Author |
Topic: Essay by Christian Freeling on inventing games (Read 539981 times) |
|
MarkSteere
Forum Guru
Finite games rule
Gender:
Posts: 289
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #150 on: May 26th, 2010, 9:36am » |
|
on May 26th, 2010, 9:29am, christianF wrote: I described the problem in a nutshell, |
| And that would be the precise nutshell to which I was referring. It's growing like Pinocchio's nose
|
« Last Edit: May 26th, 2010, 9:37am by MarkSteere » |
IP Logged |
|
|
|
christianF
Forum Moderator Forum Guru
Arimaa player #4019
Gender:
Posts: 804
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #151 on: May 26th, 2010, 11:09am » |
|
on May 26th, 2010, 9:36am, MarkSteere wrote:It's growing like Pinocchio's nose |
| Hi Mark, glad you're enjoying the proceedings so much. Seems like I didn't quite make it with my 'predictable behaviour' claims. I clearly misjudged the catenaccio approach . Are there extenuating circumstances? At the core of the game there's a simple 'king's move / knight's move' scenario, but the interaction is very complex. So it was a difficult monkey to judge to begin with. Predicting the behaviour of say Grand Chess or Dameo is simpler because there are earlier games one can rely on, or your own Oust, which is a stroke of genius, or Cage, where the 'entrance into an endgame' aspect is immediately apparent (wasn't there a rule change in Cage? - gridlock if I remember correctly? ). At the same time I don't think the problem is any bigger than implementing both criteria and calling the first one 'obstruction' and the second one 'shielding'. They're not illegal, both are red card offences, and may cost you a piece just like any other bad move. This soon learns you to keep an eye on them. They may even be committed as an actual sacrifice if a player can manage a threat to score despite any piece being sent off.
|
« Last Edit: May 26th, 2010, 11:15am by christianF » |
IP Logged |
|
|
|
MarkSteere
Forum Guru
Finite games rule
Gender:
Posts: 289
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #152 on: May 26th, 2010, 1:12pm » |
|
on May 26th, 2010, 11:09am, christianF wrote: Hi Mark, glad you're enjoying the proceedings so much. |
| Sorry, not to be celebrating your dilemma. Just goading you a little. I've committed at least my share of folly in the game design business, retracting flawed games on occasion. on May 26th, 2010, 11:09am, christianF wrote: or your own Oust, which is a stroke of genius, or Cage, where the 'entrance into an endgame' aspect is immediately apparent (wasn't there a rule change in Cage? - gridlock if I remember correctly? ). |
| Thanks re Oust and yes re Cage gridlock, my folly In a contrived position, a giant diamond of like colored checkers around the center, the checkers couldn't move inward. Fortunately a tiny fix was available to straighten out the game. There isn't always. The problem was kind of a "typo" in the game's geometry. Cage was trying to be a valid game, but I wasn't letting it at first.
|
|
IP Logged |
|
|
|
christianF
Forum Moderator Forum Guru
Arimaa player #4019
Gender:
Posts: 804
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #153 on: May 27th, 2010, 6:37am » |
|
on May 26th, 2010, 1:12pm, MarkSteere wrote: Sorry, not to be celebrating your dilemma. |
| Thanks, I'm in sheer desperation. on May 26th, 2010, 1:12pm, MarkSteere wrote:Cage was trying to be a valid game, but I wasn't letting it at first. |
