The site has moved to a new server, and there are now some issues to fix. Please report anything needing fixing with a comment to the homepage.



The Chess Variant Pages




[ Help | Earliest Comments | Latest Comments ]
[ List All Subjects of Discussion | Create New Subject of Discussion ]
[ List Latest Comments Only For Pages | Games | Rated Pages | Rated Games | Subjects of Discussion ]

Comments/Ratings for a Single Item

Later Reverse Order Earlier
Outer Space Chess. Space-themed game with hyperspace and regular space boards. (5x8x2, Cells: 43) [All Comments] [Add Comment or Rating]
Nicholas Kuschinski wrote on 2003-07-15 UTC
Now THERE is a comment I agree almost wholeheartedly with. I have long ago
given up on this game anyways, finding its flaws too great to handle. I
also agree about the math thing. I just really don't have access to any
other reasonable sort of tool (AI programming being just too damn time
consuming). It is because of this that I have pretty much abandoned game
design alltogether.

As for Duinho's comment, which you claim to be a perfect example of a way
to solve such problems: Allow me to disagree.

It doesn't solve anything. A player may still capture a king and force a
draw, that this happens to make him less likely to win . . . well, if you
really want to draw, you don't really care if you win or not. The game
still has a foolproof way of reviving a losing position and bringing it up
to a draw. It is FAR from being an actual solution to the problem.

Carlos Martín-F. wrote on 2003-07-15 UTC

Thank you for taking the time to read my comment.

I'm not discarding math as a tool to detect some problems that may appear in a game, but I wouldn't use pure mathematic to develop the game, because it may result in something that is 'mathematically correct' but not playable. I could, for example, use mathematical analysis, once I have finished my game, to check if it has errors of some kind. As for the rules that are 'forced' into a game, in order to 'fix' some inconvenience derived from another previous rule, maybe the problem is in that previous rule from which the inconvenience comes. You might find that one certain rule that you like very much provides more problem than benefit. Sometimes you can find some elegant way to elude the problem (and Fergus' example is simply perfect for illustrating this), but, there can be some cases where no appropiate solution is to be found, and then you have to remove your beloved rule and start from the beginning (this is painful but quite common when you begin designing games).

Not all the rules are valid for a game, and the fact that a rule is beautiful doesn't imply that it is suitable. I wouldn't prefer allowing players to force a draw by leaving their Galaxy on Hyperspace's black hole, instead I would check the rules about Hyperspace. Maybe some of the main rules of the game do not make much sense. Perhaps the error is in the fact that leaving your Galaxy in Hyperspace's black hole (which is such an easy thing to do) gives you such an enormous advantage. Somehow, it seems like you invented Hyperspace to let pieces be trapped on it, and later discovered that having trapped pieces was too great a disadvantage for one side and tried to fix it with another rule that overcomplicates the game and makes you have to introduce 3 more criteria to define a piece's impossibility to leave Hyperspace:

1. If a piece is not on the Black Hole's square, but is in hyperspace, and three or more moves have passed the next move of the player must be to move his piece onto the Black Hole, so long as the Black Hole square is not occupied.
2. If a piece has been in hyperspace for four moves or more, and is now on the Black Hole square in hyperspace, the next move of the player after the piece has entered hyperspace must be to move it to back to ordinary space, so long as a legal move exists that will let him do so.
3. If a piece has not made one of the above moves because it was not legal (ie: the Black Hole was occupied, or there was no legal move that permitted the piece to go back to regular space) the piece must make such a move as soon as it is legal.

This set of exceptions, and exceptions to the exceptions, and 'so long as'... is too much for me :)

PS: Please don't think that I absolutely dislike everything about your game. If you find my previous comments too harsh, you might feel glad to learn that there are certain aspects about Outer Space Chess which I do like sincerely, for example the way the 'Galaxy-Nebula system' (as you call it) works.

