Welcome, Guest. Please Login or Register.
Dec 14th, 2024, 4:17am

Home Home Help Help Search Search Members Members Login Login Register Register
Arimaa Forum « Essay by Christian Freeling on inventing games »


   Arimaa Forum
   Arimaa
   Off Topic Discussion
(Moderators: christianF, supersamu)
   Essay by Christian Freeling on inventing games
« Previous topic | Next topic »
Pages: 1 ... 63 64 65 66 67  ...  78 Notify of replies Notify of replies Send Topic Send Topic Print Print
   Author  Topic: Essay by Christian Freeling on inventing games  (Read 539980 times)
christianF
Forum Moderator
Forum Guru
*****



Arimaa player #4019

   


Gender: male
Posts: 804
Re: Essay by Christian Freeling on inventing games
« Reply #960 on: Sep 23rd, 2012, 6:40am »

The current opening protocol was chosen because it was clear from the onset that Mu could be played with more than two players. Being triggered by Martin Medema's Atlantis, how could it not.
 
There will be no multi-player applet for Mu in the forseeable future. So Velox might benefit from a switch to the pie-rule as mentioned above. It is only applicable to the current two-player version, but since multi-player Velox would probably end up in total chaos anyway, who cares?
 
Levis is another matter, because contrary to Velox it lends itself excellently for both over-the-board-play and for multi-player games, and these have no use for a pie. So we'll stick to the current opening protocol, since in Levis it serves any number of players equally well.
« Last Edit: Sep 23rd, 2012, 10:56am by christianF » IP Logged
christianF
Forum Moderator
Forum Guru
*****



Arimaa player #4019

   


Gender: male
Posts: 804
Re: Essay by Christian Freeling on inventing games
« Reply #961 on: Sep 29th, 2012, 6:54am »

So a levis game can be rather uneventful. After Red_18 it looks like Red has plenty of time to get to the conflict zone around the one-cell lake first, and Red_19 seems to confirm this view. At move 20 Red even acts on that by taking a move to activate reinforcements from the back.
 
That must have been a wrong decision. Purple_20 brings a couple of strong columns in alignment, and Purple can subsequently paste a wall - not with any piece on it, but it seals off the majority of the territory, so who needs a piece on it?
 
Contrary to that, our new game came out of the placement stage with a totally different dilemma. Red is condemned to a cut-off strategy and Purple must figure out if he can get fast enough out of a tube to overrun his opponent.
« Last Edit: Sep 29th, 2012, 6:55am by christianF » IP Logged
christianF
Forum Moderator
Forum Guru
*****



Arimaa player #4019

   


Gender: male
Posts: 804
Re: Essay by Christian Freeling on inventing games
« Reply #962 on: Sep 30th, 2012, 1:46pm »

Fritzlein is right in that you can't foresee everything. Mu's rules capture its spirit, but the opening protocol, the way to achieve an initial position, defines a strategy based on the 'speed vs position' dilemma. Two forces radiating out for a head-bang encounter, where one comes out victorious & never mind the rest of the game. The crucial phase is sifting out the right starting segment. In our current game Purple had three tubes, each of them capable of fast growth. Yet each of them would have lost to the central Red segment. The segment Purple eventually selected was much slower but better positioned. This became clear after a few moves, and the game was essentially over after that. It's the Sumo wrestling syndrome, and fun as it may be, it justifies some reflection on the opening protocol.
 
An initial position with one filled segment finds its basis in Atlantis. It was one of the anchors for the thoughts that led to Mu's discovery. There was no other basis and it works out fine if you like Sumo wrestling and take the fact that it doesn't quite seem to serve Velox in the bargain. Which I don't like to do, actually.  
 
In terms of change, my first thoughts usually are to omit something, rather than to add something. So I suggest to reflect on omitting the entire 'clearing' stage, and proceed with movement right away, that is: after the second player has placed his final segment.  
 
