Herlock Sholmes: It seems pretty dumb to me. And with a bit of style sheet hackery, one will be able to have it displayed as a regular board as well. I don't see the added benefit of having this - but perhaps Fencer can make it a user option "display my chess board in an annoying way".

Herlock Sholmes: I fail to see how your variant is different from regular chess. All you seem to be doing it cutting the display in half, and moving it to the other side - like an old fashioned TV warping its image half a phase.

Teachme2play: It's not Glicko. Your rating doesn't change much because you have played a lot. But if you now stop playing line4, wait a couple of years, and then play again, your rating will still not much. BKR considers a rating well established if you have played a lot of games. It doesn't distinguish between having played those games in the past month or five years ago. Glicko does.

Teachme2play: I think that the score calculation should be slightly different. It's Glicko, isn't it?

No, it's not. It's neither Glicko (which would take into account how recent your rating is - if you've played a lot recently, your rating will change much less (aka, is more stable) than if you haven't played a lot recently), nor is it even Elo based (which does account for how big the score was in a match - with Elo, you can actually gain rating points while losing the match: for instance, if the rating difference say you ought to lose the match 7-3, but you only lose 6-4, you win rating points; the winner of the match looses points).

What system BK uses, I do not know. But based on observations how rating points are calculated, I can say it's neither Elo, nor Glicko (note that Glicko also takes into account which how many points a match has been won).

Rainbow Road: Both requested many times, but never implemented. Now, if you can claim either of them are actually chess variants, you'll have some chance of getting it implemented. ;-)

UzzyLady: Froglet games could also benefit from a similar automatic end: there's no point in dragging the game on once the sum of the points of the remaining frogs is less than the point difference between the players. I've been playing a couple of 3-win matches that started in 2006, and an automatic end would have cut playing time by months.

ChessVariant: 9. 10x8 version with 8 knights, 8 bishops, 8 rooks, 5 queens, 5 marshalls, 5 cardinals and 1 amazon each.
10. 10x10 version: throw in some zebras, camels, nightriders and other fairy chess pieces.

And each variant can be "wild": the system determines how many of each type each player gets (still filling the board of course, and each player gets the same set of pieces): for instance 4 knights, 10 bishops, 5 rooks and 13 queens.

And then we throw in darkness (you cannot see your opponents pieces), atomic, extinction, dice, we let the knights relay, add a behemoth and a cheshire cat, and we'll have the greatest game ever!

jessica: If white starts, it cannot be white who loses all his pieces. Black will never have more pieces than white.

ChessVariant: "The player, who lost all pieces, loses the game as well. " And again, the only player who can loose all the pieces is white player

Does black start in this variant? That's not listed in the rules.

ChessVariant: game is a draw when game ends by eliminating all the pieces (which is almost unlikely)

Unlikely? Impossible I would say. Unless the last piece can take itself.

but think twice before you eliminate

Considering "elimination" (which I presume to be the same as "capturing", the term BK uses) is mandatory it's not that a player has the option to refrain from "eliminating".

It's very fast game

On BK? A game which can last 32 moves? That's not going to be "fast".

nodnarbo: Did you play test this? Extensively? We've seen several five minute ideas implemented here which either turned out to be quite boring games (Frog legs for instance), or needing (repeated) quite drastic rule changes to make the game more balanced (Cheversi for instance).

I rather see different games (I've posted many suggestions in the past) that yet another chess variant.

coan.net: That sounds like a boring game. You still need luck, and it'll create many draws as there will quickly not be enough pieces to mate.

MadMonkey: Froglet does this, doesn't it? That is, if you click on a piece, and there's only other piece it can take, said piece is taken. Note that in Froglet there is not requirement to capture additional pieces.

But checkers could be improved. Due to capturing mandatory, there are often forced moves. No human intervention is needed. The system could do it for us.

Thad: A GREAT solution would be to be able to specify the maximum number of vacation days useable in any game//tournament.