I appreciate the idea of a royal piece being a strong piece, and find quite interesting the way you achieve this by stating that only one piece may capture the Nebula, and then state that that piece (the Galaxy) cannot be captured (capturing your opponent's Galaxy would at least guarantee you a draw), but then, if you could find another way to reduce the Nebula's mobility, you wouldn't have needed to add the rules about:

-the Nebula not being able to move on 2 consecutive turns
-the Nebula not being able to move through 'attacked' squares, and
-the permission to capture your own pieces.

I'm just trying to provide a 'constructive review' with which, of course, you have the right to disagree.


Fergus Duniho wrote on 2003-07-01 UTC
I agree with you that there is nothing wrong with fixing problems with a
game. But I do appreciate the point Carlos is making, because there are
more and less elegant ways of fixing the same problem. Let me describe a
problem I fixed in one of my games that is like the one you describe with
the Galaxy. 

In Three-Player Hex Shogi, players may capture Kings and hold them in
hand, and the goal is to make all three Kings your own. The problem
here is that a player might force a draw by holding one King in hand
indefinitely. There were various ways to fix this problem. For example, I
could have said that no King may be held in hand longer than five turns.
But this would be an inelegant solution that overcomplicated the game.
Instead, I chose to provide an incentive for dropping a King back on the
board. So I added the rule that the King is the only piece a player may
drop when he has one in hand. Since dropping other pieces is normally
critical to doing well in the game, holding a King indefinitely would
impair a player's chances of winning.

Nicholas Kuschinski wrote on 2003-07-01 UTC
A lucid review if I have ever seen one. Living alone in a small college dorm room is an impediment to playtesting. For the most part, I share your point of view. Nonetheless, There are one or two points worth defending: <br> 1) Math is a good thing! Playtesting is admittedly far better, but the human mind is often nearsighted, and misses some things that pop up immediately if you do some math. I'm all about playtesting, but you can't just discard the only tool that this poor college student has at his disposal. <br> 2) If you READ the strategy and tactics section, you will realize that I have provided only extremely broad strokes. This is what I WANT my game to be, and I admit it. I didn't dare go into specifics, and at that, merely gave a quick overview of what appears to be obvious. (And about the rules making the strategy . . . How the heck would you design the rules then?) <br> 3) What exactly is it that you have against fixing problems with a game? If there is something wrong with it, you should fix it. Granted, it might make the rules more inelegant, but its better than having a game that has a gaping hole in it. Ex: You complain about my rule that keeps pieces in hyperspace, but if one were to move their galaxy onto the black whole before his opponent does, and leave it there, forcing a draw. U telling me you would rather have THIS problem, as opposed to an inelegant rule?. <br> <br> Other than this, I can say that your comment is about as lucid as they come. Since the post, it has undergone some playtesting, and other problems have shown their faces. Most notably, a disequilibrium between the sides that is difficult to attribute to any particular rule or group thereof, and therefore, really hard to get rid of(read previous comments). Surprisingly, however, I got rid of most of the nooks and crannies before playtesting began, and I was right about almost all of what I wrote in my strategy and tactics section. <br> Thank you, nonetheless, for a sincere, straightforward, and lucid response.