One advantage is that Levis and Velox would have the same protocol. Velox might even profit from a progressive movement stage: after the second player has placed the last segment, the first player makes one move, then the second makes two, then the first makes three, and so on till both arrive at 'one move per segment' as usual. Another advantage is that it is a simplification (not to mention that the current protocol could remain optional for head-banging Sumo wrestlers).
 
The big difference will be in the strategy of the lay-out phase. A player's main incentive would to keep his segments clustered to ensure the pieces can be combined into columns as soon as possible. Tubeology would hardly play a role. 'Protuberances' would be more likely, in areas that have a high density of like colored segments. Longer ones would ensure fast growth, as usual, but it would take time, too much time most likely, to assemble enough men in them to start the process.
 
The opponent of course also tries to cluster his segments. Surrounding segments of an opponent is all but impossibe, so there's no alternative. The result of a lay-out will therefore usually be a fairly compact division in two areas, each occupied by a different player, with embryonic tubes along the edges to get growth started as soon and as efficiently as possible.
 
In the initial movement stages, you'd run of course a risk of pieces being captured, so care should be taken to stay out of range of the opponent's pieces, or have pieces covered. It is not inconceivable however to sacrifice a piece for speed: once you get growth started, it may quickly be replaced by new ones.
 
As Ed remarked: the proof of the pudding is in the eating, so shortly we'll serve the new recipe and see how it works out.
« Last Edit: Sep 30th, 2012, 3:25pm by christianF » IP Logged
christianF
Forum Moderator
Forum Guru
*****



Arimaa player #4019

   


Gender: male
Posts: 804
Re: Essay by Christian Freeling on inventing games
« Reply #963 on: Oct 1st, 2012, 4:38am »

A night's sleep and and my thoughts have arranged themselves (as ever: more or less). In random order:
  • Both the new and the current protocol will be optional in the applet, because the current one works with any number of players and the new one doesn't. A multi player version will not appear in the forseeable future though. Multi player Velox would probably end up in relative chaos anyway, so we're not even planning that.
  • Two player Velox will in all probabillity profit from the new protocol, especially if placing the last segment is followed by a progressive movement phase as described in the previous post. So a prior suggestion to use a pie may be abandoned.
  • Where in the current protocol each player starts with 7 men, in the new one it will be as many men as there are segments. The 'segment base' runs from 4-10. A minimum of 3 men is needed to start growth.
  • Since clustering one's segments in the placement phase seems good strategy (and not hard to achieve), the new protocol will render far more 'compact' boards. Since in a compact environment the average capacity of the cells will be higher this slows down growth. At the same time there's more positional interaction.

IP Logged
christianF
Forum Moderator
Forum Guru
*****



Arimaa player #4019

   


Gender: male
Posts: 804
Re: Essay by Christian Freeling on inventing games
« Reply #964 on: Oct 5th, 2012, 6:08am »

Ed is working on the new opening protocol (while optionally retaining the current one). Meanwhile Benedikt and I have started a game displaying the now almost familiar 'position vs speed' strategy. Red's final removal was a choice for the former.
IP Logged
christianF
Forum Moderator
Forum Guru
*****



Arimaa player #4019

   


Gender: male
Posts: 804
Re: Essay by Christian Freeling on inventing games
« Reply #965 on: Oct 7th, 2012, 5:53am »

Barring bugs it's working. Our first game using the new opening protocol is a velox of course. A couple of things to observe are:
  • This protocol only serves a two-player game.
  • The movement phase starts progressively, that is, after the lay-out phase the first player starts with one move, then the second makes two, the first makes three, and so on, till the maximum of one-move-per-segment is reached.
  • 'Tubeology' as a means to gain speed has all but disappeared, together with heavy 'laking and fjording' in the lay-out. Expect someting far more compact.
  • There will be more 'inter' in the action.
IP Logged
christianF
Forum Moderator
Forum Guru
*****



Arimaa player #4019

   


Gender: male
Posts: 804
Re: Essay by Christian Freeling on inventing games
« Reply #966 on: Oct 7th, 2012, 9:27am »

In the fourth test round of the CodeCup 2013 Challenge there appeared a new dutch competitor named Remco Bloemen, with a C++ program named "main32". It has an interesting list of results.  
 
