[ List Earliest Comments Only For Pages | Games | Rated Pages | Rated Games | Subjects of Discussion ]
Comments/Ratings for a Single Item
Game Courier Tournament #1. A multi-variant tournament played on Game Courier.[All Comments] [Add Comment or Rating]I would suggest to make presets for all the games which appeared in the 2003 PBEM Tournament, and if need be to add them to the list before the presets are ready, except perhaps for Omega Chess in case copyright threats loom over the site.
Yes, there are some games that MUST be included in the list of electable, so there is the need of Presets for these games. By example, Rococo (Why it is not yet a preset for this game?, I'm going to construct one as soon as possible!), Marsellais (other example of a good and forbidden game in Courier!), Crazy 38, Michael's version of Luotuoqi (why not?.), Star Trek 3D (Kobayashi Maru or Warp Zone variants?), The Central squares, Ximeracak, Fischer Random Chess, Italian Progressive Chess, Korean Chess, Beautiful Sun Chess, Makruk, and many other very interesting or popular variants, actually without a preset in Courier, you can include Taikyouku Shogi (well, if there are candidates to play it in the Tournement)
Makruk already has a preset. Tony Quintanilla and L. Lynn Smith have been working on a preset for one of Smith's 3D Star Trek variants. Fischer Random Chess is currently beyond the capabilities of Game Courier. As for the others, I will let those who are interested in them make the presets.
Regarding the games from the 2003 Multivariant Tournament, the following already have Game Courier presets: Cavalier Chess, Chess with Different Armies, Grand Chess, and Hostage Chess. Those without presets are these: Chess on a Longer Board with a few Pieces Added, Extinction Chess, Omega Chess, Rococo, and Ximeracak. I think I never did a preset for Omega Chess mainly because the PBM system could not handle its peculiar coordinate system. Game Courier can now handle it with the custom shape method. But, since it is proprietary, I will hold off on making a preset for it. As for the others, I'll leave them to those who are interested in including them in the Game Courier tournament.
I just got a new idea concerning how I could handle the approval poll. Instead of having a period of time for you all to get new presets made, I could start the poll sooner by creating a polling page that would automatically update itself whenever new presets are added to the site. Basically, it would use the same database as the index pages do, so that it could automatically list all games there are presets for. As for the name of each checkbox field, I could use the ID given to the page. I may need some help from David Howe on implementing it, but I'm sure it could be done.
Because there are already over 90 presets, and it may reach 100 soon, I'm thinking of requiring votes for a minimum of 20 games. I think it needs to be this large to provide some reasonable chance of overlap between how people vote. Let me know if you think 20 is a reasonable minimum or if it should be higher or lower. I expect there are some people here who could easily pick 20 or more, while there are others who don't know that many variants yet. Would 20 be too much to ask of newcomers?
I have two ideas for how the tournament can be run. One is to do it as a set of concurrent sub-tournaments. Each sub-tournament would focus on a single game, and it would consist of a series of elimination rounds to determine a champion. The champion of the tournament as a whole would be whoever is champion of the most sub-tournaments. Prizes would be awarded for championship in each sub-tournament, and a bonus prize would be given for winning the whole tournament. By this method, winners would play more games than losers. The other idea is for everyone to play the same number of games. Points would be given for each win or draw, and the whoever scored the most points would win the tournament. This is how the last multivariant tournament was done. With each method, appropriate methods of tie-breaking would be employed if needed. Let me know your thoughts on which method you prefer and what reasons you have for favoring one or the other.
I would suggest an intermediate format. Have all contestants play a six or eight-game round-robin in a pool of nine to twelve different games, like in last year's tournament. Afterwards, use three-game playoffs in which the contestant who got the lower place in the round-robin always plays Black, either between number one and number four and number two and number three, or just between number two and number three, and then between the two survivors. But having to play many variants several times looks too rich.
