Thad: In games with less luck, it would be just as easy, or maybe even easier, to play a game against someone with a really low rating every four weeks in order to maintain a top rating using the system that is in place now than it would be to maintain a multi-game match and finish a game every month, so my proposal doesn't introduce any new risk or abuse.

You are right, the abuse thus introduced would be insignificant (btw the Glicko thingy also take cares of the technique you mentioned, since games with a lot of rating difference count for almost nothing in the activity).

I agree that a complete overhaul of the scoring system would be a benefit to all. BK could be a lot better if he'd just be willing to implement a few key things that matter to many players even though they don't matter to him.

You hit the nail on the head.

Thad: Possible abuse of your proposal : the #1 ranked player could always keep a 10-point match going in order to minimize the risk.

A solution to your issue (and to the general rating issue) would be :
1) Using the Glicko rating system in order to have a continuous measure of activity.
2) Taking in account the length of the match in the rating adjustements (backgammon sites know how to do that).

The bad news is that it will not be implemented.

Herlock Sholmes: Good one ! While one could argue that backgammon is a probability game anyway and that doubles introduce some "unsmoothness" to cope with, basically yes, they are just spices.

I would still say that in your game proposal the effect of a double sounds much more dramatic than a backgammon double - but since I am not playing any Dice Chess I could easily be wrong.

Herlock Sholmes: I agree with Abigail in that spicing up for the sake of spicing up usually leads to nothing but stomach aches.

As far as standard chess in concerned, I would say that castling and stalemate are indeed somewhat artificial spicings, while in contrast promoting and en passant are great solutions to potential game problems.

AbigailII:
1) The system is the one used by the US chess federation. It is in fact Elo-based like Glicko (Glicko being Elo enhanced with a clever coefficient handling).
2) The system is indeed not Glicko because it handles the coefficient differently. Glicko seems a bit better to me for the reasons you say, but USCF is not evil.
3) The system handles a match as one single entity. I think that is fine : accounting for the score in cubed backgammon matches would make no sense, and even in chess, playing for a draw in order to win the match is fine and should never lose rating points.
4) The system handles a match as one single game. Now that is very bad because the longer the match, the higher the probability that the better player wins. That opens the BK system to exploits.

Backgammon players have devised a fix for the point 4), which is called the FIBS rating. It is not perfect, but it does pretty well. Basically, you multiply the rating difference by the square root of the number of points required to win the match.

I explained to Fencer how it could be quite easily implemented here, and he said something like "maybe" (that was long ago).

AbigailII: If you are correct, then one could kill two birds with one stone and correct Abalone's defect, which is that it is impossible to break through the opponent's wall. This means that Abalone is a lot of fun at first, and as soon as one gets better it becomes awfully dull.

Long ago we experimented a bit with alternate rules, and the best we came up with was to change the start position by inserting two of each player's spheres in the middle of the opponent's position. The reason for it is that it seemed still impossible to break through when one sphere up, but possible when two spheres up. I am not sure of that though.

Fencer: That would be the part about balancing happy users against unhappy users

alanback: Those who have played Logik here know that minimizing the number of clicks must not be a high priority in Fencer's todo list !

coan.net: 100% agreement on all accounts.
I'd add two bits :
- Playing a forced move isn't playing either, at least it isn't for me.
- Installing a new program on my computer is already a barrier for me, even if it doesn't require any password.

mctrivia: Independently of the relative importance those four points have for me, I agree with you that they appear to be the most generally wanted, or at least the most consistently requested.

AbigailII, coan.net: I agree 100%. I would ask for the following generalization (if I didn't already ask) :

In all games, all relevant information that the rules make available to the players should be visible on the board as far as reasonably feasible. Keeping notes and tracking game scores are definite setbacks to the playing experience.

This includes the long awaited distinction between "race checkers" and "checkers on bar" in Cloning, Race and Crowded backgammon. There are probably other instances in games that I don't play.

mctrivia: LOL, never thought about that. So I did money laundering once in my life. I feel like a rebel now

mctrivia: I am no checkers players so I could be wrong, but to me it looks like there is a difference. When a queen takes, removing the checkers could allow for a loop which would be impossible otherwise, because there would be two consecutive checkers along the line.