It's not so great for games that last more than a year. I've played games that take more than 400 moves, and even if both players play 1 move/day, such games still take a year or longer. No reason people shouldn't be allowed to take vacation in the second year.

joshi: I think a black list is no good idea, people on the black list cannot get rid off that reputation.

I think a blacklist is an excellent idea. We already have whitelists (private tournaments, fellowships) where people can select who participates, creating blacklists is just another step. If I want to create a tournament without participation of a certain person whose name cannot be mentioned because the moderator of this board will modify your post without any notice, I can now create a private tournament and allow anyone in, except the person whose name cannot be mentioned because the moderator of this board will modify your post without any notice. But this is a hassle, both for me and the people signing up. Instead, if there were a blacklist one could exclude people you do not want to participate in your tournament. Or sign up only for tournaments that exclude the people you do not want to play.

coan.net: But surely the difference between "random position for each game", and "random position repeated over the whole tournament" is way smaller than the difference between "one game match" and "10 wins match" the tournament creator can make for the final match anyway. Not to mention a difference between "max 4 people per segment" and "max 20 people per segment".

IMO, only three kinds of tournament editing make sense:

1. No editing allowed at all
2. Only allow editing things that don't affect game play (name of the tournament, days after deadline, early start, add/delete game types)
3. Edit anything (time control, match types, number of participants/section, autopass, etc)

Brainking allows more than 2), but doesn't allow you to edit everything.

Could the box showing icon/name/chips at the poker tables be make a little taller? Currently (on my browser), the number of chips a person has falls halfway below the box, making the number hard to read if the person doesn't have cards, and almost impossible to read when the player has two cards (as the cards obscure the chip number).

Undertaker.: In 2008, Ladiesmanny ran out of holidays, played his moves for a few days, then extended his membership to get more holidays, to be able to resume his crawling speed.

joshi tm: I rather have games with a different game play than size variants on games, or yet another chess variant.

Some examples:

• 9 Men Morris. (Well known game)
  
• Cats and Dogs. Played on an 8x8 (Go) board. Players alternate placing stones of their colour on the board, under the restriction you cannot put a stone horizontally or vertically adjacent to stone of your opponents colour. First player who cannot move loses.
  
• Roots. Played on a 19x19 board. Players start with 8 'roots' of one stone each (roots are set up in a 4 x 4 grid, evenly spread out over the board, in alternating colours). Players alternate placing a stone of their colour on the board under the condition that each placed stone is horizontally or vertically adjacent to exactly one stone of their colour - so the groups look like root systems. If after a stone is placed one or more root systems cannot expand any further, they are removed from the board immediately. Winner is the person who eliminates all root systems of the other player.
  
• Connection games, like hex, twixt, or Y.
  
• Clobber. Played on an NxM checkerboard (N and M cannot be both odd). Initially, all white squares are filled with white stones; all black squares with black stones. One player plays the white stones, the other the black. Players alternate moves. A move consist of taking one of your stones and moving it horizontally or vertically one square to a square occupied by a stone of your opponent. The opponents stone is removed. First player who cannot move loses.
  </ul>

MadMonkey: There are zillions of freely distributable games that include Mah-Jong based games. I'd be surprised if none of them had a license that forbid using just the graphics.

For instance, the fanclub version of pysol includes Mah-Jong graphics, which are distributed under the GNU license. This means that Fencer can freely use them - the license would only mandate Fencer to do something if he was going to distribute the code of BK.

Currently, we have "random game" tournaments. The creator of the tournaments selects one or more games, and for each game between participants, a game will be randomly picked from the selected games.

I'd like to propose a variant. Instead of randomly pick a game, a match consisting of all selected games is played. For instance, one may create a "monster tournament" with "Reversi", "Reversi 6x6", "Reversi 10x10" and "Anti Reversi" as the games. All players in the same section would play 4 games against each other - all the reversi variants. You'd score 0 - 4 points (so, if it's a five player section, you could score up to 20 points). All games would start at once (otherwise, tournaments would never get finished).