I'm thinking the ideal format for the tournament will depend on how many people sign up for it. If only a few people sign up, then it will be feasible for everyone to play everyone else at all the games. If a moderate number sign up, then my sub-tournament idea would work well. It may be best suited for somewhere in the neighborhood of 8 players. With that many, each sub-tournament would have 3 rounds. If 4 rounds isn't too much of a burden, then it could handle up to 16 players. Something like what Antoine suggests may be suitable for larger numbers. I've also been thinking of a series of elimination rounds from one variant to the next, which I think was part of what he suggested. Another possiblity for large numbers would be to place different groups of 8 into different sub-tournaments.
THIS IS AN ELECTORAL ADVERTISEMENT: Vote for Kamikaze Mortla Shogi!. It is nice! Well, you have some other good options, not only in the initial list, there are good new games in the list, and it can be increased in the next days... What about Rococo, Extinction Chess, Pocket Polypiece, Star Trek 3d, Viking...? And you have still the option of increase your list!. If you have not voted yet, be encouraged and VOTE NOW!.
I mean: Vote for Kamikaze Mortal Shogi!. Mortla is an involuntary mistake
Since you're lobbying for your favorite games, I'll add a few words on some of them, then do some lobbying of my own. It's just possible that we're biased, because we created the game, but Kamikaze Mortal Shogi is truly a great game. I would even recommend it above my other games. I daresay I even find it more interesting than Shogi. Moving on to Pocket Polypiece. During the 43-square contest, I thought this game would be the strongest competition for my own entry, Voidrider Chess. I was surprised that it didn't win a placing. When I played it against Zillions, I found it to be an intriguing game with interesting new tactics. I'll start my lobbying with recommendations among my own games. Besides Kamikaze Mortal Shogi, my favorites among my own games currently include Bedlam, Voidrider Chess, Interdependent Chess, Clockwork Orange Chess, Eurasian Chess, British Chess, Grand Cavalier Chess, and Storm the Ivory Tower. Bedlam, Clockwork Orange Chess, and Interdependent Chess may appeal to fans of Shogi and Chessgi. Eurasian Chess, British Chess, and Grand Cavalier Chess may appeal to fans of Chinese Chess or Grand Chess. Storm the Ivory Tower may appeal to fans of Chinese Chess, Korean Chess, or Smess. Among other games, my favorites include Shogi, Chinese Chess, and Hostage Chess. David Pritchard called Hostage Chess the Chess variant of the decade, revoking that title from Magnetic Chess. Unfortunately, my very favorite game can't be done on Game Courier. But, for the sake of context, I'll mention that it's Knightmare Chess.
Excellent is for the tournament! If you have a game in mind that needs a Preset, reply by a comment. If you already know how to create Presets and just need help with minirules and posting, or details, send the URL of your Preset to me at my name link address. As time permits (and if feasible), I'll help -- in the order received, of course.
Aproval poll partial results must be visible in this page, I think
When there are several variants on the same preset (AKC I and II, Rococo and Rococo with mirror arrays, Takeover Chess and Takeover Chess on 64 squares), which one is chosen? (I doubt the first variant is always the one which is attracting the votes.)
In design contests, usually the main variant is assumed to be the game being considered. Personally, with regards to Takeover Chess, I think that the game in 64 squares is more interesting. However, this is not easy to clarify at this point. Perhaps if there is more than one variant the players should be able to choose the variant they wish to play.