| We seem to have in common that you perceive a game as a spirit whose accomodation is the inventor's goal . HanniBall has the spirit of a soccer game and should first and foremost be fun. A single dominant strategy, however successful, has no place in that equation. Another nights sleep brought no single gereric rule so I'll go for the double. This means that the current obstruction rule remains in place and the new one, governing piece density and distribution around the ball, is added under the name shielding. Obstruction If a player on his turn finds a position in which he has at least one piece other than a Keeper, and not one of his pieces can reach the ball in any number of moves, then the opponent has committed obstruction and the player to move may remove one of the opponent's pieces from the board as his first move. This may or may not be a blocking piece. If the obstruction is still in place at the beginning of the blocking player's next turn, it is up to him to undo the obstruction, or risk having yet another piece removed. Shielding concerns only the 9x13 fieldsquares, not the goals. Definition: the 'ballsquare' is the square where the ball is at the beginning of any given turn. A piece's or a square's distance to the ballsquare is measured as the smallest number of king's moves to get from there to the ball. Squares with equal distance to the ballsquare lie in 'rings' around it. A ring is an odd sized 'square of squares'. The first one (R1) is 3x3, the second one (R2) 5x5, then (R3) 7x7 and so on. Rings may be truncated, the remainder being outside the board. Shielding If a player on his turn finds a position in which on any particular ring the opponent's pieces occupy half the number of squares of that ring or more then the opponent has committed shielding and the player to move may remove one of the opponent's pieces from the board as his first move. This may or may not be a shielding piece. If the shielding is still in place at the beginning of the shielding player's next turn, it is up to him to undo it or risk having yet another piece removed. Very similar to obstruction. The penalty may or may not be given. It would be silly, for instance, to send a piece off at the cost of a move, when a player has a win in four moves. In terms of visualizing, obstruction is easy to spot as long as there are still Lions and Elephants in the game: only a complete seal off, orthogonally and diagonally, constitutes obstruction. In the (probably rare) event of a player having only Horses left, the opponent may cause obstruction with an enclosure that is open to a king's move, but inaccessible for Horses. The systems (Zillions, iGGC) signal such cases of obstruction. Concerning shielding, please note that I've finetuned it by changing the maximum number of pieces per ring from "not more than half" to "less than half". This reduces the maximum number of like colored pieces on rings with an even number of squares with one piece, while not affacting rings with an odd number of squares. To understand the implications, let's see how piece density and distribution is affected in R1, R2 and R3, going from the field to the edges and corners: * | R1 #sq (#pmax) | R2 #sq (#pmax) | R3 #sq (#pmax) | d5 | 8 (3) | 16 (7) | 24 (11) | c5 | 8 (3) | 16 (7) | 17 (8) | c4 | 8 (3) | 16 (7) | 11 (5) | b5 | 8 (3) | 11 (5) | 15 (7) | b4 | 8 (3) | 11 (5) | 10 (4) | b3 | 8 (3) | 7 (3) | 9 (4) | a5 | 5 (2) | 9 (4) | 13 (6) | a4 | 5 (2) | 9 (4) | 9 (4) | a3 | 5 (2) | 6 (2) | 8 (3) | a2 | 3 (1) | 5 (2) | 7 (3) | In actual play shielding in the center hardly amounts to more than checking the first ring. It's nearer to the edge and corner that a more thorough check is required. To make visualisation easier the first two rings around the ball (and therewith implicitly the third) will be highlighted in lighter shades inward towards the ballsquare. Where the rule against shielding does not prevent obstruction, thanks Gatsby for pointing that out, is guarantees a structure around the ball in which at least half the number of squares from any distances are available for invasion. Towards the edges and the corners, the number of like colored pieces allowed around the ball shrinks considerably (note: though mainly a problem of the player in possesion of the ball, the opponent may also commit shielding!). And catenaccio needs, if nothing else, a ball that's near the edge and sufficient 'piece density' around it, to proceed. So rather than facing a dilemma, I think the rule against shielding effectively puts an end to the one problem HanniBall turned out to have after its launch at iGGC.