Just don't sign up for a 20 player/section, 125 game "monster tournament", unless you can handle 2375 new games. ;-)

coan.net: So, if you think it's a bad thing a person might learn what kind of moves he has available, do you also think Brainking should no longer move a piece automatically if the piece to be moved has only one available move? Furthermore, should Brainking make all pieces clickable, regardless whether they can actually be moved, to prevent someone finding out he has more moves available by either hoovering with the mouse, or by using the tab key to cycle over the clickable pieces?

Vikings: Well, all 'information' such a button (which would probably never materialize) would reveal is "you only have one move available', or, in its absence, 'you have more than one move available'. It's not going to reveal anything about opponents pieces, or whether you have more than two moves available.

Not that I think it harms the opponent to show a player the list of available moves. A determined player can already find out all moves he/she can play; first by hoovering over his/her pieces (to see which pieces can be moved), then by clicking all moveable pieces and see which fields are clickable (destination squares for the pieces).

Thad: But they can already find out now. Pieces that do not have a move available are not clickable (and hence will show a destination in the status bar when the mouse hoovers over it). And a piece that only has one available move already does that move when clicked.

Well, a new year. Since you never know, it may be a lucky year for me.
So, here's my list of "wishes" - some of them I've wished for for a long
time.

First, changes in apparance to make playing some games more enjoyable:

• Differentiate between 'racing' and 'non-racing' checkers in the
  gammon games where it matters.


• Highlight captured (but not yet removed) pieces in the various
  checkers and camelot games.


• In the Frog Finder/Frog Legs, mark squares that have been guessed,
  but do not have a frog.

Wishes that reduce the number of mouse clicks, and hence the number of
server requests:

• Make shooting the default in Frog Finder/Frog Legs; require only an
  extra click if you want to guess (as shooting occurs far more often
  than guessing). If you make guess/shoot a radio button (with 'guess'
  the default) one never needs two server requests to make a single
  move.


• Use some Javascript to allow people to select their colours for a
  guess in Logic. This reduces the number of server requests by nine
  for each turn.


• When setting up a position in (Crazy) Screen Chess, keep the current
  placed piece "active" after placing - if there's still such a piece
  in hand. This allows one to set up their pawns with 9 server requests
  instead of 16 requests.


• Don't force people to roll the dice in gammon games if they have a
  non-racing checker on the bar, and the opponent keeps their entire
  home blocked with at least 2 checkers on each point (that is, don't
  force the die to be rolled if there's no way a checker can be moved).

Autoplay.

• A full autoplay (that is, the computer moves (or passes) on behalf of
  the player if the player has just one possible move. Useful for games
  like Ludo, the various Gammon variants, the Checkers variants (with
  their forced capturing), Camelot variants and to a lesser extend for
  the Chess variants.


• Since the above is unlikely to happen, how about having a button
  "Play forced move" which will be there when you only have one move
  available. Then you can play your forced moves with just one click.


• Better even, have a button "Play all forced moves" on the main page
  that playes all forced moves.

Tournaments.

• Being able to create a tournament with "all games", where "all games"
  means "all games as exist when the tournament starts", as
  opposed to the current "all games as exist when the tournament is
  defined". If you have dozens of open tournaments, it takes
  a lot of work to add a game to all the tournaments.


• Elimination tournaments that allow for a number of players that isn't
  a power of 2. That way, you can have a large elimination tournament
  without having a high risk it gets deleted because you didn't guess the
  number of willing participants correctly.

Messages.

• I'm only interested in a small fraction of the automated messages I get.
  I can easily do without the message you get when a game in a multi-point
  backgammon match ends (just notify me when the match is finished). I
  don't care whether a next round of a tournament I created starts. I'm
  not interested in the couple of pawns that sign up for a million tournaments
  and get booted off when the tournament starts because they don't have
  enough slots (how about removing pawns from *all* open tournaments when
  they filled up their slots?). I can do without messages about postponed
  or deleted tournaments. I'd like to see some configuration options about
  what kind of automated messages you want to get, and which ones you don't
  want.

Hrqls: Board game.

Hrqls: I own a copy of the game.