Answering  questions about strategy appears to be extremely difficult because Symple has a strategic dilemma embedded in its object. Watching this program play is fascinating, and will most likely provide the very tools that are needed to carve out the main strategic principles. Since I'm already struggling to keep up with its opposition, I have no doubt that I'd be on the losing end against main32 for the foreseeable future. I'm looking forward to playing against it.
 
On another note, Luis has started a thread (at BGG and Lifein19x19) about what he describes as "an offbeat generalization of the Symple balancing mechanism", where I'd rather call it "a general balancing mechanism", inspired by the one of Symple.
 
To remind viewers of the essence of this balancing mechanism I replied thus:
Quote:
On a more general note, Luigi's idea may well be worth considering, given the response, but it emerged in Symple and was immediately afterwards used in Sygo. Both games allow multiple placements per turn (with each placement the growth of an existing group), OR a single placement (which must be the start of a new group).
 
The 'special move' is that as long as no growth has taken place the second player has the right to do both in the same turn. The first player can prevent him from doing so, by growing his groups, which is a regular move.
 
The balance is in the measure of initial growth one player will be willing to sacrifice to the other for turn initiative - the principle grows out of the multiple placement character of the move protocol, and cannot be embedded in a single placement game in this particular way. By lack of the 'measure of growth' as a balancing criterion, in a single placement game like Go it must be the measure of something else.
 
That's why I don't believe Luigi's suggestion, however worth considering, is a generalization of the Symple balancing mechanism.

And looking at main32's results, there at least not much indication that Symple's balancing mechanism works other than intended Wink .
IP Logged
christianF
Forum Moderator
Forum Guru
*****



Arimaa player #4019

   


Gender: male
Posts: 804
Re: Essay by Christian Freeling on inventing games
« Reply #967 on: Oct 9th, 2012, 4:17pm »

Given the development of this velox game I'm tempted to let levis keep the original opening protocol, while velox gets the new one. The new one is too slow for levis, the old one too fast for velox and I like things to be simple. Life is complicated enough as it is Cool .
IP Logged
christianF
Forum Moderator
Forum Guru
*****



Arimaa player #4019

   


Gender: male
Posts: 804
Re: Essay by Christian Freeling on inventing games
« Reply #968 on: Oct 14th, 2012, 6:28am »

It's definitive, our latest velox game convinced me that the 'reduced' opening protocol serves the game perfectly. So Velox and Levis have drifted apart somewhat, despite having the same solid core:
  • Velox is a two player abstract that starts out with as many men as the players have segments (while Levis starts with 7 men each, one filled segment).
  • Movement in Velox starts after the second player has placed the last segment and proceeds progressively: one move for the first player, two for the second, three for the first and so on, till the number of 'one move per segment' is reached. In Levis, after placing the last segment, there follows a 'clearing' stage to determine each player's starting segment, and of course Levis allows only one move per turn.
  • Velox boards tend to compact shapes because moves can be combined quicker to columns that are capable of explosions and therewith growth. Levis boards develop 'tentacled'.
  • Velox cannot be played 'over the board' because moving implies trial and error, and only an applet can take trial moves back effectively. In return it gives a gaming experience that is brand new. I don't mind the qualification "a hybrid between an abstract game and a pinball machine", in fact I coined it. Levis remains exactly what it was, multiplayer and easy to play over the board, provided one has the fairly simple materials.
At the moment I don't have the time to put the new Velox protocol in the rules, but the above are the ones implemented right now Smiley .
 