For instance : queen on a1 against checkers on d2-c3-d6-b6 (please pardon me for the chess notation).
By removing the checker on c3 after taking it, the queen could take all pawns : a1-e5-c7-a5-e1.
If the checker on c3 stays, the queen is blocked on a5 by the two consecutive checkers c3 and d2.

AbigailII: any rating system worth its salt should not consider a match won 6-4 to be equivalent to a match won 10-0

That is wrong when cubed backgammon matches are concerned - and similarly might not be true in games that have more than two possible results (in chess, what about the player who offers draw in a winning position in order to ensure the match win?).

All the interest of cubed backgammon matches comes from maximizing one's chance to win the whole match, while not caring about what the exact score will be. For that a the "FIBS" rating system has been devised, which basically uses the Elo formula, but only after multiplicating the rating difference by the square root of the length of the match. That adjustment is said to be perfect for n-point matches with only 1-0 and 0-1 results, but I don't know why.

I have offered quite some time ago to help Fencer implementing the FIBS adjustment on Brainking (which is not a hard thing to do). Perhaps it must be said that Glicko, BKR and FIBS are all enhancements on the Elo system, and that the FIBS adjustement is compatible with either Glicko or BKR.

CryingLoser: I see your point, but I fully agree with Ceiter's and Abigail's objections. A common practice in challenges for top position is that the leader can choose the type of match ; if he chooses a 21-point match with 14 days per move, it is pretty equivalent to declining as far as a change in ranking positions is concerned. And one should not try to make rankings appear more meaningful that they are. They are not ladders, and they are not even always fair (the problem of multi-points matches has never been addressed). The #1 is not necessarily the strongest player.

Still, you are right that the tendency to sit on a top position (or whatever position) is bad for competition. To avoid that, I would suggest setting as default a ranking list of really active players - that is, lowering the 6 months limits. This does not exclude the keeping of a ranking list of less active players.
As for people keeping artificially active by playing clones, this should be forbidden by the user agreement and people doing that should be rating-banned. I feel that the playing of clones is mostly used for an artificial rating increase anyway.

There is still the possibility of playing only very low-rated people, keeping active at low risk. To avoid that, the answer would be the tried and true Glicko rating system, which keeps a more realistic measure of activity ; in that system, when the rating difference is high, the measure of activity increases by almost nothing.

Ceiter: You are right, it would be very useful. Using the numbers (24764) is another entry possibility.

MadMonkey: The variant where the player who can't play takes all the remaining stones is supposed to be a lot more balanced, and just as easy to implement. Once this is done, I have then nothing about a random variant - which might in fact be less balanced

AbigailII: As for which games are or aren't suitable on a turn-based site, I would say that you hit the nail on the head. Rephrasing, I would say that the board image should include all relevant information.

This is unfortunately not the case of the current implementation of Crowded / Cloning Backgammon (checkers on the bar that are undistinguishable from checkers on the 0 point, stacks of more than 15 checkers which use a 15-checkers image).

Logik is a great game, but pllllllease Fencer, could we have a piece set with the letters A to H instead of colours ? I thought I was the only one who hardly distinguished some of the colours, but when I asked about it in the Logik board, it sounded like most people had some problems too. If you don't do that, I will need new glasses very soon

As for why letters would be better than figures, imagine yourself thinking about "When I played three fours and two ones, I got one black clue and two white ones, so...".

The "like at home" board with coloured pegs you created should of course stay, but even here it would be a nice addition to use the title attribute of the image in order to display the name of the colour (resulting in a nice popup help in all good browsers).

MadMonkey: The three existing ones with a doubling cube - that's three new boat variants !

wetware: Same with me, and the use of a doubling cube would renew my interest in some of the "silly" luck games there are on this site - I thought Dice Chess was one of them, but I might be wrong as I never tested it :-)

AbigailII:The doubling cube is not meaningless when there are draws, one must just decide what to do with draws. I thought that we would simply discard them and count them as 0-0 (like in a n-wins match). Note that it doesn't make draws meaningless. The draw rate gives more incentive to take the cube.