|
« Last Edit: May 27th, 2010, 1:30pm by christianF » |
IP Logged |
|
|
|
christianF
Forum Moderator Forum Guru
Arimaa player #4019
Gender:
Posts: 804
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #154 on: May 28th, 2010, 8:37am » |
|
The rule against shielding will be implemented soon at both iGGC and Zillions. Highlighting the rings may take a bit longer because it may require a change in the interface. The alternative can be to make the rings part of the ball. Moving the ball would then imply moving the rings, and the highlighted squares would cover other pieces, requiring a rewrite of the whole position. To get around this, only the lines between the squares could be highlighted. The ball would then look something like this: And in the initial position it would look like this: The problem is that the rings will go over the edge of the board. To solve this, Ed will employ only two horizontal and two vertical lines ... ... (using one pair for each shade of yellow) and project the required ringlines with them for every possible square. Bit of work, but easier than changing the interface. Predicting behaviour This is the epilogue I added to a late arrival: Playtesting for a week or two at iGGameCenter revealed that the game's tactics satisfy its spirit. However, a not anticipated problem emerged, in terms of strategy. Arty Sandler was the first to formulate it: "Get the ball (black can get to it first), bring it to the left or right backfield and build a 'narrow passage' along the b- or h-column where you keep the ball save from invasion by a knight's move. To get in, the opponent would need a Lion or an Elephant, and a lone invader runs the risk of being captured. Now here's the puzzle: move the whole narrow passage towards the opponent's side, taking the ball along, till you're close enough to the opponent's goal to make a break for it with a Lion and the ball. That's it in a nutshell. It's been coined catenaccio, and though it revealed no inconsistency in the rules, it wasn't the way the game wants to be played. It clearly needed a rule to limit the number of pieces and their distribution around the ball. Thus the rule against shielding emerged. The point being that I did 'feel' the game's behaviour the way it now has become apparent, in it's immediate movements and interaction with itself, but missed the emergence of a strategy that clearly is not 'in the game's spirit'. I think with the rule against shielding the rules have now 'accomodated the game's spirit' adequately. As an afterthought, note that shielding can be used in a tactical way. A player can leave the ball in a position where not he, but his opponent commits shielding. Of course shielding is only considered at the beginning of any given turn, and when a player finds that his own pieces commit shielding at the start of his turn, it is up to him to undo it in the course of his own moves.
|
« Last Edit: May 30th, 2010, 2:37am by christianF » |
IP Logged |
|
|
|
MarkSteere
Forum Guru
Finite games rule
Gender:
Posts: 289
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #155 on: May 29th, 2010, 9:00am » |
|
I'm vicariously relishing your patient, understanding audience
|
|
IP Logged |
|
|
|
MarkSteere
Forum Guru
Finite games rule
Gender:
Posts: 289
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #156 on: May 29th, 2010, 9:09am » |
|
My ashes would be blowing out of a cold charcoal pit by now. lol
|
|
IP Logged |
|
|
|
christianF
Forum Moderator Forum Guru
Arimaa player #4019
Gender:
Posts: 804
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #157 on: May 29th, 2010, 9:49am » |
|
on May 29th, 2010, 9:09am, MarkSteere wrote:My ashes would be blowing out of a cold charcoal pit by now. lol |
| Mine are not, on the contrary. I have enjoyed many of the subject related comments here and I think the 'audience' has shown a fair share of patience with me, as well as providing some much appreciated help in shaping HanniBall's rules .
|
|
IP Logged |
|
|
|
christianF
Forum Moderator Forum Guru
Arimaa player #4019
Gender:
Posts: 804
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #158 on: May 30th, 2010, 1:54pm » |
|
Controversy rules The chat at iGGC suffered overload, so at least HanniBall is capable of causing controversy. Several alternatives were suggested and I was subtly remembered that my previous attempt at a fix failed miserably. Not surprisingly so, because it didn't address the problem, just shifted the game in slightly less positional surroundings. I had jumped at a solution just like some of the chatters today, and I jumped back rather quickly. I believe the rule against shielding does precisely what it has to do: it regulates both density and distribution of like colored pieces around the ball in a generic way. Yet Arty's suggestion to simply disallow three like colored pieces in a straight and unbroken line might work too. Or it might not, I haven't had time to let it sink in yet. The rule against shielding doen't prevent obstuction, but it might prevent a special 'undetectable' kind of obstruction Arty came up with: The system wouldn't detect 'obstruction' here because there's a black horse in the ranks, and if it would move, the position is open king's move wise. But the horse can't leave its position. The readers here are evidently good puzzlers, so here's a puzzle: Can a similar indirect obstruction be constucted with the shielding rule in place? In the example white is two pieces over the limit on the first ring. Arty had another interesting example: it is a shielding situation on R8 and it certainly doesn't look like shielding: Here it is important to realize that shielding and obstruction aren't forbidden: they're red card offences, so black could lose a piece here. No big deal, if your pieces are in this position you're probably so bad at the game that you're quite used to losing pieces. I don't really see the probem arising, and if it does I don't really see the problem, period. Generally: the difference between disallowing something and making it a 'red card' (that is: allowed but punishable) is that in the latter case the offence can be made on purpose. Like not minding the punishment if it gets you two pieces in a scoring position. And the punishment in turn is an option: like not capturing a piece at the cost of a move, because you have a win in four moves. So they're part of the possible tactics.