Carlos Martín-F. wrote on 2003-06-30 UTC
<font size='2'><p> After having a look at your game, I'm afraid that we have very different points of view about what a chess variant game should be (still, I'm not saying this is a problem, the world would be rather boring if we all had the same opinions!). <br><br> You seem to develop your game from a strictly mathematical basis. As a result, I sincerely wonder if the game is playable at all. I was disappointed at reading things such as <i>'The game has not yet been playtested, so most of my discussion will be coming either from rigorous analysis, or from the strategies that I tried to embed into the rules.'</i> I think games should be developed through playtesting, rather than from abstract mathematical theories, and I firmly believe that it is not the strategies that make the rules, but vice versa (although you have the right not to agree with me). I am absolutely surprised that you even dare write a 'Strategy and Tactics' section, when <i>'Not actually having played this game, some of this might be wrong'</i>. <br><br> I also do not see what the 'impediment' is about 43 being a prime number, unless the only shape you can think for a board is the rectangular one. And as for the notation of the cells, I don't think there's a need to add one number to identify boards when you only have two of them: let's call cells in ordinary space from <b>a1</b> to <b>e8</b> and cells in hyperspace simply <b>a</b>, <b>b</b> and <b>c</b>. <br><br> Finally, a word about simplicity. When I first tried designing Chess variants, I had the temptation of including a lot of special rules, and have lots of different pieces with complicated moves... then I learned that a game is not a recopilation of rules, and that a complex game is not necessarily a good game. Actually, when things are too complex, it is harder to keep all the rules in mind, and it results in a lack of playability. Some of my favourite Chess Variants are as simple as <b>Mono-dimensional Chess</b> by Luiz Carlos Campos, whose board has only ten squares. <br><br> This refers to some of the rules you include, such as <i>'No piece may remain in hyperspace for more than five consecutive turns unless it is impossible for them to leave'</i> (which makes you have to keep track of how many turns a piece has been in hyperspace), <i>'Nebulae may not move twice in a row. If a player is forced to move his Nebula in two consecutive turns, he must pass on the second'</i>... all this seems as if every time you find a flaw in your game, you fix a new rule to 'mend' the error somehow (an example of this would be tha last rule I mention, probably invented after suddenly finding that the Nebula was too mobile). I don't think this is an appropriate way of writing rules for a game. <br><br> (I hope I'm not offending anyone, since it is far from being my intention!) </p></font>

Nicholas Kuschinski wrote on 2003-04-14 UTC
I'm sorry if people feel offended by my dislike of this software. Please don't take it personally. Yes, Zillions does use a bit of programming to work, but that's not really what I was referring to. You can't use math do develop a game, but you can use it to test for various situations in ways that most other systems fall short. It is true that it often misses all sorts of problems, and that it depends on what form of assumptions you make, but it will catch a significant portion of any problems a game has far more eficciently than any other approach, you just have to be clever. Playtesting with humans is far more controlled than with any sort of AI system: the advantage is that of being able to stress anything that pops up that may look like a problem. You can deliberately play in such a way that your rules are almost likely to break down. Its not really just trial and error. I do not condemn ZoG entirely, and I am not in any way trying to cause any sort of disturbance. I simply don't want to hear comments saying 'Your game should be changed in such and such a way because that would make it easier to implement in ZoG'. No matter how much you support or love ZoG, the goal of game design should never be to make ZoG happy, but to make a good game. I refuse to make any modifications to the rules merely on that basis, and although I have probably made a rude exclamation or two when it wasn't actually called for, I think that all serious people here will be able to sympathize. If it seems to anyone like I have insulted you, let me assure you that I did not mean to. Could we please stop talking about ZoG now and move on to something more interesting? (like chess variants?)

Fergus Duniho wrote on 2003-04-13 UTC
You can't design a game with Zillions of Games without doing computer programming. So I presume you are misinformed about how this program is used. I'm curious about your claim that one can use math to develop Chess variants. Perhaps you would be kind enough to explain how this is done. In the meantime, let me explain how Zillions of Games is used for developing games. The first step is to describe the rules in a Zillions Rules File using a special macro language for describing rules. The next step is to provide graphics. One may use the graphics that come with Zillions, graphics created by oneself, or graphics created by others. It enables playtesting in three ways. You may play against the computer, you may have the computer play against itself, or you may play against a human opponent online. Having the computer play both sides is useful for testing for defects that are most apparent between equally strong opponents, such as testing whether a game is drawish or unbalanced. Zillions is also good for getting nice diagrams to illustrate a web page on the rules.