When I begin the next poll, I will be listing variants by name instead of by preset page ID. Since the method of voting I'll be using is cloneproof (explained below), it isn't a problem to include multiple versions of the same game in the next poll. I already plan to exclude very similar games from both being played in the tournament. If a pair of very similar games both ranked highly in the next poll, only the higher ranking game would be included in the tournament. My reasoning behind this is that wanting one thing and wanting another doesn't always imply that you want both together. Presumably, we want a bit of variety in the tournament rather than close variations on the same game. So, for example, Shatranj and Chaturanga will not both be played in the tournament, and Yáng Qí and Eurasian Chess will not both be played in the tournament. However, I am not counting two games as very similar when one is a standard regional variant and the other is a variant of it. So, for example, Yáng Qí and Chinese Chess could both be played if they get enough votes. Here's the significance of a cloneproof voting method. Consider the Borda Count method, which is not cloneproof. It gives one point to each preference in last place, 2 points to each in next-to-last place, and so on going up. Using letters to consider preferences, consider these votes: 60 ABC 50 BCA A majority prefers A to both B and C, meaning that A should win. But B has a higher Borda Count than A. It has 270 points vs. A's 230 points. The reason it has so many points is because it is part of a clone-pair with C. Suppose that B and C are two very similar variants, such as Shatranj and Chaturanga, while A is something very different, such as Shogi. The votes come out like this, because those who like Shatranj also like Chaturanga nearly as much. This sort of thing would screw things up if the voting method wasn't cloneproof, but the method I'll be using is. Given these votes, it would give the win to A.
I had been thinking the same thing about Gothic and Grand Chess. They use all the same pieces, and they differ only in board size, setup, and rules concerning Pawn movement and castling. They're more alike than Eurasian Chess and Yang Qi. Barring any serious objections, I'll add this pair to the list of very similar games.
Any thoughts on how many games it is reasonable to expect each person in the tournament to play? Let me offer a suggestion and get feedback. The tournament could be set up so that each entrant will play in five subtournaments, each subtournament consisting of eight players. With eight players to a subtournament, each subtournament would have three rounds, with four eliminations in the first round and two in the second round. If each entrant played in five subtournaments, his games would fall between a minimum of five and a maximum of fifteen. As for the logistics of deciding which games entrants would play, the top five ranked games would be played by everyone if exactly eight people signed up. If more than eight signed up, then there would be additional subtournaments, but not everyone would play in each one. Assignments to subtournaments would be made based on a list of ranked preferences from each entrant. Anyway, deciding the exact logistics of the tournament is not as important right now as it is to decide on a reasonable number of games to be played. A minimum of five, a maximum of fifteen, and an average of ten seems fine to me. But I would like to know what other people think.
After doing a bit of math, it looks like the subtournaments method of running the tournament wouldn't work out well unless we had exactly eight people in the tournament. For example, if we had nine people, we could not evenly distribute people among six eight-person subtournaments. At best, six people would each play in five subtournaments, and three would play in all six. One possible alternative is to stick with subtournaments but to put seven people in some subtournaments. In the case of nine people, three of the subtournaments could have seven. The one drawback to this alternative is that some people would have to sit out the first round in a subtournament, then play against one of the three winners in the next round. A second alternative is to have some preliminary elimination rounds that reduce the number in the tournament to eight, then let the eight people compete in a few subtournaments. A third alternative is to forego subtournaments and do something else. One possibility is to just have everyone play so many games, each against a different opponent, moving first and second in an equal number of games. Points would be given for wins or draws, and the tournament winner would be whoever gets the most points. Ties would be broken with an extra game. These alternatives have been for eight or more. If we get exactly eight, a set of subtournaments would work well. If we get fewer than eight, then it may be best to just have everyone play everyone else at something, then total up points. Depending on how many signed up, people might play two or three games with each of the others.