P.S. We've started another one. You guys may prefer to keep watching, but the aliens have landed, they're not going to crash anymore - you can actually play them Grin .
« Last Edit: Oct 14th, 2012, 7:00am by christianF » IP Logged
christianF
Forum Moderator
Forum Guru
*****



Arimaa player #4019

   


Gender: male
Posts: 804
Re: Essay by Christian Freeling on inventing games
« Reply #969 on: Oct 17th, 2012, 12:53pm »

on Oct 14th, 2012, 6:28am, christianF wrote:
P.S. We've started another one. You guys may prefer to keep watching, but the aliens have landed, they're not going to crash anymore - you can actually play them Grin .

Yeah, right, players may crash though. Ed frankly admitted that this catastrophy was more due to his initial lay-out (no 'protube' like Red) and less than effective initial moves. So now we're warned that there's some 'opening theory' involved in the lay-out as well as the most effective starting moves in the 'progressive' early stages.
 
P.S. Note that Ed plays second again, at his own request.
« Last Edit: Oct 18th, 2012, 10:45am by christianF » IP Logged
christianF
Forum Moderator
Forum Guru
*****



Arimaa player #4019

   


Gender: male
Posts: 804
Re: Essay by Christian Freeling on inventing games
« Reply #970 on: Oct 20th, 2012, 3:45pm »

on Oct 17th, 2012, 12:53pm, christianF wrote:
P.S. Note that Ed plays second again, at his own request.
So as to show that moving first or second has absolutely nothing to do with it. This purple catastrophy was more due to Ed's initial lay-out (no 'protube' like Red). This red catastrophy shows how Ed adapted by making tubelike shapes. So did I, but I next chose a 'one center of growth' strategy, while Ed divided his forces to make two such centers. That proved the better strategy.
 
So 'tubeology' is back. A boardshape doesn't have to be 'compact' to assure good cooperation of men in the progressive movement stage, as I had assumed for convenience. These insight will go into our next game, and who knows what will happen next Smiley .
IP Logged
christianF
Forum Moderator
Forum Guru
*****



Arimaa player #4019

   


Gender: male
Posts: 804
Re: Essay by Christian Freeling on inventing games
« Reply #971 on: Oct 24th, 2012, 7:26am »

on Oct 20th, 2012, 3:45pm, christianF wrote:
So 'tubeology' is back. A boardshape doesn't have to be 'compact' to assure good cooperation of men in the progressive movement stage, as I had assumed for convenience.

I lost on a less than intelligent lay-out of my segments, leading to an arrears in development and initial growth. In Mu positive feedback is quite enough in such a case to trigger a tsunami by the opponent.  
 
A regular and usually achievable starting configuration to start growth with four-segments is shown top-left, the goal being the move combination top-right. Note that it involves moves from three different segments, two of which already saw a move in a previous turn. The rightmost segment is the 'tube' or spawning chamber, a segment with a low average capacity, making it the best starting area for chainreactions and the growth implied.
 

 
The bottom-left starting configuration is more effective, but maybe also less achievable against an opponent who is aware of it, as most would be, considering I'm pointing it out here. The number of moves is the same, but a short distance down the line, the chain reaction will spill over to the second segment, that has a lower average capacity than its counterpart above, and thus will speed up growth more effectively.
 
These are only a few of a number of possible configurations that allow a quick start and that show that tubes keep playing a key role in both Levis and Velox, despite their differences in move protocol.
 
P.S.
Note that the bottom one has a disadvantage too. In the top configuration, once the second segment blows, it will easily spill over to two more segments. In the bottom one it only can reach one (albeit faster). That means that the bottom configuration will at a certain point, and for a short while, have less combinatory power.
 