Peter Aronson wrote on 2003-04-13 UTC
Nicholas, <p> Your tone could <em>really</em> use some work. The CVP is generally a polite, <strong>adult</strong> discussion environment, and we like it that way. <p> As for Zillions, have you examined it in any detail? I am professional software designer and programmer with a degree and 27+ years of programming experiance, and <em>I</em> am really impressed by what Zillions is and does. The AI is not enormously strong, but it doesn't make assumptions, and the macro language is very, very impressive (having implemented several macro languages myself, I know what is involved). <p> Implementing a game with Zillions isn't as good as playtesting with a human, but doing both is better still. Game designers will often have blind spots about their own designs, and their playtest opponents may or may not spot those blind spots. I have seen Zillions, which after all does not play like a human, find these blind spots very quickly after human to human playtesting (often fairly considerable amounts of it) have failed to find them. In playtesting, having players with different play styles is invaluable. Zillions provides a playtester with a very different style. <p> As for using math to analyze games, well, I have seen far more failures than successes at that. It is all a matter of assumptions, and if you make the wrong ones, well, GIGO. <p> I have also found that coding a game for Zillions (and by the way, writing a ZRF <em>is</em> coding) to be one of the best ways I've ever found to analyze a set of rules, since it requires you to resolve all ambiguities, and determine in detail all of the piece interactions.

Nicholas Kuschinski wrote on 2003-04-13 UTC
Only for those people who refuse to do any of the following: math, computer programming or playtesting (with humans). All of them are certain to give better results and provide reliable answers. I usually choose the first of these. I believe that the best is probably the last. Programming computerized applications and trying to create an AI has (in my experience) served mostly for games in their earlier stages. Zillions might be good for those people who are to dumb to do any of these, but I can't really see any other reason to resort to it.

Fergus Duniho wrote on 2003-04-13 UTC
Zillions of Games is not cheesy software. It is an invaluable tool for designing, testing, and sharing Chess variants.

Nicholas Kuschinski wrote on 2003-04-13 UTC
Sorry. I was angry at the time, and felt like I had to vent it somehow. I
certainly did not mean to seem insulting to you (although I refuse to say
the same about cheesy software), and I assure you that I read your comment
in full. I will now respond to comments about the game: I feel that a
completely disinhibited nebula isn't much fun as a royal piece. Its far
too free and keeps the game from looking (intuitively at least) like it is
headed towards a conclusion. I feel it is necessary to keep it on a leash:
It increases the tactical element rediculously, as it permits interesting
combinations to be used to hold it in place, and makes defensive play much
more urgent. I had intended to create an extremely defensive game that had
a back-heavy tactical weight: Keeping the nebula surrounded. You could
certainly try this as a variant, if it suits you better, but this brings
up a sharp contrast with my original vision of the game.
<br>
<br>
As to Antoine's comment: It seems black is behaving more agressively than
it should. The way to play this game (at least I hope so, I haven't had
the time to look into this more thoroughly) is to keep everything tightly
knit and well defended in the back. Breaking up a good defensive structure
is probably more dangerous here than in any other chess variant I have
ever seen. I have specifically chosen the lineup so that a good defensive
structure can be obtained before there is any real danger, but it must be
done quickly (of course, if your opponent is making dumb moves, it is
still possible to attack quickly and take advantage of the situation).
This analysis is rather disturbing however, and I fear that even with the
new rule that permits black to keep up with white in hyperspace the sides
may still be unequal. I am fairly certain that this game is still flawed
(most games are) but I was hoping it was less so than before. At this
stage I fear there is little I can do about it, but I may yet come up with
some burst of brilliant insight in the near future, it is hard to tell.

Michael Nelson wrote on 2003-04-12 UTC
Nicholas,

Could you please express your (often quite accurate) comments in a less
insulting fashion?  If you had actually read my comment instead of just
observing the word 'Zillions' and dismissing my idea (and me) out of
hand, you would have seen that I am equally concerned with playability by
humans.  If Zillions can't be programmed to play something legally (as
opposed to playing it well), generally there are playablity issues for
humans as well.  I can't visualize the Nebula rules on the board as you
have them now--I'm sure I am not unique in this respect.  Even if you
think my observation is entirely erroneous, you could express yourself in
less abrasive, attacking language: why don't you?