Ideally, everyone would get to play everyone else. This is desirable from a social perspective, because it allows everyone to meet everyone. It is also desirable from the perspective of fairness, because it is more fair to declare someone the winner if he has played everyone else. This means that he has played the same opponents as others and that he has played against each of the losers. With large numbers of players, having everyone play everyone else becomes less desirable, because it may overload the players with too many games to play. In that case, the social goal may be compromised, since it is not critical. Instead of fully meeting it, it may be maximally met by giving a player a new opponent for each game. But the goal of fairness still remains as important as before, and it should not be compromised if at all possible. The advantage of the subtournament method is that it works best for maximizing the fairness of the tournament. But since it is not doable with most numbers, I have proposed a different method for more than eight players. It is a three-round method with eliminations after each of the first two rounds. Four games are played in the first round, four more in the second round, then three in the last round. The first round reduces the players to eight, and the second round reduces the players to four. In the first round, it would not be fair to immediately give the win to the highest scoring player, because there would be some players who have not played each other, including some who have not played the highest scoring player. But it does seem fair to eliminate all but the top eight players. Odds are good that the best player will be among these eight. In the second round, players play four more games, as much as possible against new opponents. At the end of this round, it is still possible that some of the players in the second round have not played each other, though it is no longer an inevitability. So, while it may not be fair to declare the highest scoring player in this round the winner, it does seem fair to eliminate the four lowest ranking players. Finally, in the third round, all remaining players play each other at one game apiece, and the winner of the tournament is the winner of this round. This method is also described in the text I am adding to this page, and it includes details on tie-breaking, which I have not included here. In the comments section, I am focusing more on the justification for this method.
I agree that players should be given a choice of games in the first round. If we go with the three-round method I described, then I would recommend letting players rank their preferences among the top 11 games, then do my best to assign everyone his top four games in the first round. Assignment of opponents would be based on who shares your top choices. As players continued through the rounds, they would eventually have to play some of their less preferred games. To retain some choice of games in the last round, each pair of players would have the option of playing any game they had both won against other players in the tournament. This option could even be given during the second round. Before the tournament begins, I plan to make Game Courier keep track of how much time each player has taken. This could be done by creating a timestamp list that parallels the movelist. I could also create a timelimit function that causes a player to automatically lose when his time runs out. So that all games in the tournament could be manually checked to see whether the timelimit function was being used, I would also include information on its use in View move. As for the timelimit itself, how about 30 days per player. If both players kept an even pace, this would allow a game to last as long as 60 days. Then, each round would be given two months to finish. If everyone happened to finish the round sooner, then the next round could begin sooner. With three rounds, the tournament would last up to six months but could take less time. I think four games at once may be doable. Game Courier makes it easy to play multiple games at once, but most people will be playing in their spare time, and they shouldn't be too overloaded with games. Playing four games at once, most people may be able to manage a pace of two moves a day in each game.
Fergus wrote, 'most people may be able to manage a pace of two moves a day in each game.' Most, but not all. I work strange hours, and am usually online between 23:00 and 01:00, when most honest folk have gone to bed. That makes more than one move per day problematic. From my point of view, more simultaneous games (say, 8) on a slower time control would be better.
I will note that itsyourturn.com normal tournaments allow 48 hours to respond to a move for normal tournaments, and 28 hours for fast tournaments (with the clock off on weekends). Some of us have fairly full lives, and two moves per day in multiple games is flatly unreasonable. I do not play from work, and I do not usually play until the kids go to bed. That gives me time generally for one set of moves per day, if I am not (as it occasionally happens) completely exhausted.
The focus has been on Fergus' comment about making 2 turns per day in four games. I think that the overall time-limit idea is very valid. This is parallel to the idea of a chess clock, of course. From what Fergus says, it seems each game has its own time limit. 30 days per game may be a bit tight, how about allowing 60 days, or maybe even 75? Many games end in about 30 moves, so that means that a pace of a move every 48 hours, as Peter suggests, may be doable. On an active day, one can pick up the pace. On a slow day, or week, run the clock a bit. Some games may take longer to resolve than 30 moves, for example the larger games, so a more liberal time limit would be adviseable. Also, chess variants players are a select group, let's not force players out by overstressing the time limit. Its better to have players than a fast tournament. Note that up to now CVP tournaments have had at most 5 registered entrants! The 14 current voters is a fantastic possibility. They should be encouraged to participate by allowing ample time to play.
25 comments displayed
Permalink to the exact comments currently displayed.