P.P.S.
I'll have to catch up on some of those ever accumulating requirements in and around the house Tongue and do some work on mindsports too, so we've put Mu on hold for the moment. If you don't want to have some fun, don't try it Wink .
« Last Edit: Oct 24th, 2012, 7:53am by christianF » IP Logged
christianF
Forum Moderator
Forum Guru
*****



Arimaa player #4019

   


Gender: male
Posts: 804
Re: Essay by Christian Freeling on inventing games
« Reply #972 on: Oct 25th, 2012, 6:12am »

I admire the best work of some of the more prominent inventors of abstract games, if that be the word to use in a realm that is largely neglected in the first place. One of them is Luis Bolaños Mures who among others invented an original and highly organic 'dynamic goal' connection game called Yodd and a beautiful and equally organic quintessential 'approach' game called Ayu.
 
When he's at his best, simplicity, originality and elegance are Luis' hallmarks. He's quite driven though and sometimes publishes prematurely. Well, who doesn't? He also leans towards the dogmata of the Church of Cylophobia and Hard Finitude. Games should be free of cycles, finite, drawless and perfectly balanced in terms of turn order advantage. Some members require, as an additional condition, that they should not be by me Wink .
 
Don't argue that some of the worlds most prominent abstracts like Chess, Shogi, Xiangqi, Go and Draughts don't comply. The Church agrees and thinks these should be banned.
 
Now it turns out that Ayu is not theoretically finite. Luis starts with "Bad news". The choir sings "a stab in the liver!" and "dammit!" and "it was our favorite of your games!" and "what now lord, what now?".
 
Because of what? The cycle discovered in Ayu requires cooperation between the players to emerge. That's called "soft finite". It means that you would need two church members cooperating to convince one another of something they agree upon in the first place. You cannot play to win and cooperate to prove the game isn't finite at the same time.
 
The discovery of a 'cooperative' cycle takes nothing from the game in terms of simplicity of concept, elegance, quintessence or playability. Games are the essence, not dogmata about them.
 
« Last Edit: Oct 28th, 2012, 7:19am by christianF » IP Logged
christianF
Forum Moderator
Forum Guru
*****



Arimaa player #4019

   


Gender: male
Posts: 804
Re: Essay by Christian Freeling on inventing games
« Reply #973 on: Oct 27th, 2012, 9:14am »

We've modified the rules of Mu to reflect the differences between Levis and Velox as they crystallized in the past weeks. We've also widened the choice of the number of segments to 12 (which makes te maximum boardsize one less than a 13x13 Go board). Velox benefits from a somewhat larger board, if only to give the progressive start of the movement stage more combinatory scope.
 
So we're engaged in a 12x12 Velox game now, enjoy Smiley .
 
Edit:
This appears to be the first Velox game that is not characterized by a gross lack of insight regarding possible lay-out strategies.
« Last Edit: Oct 30th, 2012, 12:40pm by christianF » IP Logged
christianF
Forum Moderator
Forum Guru
*****



Arimaa player #4019

   


Gender: male
Posts: 804
Re: Essay by Christian Freeling on inventing games
« Reply #974 on: Oct 30th, 2012, 3:20pm »

So what did we learn of this one? For starters you need to build fast access routes to low capacity 'tubes'. You need the cooperation of at least 3 men to initiate growth, so with 12 segments you can choose four growth centra with 3 men each, or three with 4 men each, or two with 6 men each. In this game Red chose the latter against Purple's three starting areas.
 
Purple's growth was faster, which is good, but fast growth isn't the whole story. In the inevitable encounter Red, for reasons not entirely clear, managed to occupy all sections of the wall, while keeping routes between his two sections open. This allowed him to pit his forces against the smallest of Purple's sections, while locking the rest of Purple's forces out. Invading and conquering this section turned the tables in terms of territory.
 
So having more growth centra appears to bee advantageous in terms of growth, but leaves a vulnarability because each section in itself may be harder to hold on to.
 
Let's see how it works out in the next one.
IP Logged
Pages: 1 ... 63 64 65 66 67  ...  78 Notify of replies Notify of replies Send Topic Send Topic Print Print

« Previous topic | Next topic »

Arimaa Forum » Powered by YaBB 1 Gold - SP 1.3.1!
YaBB © 2000-2003. All Rights Reserved.