Pedro Martínez: Perhaps because the core of the copyright rules are international? Berne Convention for the Protection of Literary and Artistic Works is the bases of the copyright laws of all countries that signed the convention. Both the US (since 1989) and the Czech republic (since 1993) are parties to the convention.

mctrivia: This page on a US government site dealing with copyright starts with:

The idea for a game is not protected by copyright. The same is true of the name or title given to the game and of the method or methods for playing it.

Copyright protects only the particular manner of an author’s expression in literary, artistic, or musical form. Copyright protection does not extend to any idea, system, method, device, or trademark material involved in the development, merchandising, or playing of a game. Once a game has been made public, nothing in the copyright law prevents others from developing another game based on similar principles.

Rainbow Road: It has always been my understanding that a game doesn't have a copyright. Certainly, a name can be trademarked, and a description of the rules has a copyright. But you're free to make a game with the same mechanics. Just don't copy the rules verbatim and use a different name.

Currently, when you go to someones profile, you see the number of wins, draws, losses and tournament wins. It would be nice to show the number of times a person has timed out.

rod03801: People don't get better over time necessarely. But whether they get better (or worse) isn't the point. Ratings are used to give an estimate of ones true strength. The argument to use recent activity is if a rating is based on recently played games it's likely to give a better estimate of ones true rating than if it's based on games played 5 years ago. Just as a rating based on many games is likely to give a better estimate than a rating based on a few number of games.

A rating system like BKR uses some measurement to determine how likely the rating reflects your true strength. The more likely (according to the measurement) the rating reflects your true strength, the lesser your rating changes after a result. The more likely your opponents rating reflects his true strength, the more your rating changes after a result. In BKR (but also with ELO), the only thing considered for this measurement is the number of games played. Other rating systems, like Glicko based systems, also consider recent activity. The more recent you played your games, the more likely your rating reflects your true strength, and the lesser your rating will change after a result.

Here's an example why recent activity matters. Suppose Bobby Fischer was still alive. Would you consider him stronger than Anand? Fischer played his last official games in 1972, giving him a rating of 2785 - a rating he would still have under BKR rules. On the most recent FIDE list, Anand has a rating of 2783. Personally, I'd consider the 2783 which is (partially) based on games played this year to be much more relevant than a 2785 rating based on games played way back.

MadMonkey: In the way BKR is calculated, the only factors that are relevant is the rating difference between players, and the number of games. If you have played 100 games, it doesn't matter whether you've played those 100 games all in the last month, or whether your last game was 5 years ago. It's just too bad there's no factor for recent activity.

But this is the way it is, and Fencer has said many times he has no intention to change the way BKR is calculated.

Fencer: 8 months ago, I asked Would it be possible to mark captured, but not removed checkers? Five minutes later, you responded That could be done.

Made any progress in those 8 months?

MadMonkey: svg images are XML documents. They can take style sheets, including external ones. So, Fencer could use svg images for Five-in-Line/Line4/etc images and tie it into the "external style sheet" option we already have.

Then everyone can define his/her own colours.

MadMonkey: I've made some simple graphics in a few colours, and a couple of shapes. I've included the dark purple and green from IYT. They are vector graphics, so they can be scaled to any size without problems.

trohat: Imagine you play e.g. 5-win match and because of bad time management (reason can be some family business or easily some unexpected party with friends) you lose one of those games on timeout. Would you like to lose the entire match ??

Well, yes, I would like that to happen. You'd lose the match if it's a 1-game match anyway - family business or unexpected parties or not. I don't see why a multi-game match should be an exception. Heck, I can argue that in BK, it would even make more sense: considering that for ratings, win/loss ratio and tournaments a 1-game match has the same weight as a 21-point cube match or a 5-win match, BK treats a multi-game match to be a single game, so it kind of make sense to do that for timing as well.

trohat: To make my argument stronger, imagine that this game is with a very weak opponent and the loss would ruin your BKR

It wouldn't ruin it any more than losing a single game match on timeout, so I don't see why it would be 'too hard'.

I forget to add two more points.

