to remove that little "only one edge point per channel"
rule. It wasn't necessary to the game, and only added complication to
the rule sheet, the gameplay, and Levee programs. I don't think
anyone will even notice the change, much less be affected by it. If
anyone was affected, my apologies.
Thanks for calling this to my attention though. It led to a slight
improvement in the game.
Mark
Bertram Felgenhauer
> On Jun 2, 12:23 pm, "Mike Terry"
> <news.dead.person.sto...@darjeeling.plus.com> wrote:
>>
>> It seems to me that the first player has an easy win by simply playing all
>> moves on her opponents home intersections.
>>
> Boy that'd be embarrassing, wouldn't it? I hope I never release a
> game with such an obvious flaw.
I'm afraid it's really that easy.
> Please see:
> http://www.marksteeregames.com/Levee_Edge_Fill.pdf
Ok, let's discuss this for the 7x7 board. Assume that black only plays
on the left and right edge.
Note: For this article, the egdes don't include the four corners of the
board. So the left and right edge consist of 10 points in total.
The trouble is that White needs at least 7 moves to win on an edge
(namely, three moves to close the channel above, three moves to close
it below and one move to close one edge), while black needs only
5 moves to completely block that edge.
Here is a winning strategy for black:
a) on the first move, play on the left edge
b) in reply to white's move, if possible:
- if white moves on the center column, play on the right edge
(i.e. the edge opposite to the first move)
- if white plays on the center row, play on the opposite edge
(this is the key trick)
- otherwise, play on the edge on the same side of the board white
just played on.
c) if this is not possible, play on the other edge
d) otherwise, play any valid move.
Note that once both edges are completely filled, white can no longer
win, and, because draws are impossible, black will win eventually.
If black follows that strategy, white can not win. There are two
scenarios to consider:
1) white forms a channel on one edge without using black's stones on the
other edge. To do that, white must play at least two stones in the
center column (meaning that including the first move black has played
at least one stone on the edge the channel is on), one stone in the
center row, on the side opposite of the channel, and four more stones
on the side the channel is on. By the strategy above, black has
already played at least 1+1+4-1 = 5 stones on the edge the channel is
on at the time white would play the last of these stones, blocking it
completely.
2) white forms a channel by blocking it above, using some of black's
stones on one edge and blocking it below. So he has to play at
least 5 stones below and 5 above the channel, but before that happens,
black has already completed 10 moves and filled both edges completely,
preventing a win for white.
So white can not win either way.
The argument carries over to larger boards.
regards,
Bertram
Aaron
sort of pie rule not fix this? After the first, second, or third move
for instance?
Aaron Dalton
http://superdupergames.org
Christopher Night
Can anyone who remembers them post a summary of the rules of Levee? This
seems like an interesting discussion, but I got to it too late to
understand it, because they've apparently been withdrawn.
-Christopher
Ed Murphy
.AAAAA.
B.....B
B.....B
B..C..B
B.....B
B.....B
.AAAAA.
The board starts out empty. Squares with game significance
are labeled.
Players A and B alternate placing stones on empty points. Adjacent
stones (orthogonal or diagonal) are linked by walls. Neither player
may play a stone on C, or play a stone that forms a closed loop of
walls around C. A player loses if it becomes impossible to travel
from C to one of that player's edge squares without crossing a wall.