Nicholas Kuschinski wrote on 2003-04-12 UTC
To Antoine Fourrier and Michael Nelson (both): Zillions can kiss my behind! I, quite frankly, don't really care if my game can be made more easily playable by some sort of weird computerized interface sold by a trashy software company. This was NOT the goal. I much prefer your coments on the game (which have been extremely valuable) rather than comments on how to port it to Zillions.

Antoine Fourrière wrote on 2003-04-12 UTC
1)'Nebulae are permitted to land ON squares controlled by other pieces,
just not move THROUGH them'. That is what I understood when I playtested
your game. Since all the neighbor squares of a Nebula are controlled by
friendly pieces, it follows that the Nebula moves every other time as a
Wazir. (If they weren't controlled, the Galaxy would soon capture the
Nebula.)
2)I indeed tried to copy the British Chess macro.
But since the Nebula is immune from capture,
(verify (and empty? not-attacked? not-defended?)) doesn't work,
and I had to write a verify for each square on the Nebula's path, with
the enemy Nebula controlling the squares as a Rook.
3)Zillions often loses by playing 1.. Star Cluster 1c5.
For instance:
1. Galaxy 2c - 2b = Galaxy2
1... StarCluster 1c6 - 1c5 ; Galaxy2 on 2a
2. Planet 1d1 - 1e3; Galaxy3 on 2b
2... Planet 1b8 - 1c6; Galaxy3 on 2a
3. Galaxy3 2b - 1d6 = Galaxy; Galaxy on 1d6
3... Planet 1d8 - 1e6
4. Planet 1e3 - 1d5
4... Galaxy 2a - 2b
5. Planet 1d5 x 1c7
and a slow win for White.

Michael Nelson wrote on 2003-04-12 UTC
With regard to the Nebula movement limitation, I believe it would be better if the enemy Nebula's move were considered without the limitation (as if it were a Rook). This non-recursive rule simplifies the Zillions implementation and human players' thinking. A good example is found in the check rules of <a href=http://www.chessvariants.com/large.dir/british.html>British Chess</a>.

Nicholas Kuschinski wrote on 2003-04-11 UTC
Nebulae are permitted to land ON squares controlled by other pieces, just not move THROUGH them. This assures that no matter what happens they have at least the mobility of star clusters. Thanks for playtesting my game! You have provided valuable information, but it seems that whatever you have done, you overlooked this important rule about the nebula. Also, Hyperspace wasn't meant to house galaxies, so much as to swallow spaceships. Since spaceships can go neither onto nor through a black hole without getting sucked in, it should be quite difficult to move spaceships around without going through hyperspace, which is why hyperspace is there. It may be the case that it is usually a better idea to avoid black holes altogether, but I feel that the existence of hyperspace should certainly affect gameplay anyways. I can think of a couple situations in which it could at the very least be used to permit planets to move onto black holes, thus gaining some mobility in special circumstances (if nothing else). I have no problem with pieces hogging hyperspace, so long as it can't be used to force a draw, I don't see any problem with this. As for galaxies blocking spaceships, that was the point. I'm glad it worked, as it took me a long time to figure out how to restrict a piece with rifle capture. I have stated in my presentation of the rules that I didn't think planets were necessarily more valuable than star clusters at all. On the whole, your results please me. This is what was supposed to happen! Thank you for the time you have spent working on this game. I owe you a great debt. Rock on!