• When playing a multi-game match (being it a cubed match, and N-wins match, or an N-game match), the time control is for the entire match - no resetting between games. So if you saved up a lot of bonus time, you carry it over to the next game. And if you time out on a game, you lose the entire match. No more waiting for many months if someone stops playing in a 21 game gammon match.
  
• You don't get a whole lump of vacation days at the beginning of the year. A new account gets a few days to start with, and one gets a new vacation day every N days you make moves. (So non-active people don't accumulate vacation days, and you don't have time-out free months early in the year. No strategical buying of membership either). N would vary on the membership levels. In BK's terms, I'd set N = 35 for pawns, N = 15 for bishops/knights and N = 10 for rooks. Of course, there would also be maximum of saved up vacation days.
  </ul>

trohat: As I said, if you're playing in such a style that you need to move in the few hours you have available, or else lose the game, you're doing that at your own risk. Note that if you have a 24h time grace period, a max pool time of 24h (1d), and some kind of bonus time, you have to play faster than even the very fast stairs on BK, and you still won't time out if the server is unavailable for one evening.

And note, my post didn't contain a proposal. It just something I would do if I had the chance.

mctrivia: What I would do if I had the power to decide timing:

• No weekends. No special days (holidays) without time outs as BK used
  to have.
  
• Time control consists of:
  
  1. A grace period.
     
  2. Pool time, both initial and max.
     
  3. Move time.
     
  4. Bonus time.
     </ol>
     You loses a game on time if your pool time goes below 0.
     
     
     If it's your move, you first have "grace period" to make your move before
     it starts eating away from your pool time. Because people live in different
     timezones, grace time should be at least 24 hours. I'd require a minimum
     24 hour grace period for rated and tournament games. In the BK system,
     playing a non-Fisher clock game with 3 days/move equals having a grace period
     of 3 days, and no pool or bonus time.
     
     
     Pool time is the amount of time you have to make all your moves (after
     running out of your grace time). This is the time you have in a Fisher
     clock game on BK. There's an initial pool time, and a max pool time;
     equivalent to the first and third components of BK's Fisher clock.
     Pool time runs down when it's your turn and haven't moved within the grace
     period, it can be added to if you make a move with "move time"; the amount
     added to the "bonus time". If "move time" is set to infinity it means you
     get your bonus time for each move, just like bonus time does in BK's Fisher
     clock games. Pool time can not exceed its set max, except for the case
     explained below.
     
     
     Note that both current clocks can be simulated with the system I described
     here.
     
  5. Vacation days must be set in advance; they won't be automatic. If you
     set a vacation day, all your games get 24 hours added to their pool time,
     regardless whether it's your turn on not. (In this case, pool time can exceed
     their set maximum). You cannot make any moves while on vacation. If you
     come back from vacation and make a move, if after your move your pool time
     exceeds its max, it's set to the max.
     
  6. Short server unavailability isn't a reason to grant another 24 hours to
     your pool time; if you time your games you'll time out on a short
     unavailability, it's your risk. In general, if the server hasn't been
     available for X hours, everyone whose turn it is gets X hours added to their
     pool time.
     </ul>
     Now, this isn't a proposal. Fencer wouldn't want to implement it anyway.
     

     
     BTW, Fencer, you need to clean up your HTML parsing. The system doesn't recognize </ol> and </ul> tags. Probably caused by the misbelieve the closing tag of the LI element is mandatory, judging by the closing tags the system adds at the end of my posting.

coan.net: 5 days will work for many Americans, but among Europeans, a three week vacation is rather common.

But I don't mind people being on vacation. I don't mind people taking the full amount of time sometimes. What I do mind is people that are on-line almost everyday, and for almost all of their moves wait till it's almost the deadline. Well, coan.net, you know which class of people I mean, don't you?

I'd like for games to progress steadily on average, while still allowing people to take vacations or the occasional break. A time control that combines Fisher + vacation days is what I'd like. Unfortunally, BK doesn't offer that possibility.

mctrivia: I very well know how time controls work. But not having vacations doesn't work for me. I will be away for a week or more two or three times a year. And you never know when a second round of a tournament starts.