|
« Last Edit: May 30th, 2010, 2:46pm by christianF » |
IP Logged |
|
|
|
MarkSteere
Forum Guru
Finite games rule
Gender:
Posts: 289
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #159 on: May 30th, 2010, 3:05pm » |
|
on May 30th, 2010, 1:54pm, christianF wrote:Generally: the difference between disallowing something and making it a 'red card' (that is: allowed but punishable) is that in the latter case the offence can be made on purpose. Like not minding the punishment if it gets you two pieces in a scoring position. And the punishment in turn is an option: like not capturing a piece at the cost of a move, because you have a win in four moves. |
| Your nutshell is growing, Pinocchio
|
|
IP Logged |
|
|
|
Arty
Forum Full Member
Arimaa player #3423
Gender:
Posts: 16
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #160 on: May 30th, 2010, 7:41pm » |
|
Quote:Yet Arty's suggestion to simply disallow three like colored pieces in a straight and unbroken line might work too |
| Well, it's not about "straight and unbroken line" but any group of orthogonally adjacent pieces of the same color. So "L"-shaped group of 4 same-colored pieces will be punished too. I also got an idea to include the Ball into the counting. So putting three pieces orthogonally adjacent to the Ball (either directly or forming a "chain", touching the Ball) can be punished too. In general, this rule punishes "clustering together", which is the "core" of the Catennacio. on May 30th, 2010, 1:54pm, christianF wrote:The readers here are evidently good puzzlers, so here's a puzzle: Can a similar indirect obstruction be constucted with the shielding rule in place? |
| Yes It took me a minute to construct it. I forgot to put other white pieces on this image so put one white Lion somewhere on the board. The position does not fall under the "shielding rule": 1st ring - 3 squares, 1 black piece - no shielding 2nd ring - 5 squares, 2 black pieces - no shielding 3rd ring - 7 squares, 2 black pieces - no shielding 4th ring - 9 squares, 1 black piece - no shielding. Both white horses cannot move. The white Lion cannot reach the Ball. So this is a clear "Obstruction", which won't be detected with the current algorithm. By the way, my variant of "shielding" rule "punishes" this position - Black player formed a group of 4 pieces. Removing one of them either opens the way to the Ball or unblocks one of the white horses.
|
« Last Edit: May 30th, 2010, 7:49pm by Arty » |
IP Logged |
Only those who attempt the absurd will achieve the impossible.. (Escher)
|
|
|
christianF
Forum Moderator Forum Guru
Arimaa player #4019
Gender:
Posts: 804
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #161 on: May 31st, 2010, 2:13am » |
|
on May 30th, 2010, 3:05pm, MarkSteere wrote: Generally: the difference between disallowing something and making it a 'red card' (that is: allowed but punishable) is that in the latter case the offence can be made on purpose. Like not minding the punishment if it gets you two pieces in a scoring position. And the punishment in turn is an option: like not capturing a piece at the cost of a move, because you have a win in four moves. Your nutshell is growing, Pinocchio |
| I'm not sure what you mean nor what your 'pinocchio' association means, but it is evident that you're enjoying youself in your own way. The remark I made about disallowing something or to make it a 'red card' offence is a general remark and might be worth thinking about, even by a brilliant inventor such as you, because, frankly, I don't think you quite understand. We're still eagerly awaiting your insights regarding the game, instead of constantly repeating your smallminded remarks from the sideline .