Antoine Fourrière wrote on 2003-04-11 UTC
If my understanding of your rules and my implementation of them in a
Zillions file is correct -- anyway, it prevents Nebulae from crossing
controlled squares, including the squares which are controlled by the
other Nebula, but I haven't been able to take into account that same
limitation for that other Nebula, to limit the Nebula-controlled-square
limitation -- I reach the following conclusions:
1)The Nebula is the least mobile piece. Indeed, I have never seen a Nebula
move more than one square: there is always a friendly piece which prevents
the Nebula from crossing anything. Without a pack of friendly pieces, the
Nebula would be immediately captured.
2)The Galaxy is quite mobile. It is also very useful in protecting a
diagonal against a Spaceship. Maybe slightly too useful.
3)2a and 2c never attract a Galaxy or Spaceship, and 2b, hardly ever. Your
new rule doesn't prevent a Galaxy or Spaceship to hog the hyperspace on
2b, but it doesn't seem to matter. The Galaxy and Spaceship are simply
too valuable in the regular space.
4)It seems a Planet is not significantly stronger than a Star Cluster.
5)Your game is playable as it is, and even enjoyable, which is not so easy
when there is rifle capture, but only in the regular space.
Maybe your black hole on 2b should be replaced by a gigantic black hole in
2a-2b-2c, which could only be entered on 2a and exited on 2c or the
reverse, or maybe Galaxies should be able to send a neighbor  -- any
neighbor or just a Galaxy or Spaceship? -- on 2a or 2c to the price of
going themselves on 2b, or maybe some eleventh piece should be added to
each army, I don't know, these are just ideas that I throw out, I
haven't really thought about it.

Nicholas Kuschinski wrote on 2003-04-10 UTC
Damn!! You are absolutely right! I can't actually apply the rule only to pieces which enter hyperspace either, because in that case whit could play galaxy to 2b as his first move, and never move the galaxy again, forcing a draw . . . Damn!! OK, I haven't completely analysed this solution yet, but how does this sound? A piece has to go back in five moves so long as there is another piece in hyperspace. A piece that is in hyperspace ALONE is not forced to leave. As soon as a piece no longer has any neighbors in hyperspace, it is no longer foced to leave. This would get rid of the easily forced draw just as well as the other set of rules, and wouldn't let white force so many moves on black. Upon testing the old rule, I kept looking for ways to force pieces into hyperspace in such a way that you could gain some ludicrous advantage, and nothing showed up, but it never occured to me to worry about pieces that are already there to begin with!!! That was dumb! The new version of the rule would also make 2a and 2c more attractive, as any piece on 2b is forced, by this rule, to exit hyperspace FIRST. Thanks for poitning out the mistake. Unless you can find something wrong with the newer version of the rules, I will declare this addition to be official. HOW THE HECK DID I MISS THAT???!!!

Antoine Fourrière wrote on 2003-04-09 UTC
Oddly, my comment on your game quite matches yours on mine.
There is a lot of imaginative stuff.
Otherwise, I wouldn't have spent the day to devise an incomplete zrf to
implement your rules.

But I'm afraid that White has an extraordinary advantage, as suggests:
1. Galaxy 2c - 2b = Galaxy2
1... Planet 1d8 - 1e6; Galaxy2 on 2a
2. StarCluster 1c3 - 1c4; Galaxy3 on 2b
2... Planet 1e6 - 1d4; Galaxy3 on 2a
3. Spaceship 1c2 - 1c3; Galaxy4 on 2b
3... Planet 1d4 - 1e6; Galaxy4 on 2a
4. Galaxy4 2b - 1d6 = Galaxy
4... Galaxy4 2a - 2b = Galaxy5 (forced)
5. Galaxy 1d6 - 1e7
5... Galaxy5 2b - 1d3 = Galaxy (forced again)
6. Galaxy 1e7 - 1d8
6... Nebula 1c8 x 1b8
7. Galaxy 1d8 - 1c8
7... Nebula 1b8 x 1a8 (regardless of the twice-in-a-row limitation)
8. Galaxy 1c8 - 1b8#

Well, it should be OK if you limit the five-turn rule to pieces which
enter the Hyperspace, but having said that, I can understand why a player
would send a Galaxy or at least a Spaceship on 2b, but I wonder why he
should bury them on 2a or 2c.

19 comments displayed

Later Reverse Order Earlier

Permalink to the exact comments currently displayed.