I don't mind if someone takes 7 days to move when the time control is 7 days or if they are on vacation - as long as they don't for every move, while being online almost every day.

If we would have a time control that combines the Fisher clock with vacation (preferably a "strict" form of vacation that won't allow you to take vacation on a day you also made a move), I'd start using that time control immediately. Currently, I never use the Fisher clock. Sure, I could easily be able to make my moves for the next 6 months on a 5/1/5 time control, but the game or tournament may last for longer than that. And then I would lose my game, because I will be away on vacation.

trohat: A three player game? With the number of people on brainking that consistently use up almost all their time for every move I often find two players already too many players. I don't think I'd fancy any three or more player game where moves cannot be made simultanously on brainking, no matter how fine OTB game play might be.

Nirvana: From the link: Games of Yavalath tend to be short, decisive and dominated by passages of forced play.

Forced plays are ok when doing an OTB game, but, IMO, horribly in turn based site. It can take months before you're facing a real decision in a game.

Fencer: I hope I have some time for feature requests now.

Please give us different 'race' and 'regular' backgammon pieces.

Andersp: Yeah, I remember Fencer once writing on his blog on brainking..info he doesn't like doing simple things a computer can do. "Why don't we have real autopass and automove then?" was what I thought when I read it.

As for the list that started this thread, distinguishing race checkers from non-race checkers on the bar has been a recurring wish as well.

rod03801: Oh, sure it can done using notes, or post-its, or using a legal pad.

But that 1) requires to have a notes entry field, which I rather keep as tiny as possible (or rather, not show at all), as it takes up valuable screen estate on the games I don't need to keep notes; 2) requires one to consult the notes each and every time; 3) make the mental adjustment between algebraic notation and the game field; 4) have to do work which is better done by computes (and didn't Fencer write once on his blog he doesn't like to do work that could easily be done by computers?).

My request isn't that it's impossible to keep track of guessed squares. But that game play would be so much more enjoyable if it was shown on the field. (Considering that there's little to gain to guess again on an already guessed field, Fencer may even go the extra step and make an already guessed square unclickable).

Fencer: What makes you think "Frog Legs" can't be played OTB? You'd just do what they did when they played "Kriegspiel" long before computers where common place.

In the case of Frog Legs, all you need is three people and some scraps of paper. One referee that will assign the frogs and call out the hits/number of adjacent frogs, and two players.

But since this is a feature board, here's my frog legs/finder feature request: can squares that have been "guessed" (but don't contain a frog) marked in some way? With players that do one shot/week consistently (hi coan!) games take several months to finish, and it's impossible to remember which squares have been guessed before, and rather awkward to scan to move list.

coan.net: ....even though it is a much funner game when it is not played that way

Yeah, and some people think chess is a much funnier game if you let capture you queen.

With frog legs, revealing anything may cost you the game. Which is why I prefer to play moves that do not reveal anything. I agree that this is incredibly boring, but that, IMO, just shows games shouldn't be added to this site without proper testing.

P-G: 1) I would not like to change the rules of chess. 2) A bishop that adds rook movements to its existing movement is called a queen. 3) As a variant, I doubt it'll see much of a fan base; there are already many chess variants.

On an chess server I used to play, I liked to play "wild chess" (to be pedantic, "wild chess style 3", as it has several different variants, all called 'wild chess').

The rules are as follows:

1. Pawns start in their normal position (second row for white, seventh row for black).


2. A random set of 8 chess pieces is generated, subject to the constraints that the set includes exactly one king, and no pawns.


3. This set is placed randomly on the first row for white; blacks setup mirrors white (just like in regular chess).


4. There's no castling.


5. All other rules are the same as regular chess.

In this game, it might happen that you play with three queens, or four rooks.

One might even think of a 10x8 wild variant, where the set of pieces to choose from includes the Marshall and the Cardinal. (Or even the Amazon).

See also the wildchess wiki (style 3).