|
« Last Edit: May 31st, 2010, 2:36am by christianF » |
IP Logged |
|
|
|
christianF
Forum Moderator Forum Guru
Arimaa player #4019
Gender:
Posts: 804
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #162 on: May 31st, 2010, 2:32am » |
|
on May 30th, 2010, 7:41pm, Arty wrote:Well, it's not about "straight and unbroken line" but any group of orthogonally adjacent pieces of the same color. So "L"-shaped group of 4 same-colored pieces will be punished too. I also got an idea to include the Ball into the counting. So putting three pieces orthogonally adjacent to the Ball (either directly or forming a "chain", touching the Ball) can be punished too. In general, this rule punishes "clustering together", which is the "core" of the Catennacio. |
| Ah, three max. I must admit that your solution has the lure of simplicity and I can't see anything wrong with it just now. I'll have to chew on this one a bit more, but you might be on to something. Implicitly, my concern is the best possible accomodation of the game's spirit, not the best possible preservation of my ego, you can understand that Mark, can't you? So I'll consider it for a few hours and see how it feels. Up front, and especially for Mark: it should still be a red card offence. If you disallow it, the system would disallow it at every move, i.e. a square the occupation of which would violate it, would not even show up as a target square for the selected piece. The game would thus be hampering itself. As it is, the system checks the position after submitting and doen't prevent the player from violating it at any stage, but simply gives the opponent the right to punish it. The opponent may find reasons to violate it despite the punishment, as pionted out previously. Sacrificing a piece that way is no different from sacrificing a piece in chess. Likewise an opponent may choose not to accept the sacrifice (for instance if he has a win in four). All much simpler and more 'fluid' and far more interesting than disallowing the configuration. on May 30th, 2010, 7:41pm, Arty wrote:I also got an idea to include the Ball into the counting. So putting three pieces orthogonally adjacent to the Ball (either directly or forming a "chain", touching the Ball) can be punished too. |
| You're aiming at a rule that implies obstruction, right? With your rule, the only possible obstruction configuration would be around a ball in a corner. Making the ball a 'piece' would prevent that configuration from going unpunished, right? With a bit of luck this may turn out to be the one generic rule I was looking for to cover both obstruction and shielding. One rule against 'clustering'. Does making the ball a 'piece' in Arty's concept really prevent obstruction as it appears to do? 'Prevent' in the sense that every form it can take will be punishable? I sincerely hope so.
|
« Last Edit: May 31st, 2010, 3:45am by christianF » |
IP Logged |
|
|
|
Arty
Forum Full Member
Arimaa player #3423
Gender:
Posts: 16
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #163 on: May 31st, 2010, 4:07am » |
|
Quote:You're aiming at a rule that implies obstruction, right? With your rule, the only possible obstruction configuration would be around a ball in a corner. Making the ball a 'piece' would prevent that configuration from going unpunished, right? |
| Unfortunately, this rule doesn't include all kinds of Obstruction. With the Ball on A2 and two pieces on C3 & B4, an opponent's Horse cannot reach the Ball. So if this is the only kind of pieces the opponent has then the situation is Obstruction. Besides that it seems so that my variant includes all other kinds of Obstruction.
|
|
IP Logged |
Only those who attempt the absurd will achieve the impossible.. (Escher)
|
|
|
christianF
Forum Moderator Forum Guru
Arimaa player #4019
Gender:
Posts: 804
|
|
Re: Essay by Christian Freeling on inventing games
« Reply #164 on: May 31st, 2010, 5:58am » |
|
on May 31st, 2010, 4:07am, Arty wrote:Unfortunately, this rule doesn't include all kinds of Obstruction. With the Ball on A2 and two pieces on C3 & B4, an opponent's Horse cannot reach the Ball. So if this is the only kind of pieces the opponent has then the situation is Obstruction. Besides that it seems so that my variant includes all other kinds of Obstruction. |
| Let's first establish that the shielding rule does require the obstruction rule too. The 'cluster' rule is much simpler and seems for all intents and purposes to lead to a similar spreading. Moreover, if indeed your variant includes all other kinds of obstruction, it will allow the formulation of a much simpler obstruction rule, that systemwise only needs to kick in after all king's move fieldpieces of either side have vanished. And that's not the most common of situations, most likely. For the Zillions machine it would mean less checks and more speed too (although branch density makes HanniBall a tough case for computers, regardless). So let's go for it
|
|
IP Logged |
|
|
|
|