It sounds also possible to count a draw as 1/2-1/2, subject to doubling (a doubled draw would be 1-1, etc.). This raises the question of what to do with the Crawford rule. I suggest that when the first players reaches 1 point or half a point away of the goal, none of the players may double in the next game (note that the leader could want to double if the position is a certain draw, because a doubled draw would win the match). That variant seems to require play-testing, so the simple variant should probably be retained.

dicepro: I like your idea a lot, but I don't think you fully realize what the doubling cube is about. It is NOT just a matter of rising the stakes. This is not a money site, so there are no stakes, just a match to be won. In backgammon, a match won by 21-0 is NEVER worth more than a match won by 21-20.

The point of the doubling cube is to force oneself to assess the positions in regard to the current score, in order to optimize one's chances to win the match. So if you are playing chess against a much stronger opponent, it is indeed a sensible, and certainly not abusive strategy to double at move 1 in all games (it is probably not an optimal strategy either, but I don't want to get too technical) . By making the match shorter, you will need less miracle wins.

Of course, that example shows a somewhat trivial use of the doubling cube, which means the cube is best used where the level difference is not too big. And of course in games with a random element and/or incomplete information.

dicepro: I second that 100% ! The doubling cube would enhance many games other than backgammon, and even if one thinks that there are games that don't play well with it, what the heck ? Just play those without the cube !

The doubling cube is especially interesting to use in games where there is a random element. My first thought is that it would be a bit weird to use the doubling cube in a game of incomplete information like Logik, but after a second thought, in fact it seems a great idea : in Logik, that would force to look at what the opponent is playing, thus bringing real interaction in the game (which for the moment is almost a double solo game).

Another good reason to do it is homogeneity. Rules that apply only to some games are harder to maintain.

pauloaguia: I support that 100%. Even when one is a black rook it is a bit frustrating to wait so long before the next round of a tournament when the last games have no importance.

joshi tm: We hardcore Scrabble players will probably prefer the official dictionaries :-). The French one is available for download, the others probably too.

MTC: I third that, but it has already been asked for many times to no avail.

For some time I thought I was the only colour-blind person here, but in fact there seems been a consensus on the Logik board for an alternate piece set which is not colour-based.

Numbers or letters written on circles have been suggested. Letters would probably be best, as numbers might get confused with the number of clues or the number of pegs of one colour ("three Bs" is easier to manage than "three 2s").

I like Logik, but with the present colours the game is barely playable for me. I am either destroying my eyes, either right-clicking on all grey and pink pegs, either both.

tonyh: I didn't say anything against your suggestion, actually I find it valid. Using the info as soon as it is available should be more accurate. Probably it is just more complicated to implement.

fakar10: You got me, I am too verbose :-)

MadMonkey:
>Maybe a solution to this is games Won or Lost by timing out should not be counted by BKR.

No way ! I know a lot of players who would deliberately time out in a lost position if it was the case.
I don't think there is any good solution to the massive time-out problem. My philosophy would be that your opponents were going to win their BKR points back soon if they were worth them. It sometimes happen that you play underrated opponents, but don't forget that it also happens that you play overrated opponents - e.g. the ones you timed out against :-)

rednaz23: That had me puzzled for some time too. That different games have different rating scales is absolutely normal, those depend on the skill differences (new games have few experts) and the "skill factor" of the game (a very skillful - and long - game will have bigger rating differences). But the average rating of all players should remain approximatively equal to the entry rating (1400 if I remember correctly).

What happens in backgammon is probably that many players have hundreds of started games when they suddenly leave the site and lose all games on time. As a result they fall to a ridiculous rating, but this rating disappears from the lists because they are not active any more. An average on all players, including the inactive ones, should be much closer to 1400 than the average you see on the ranking lists.

The ratings would be more realistic if they were shifted from time to time to center them around the 1400 value. But it is not really required if you keep in mind that absolute ratings do not necessarily have any meaning, only rating differences have.

wetware : It is not correct that rating differences are capped to 400 in the USCF=BK system.

AbigailII: I agree, as much as I am against using external help, I don't see any harm in using a dictionary, as long as it is an usual alphabetically ordered dictionary and not an anagram dictionary. Using the latter would take 50% of the skill off the game and should be banned in the same way as computer help is.

Scrabble is probably not in the top 100 of my wish list, I would play some if it was here though. I have not been too enthousiastic about the latest games added on Brainking. More and more games with a heavy luck factor, and some of the more interesting ones (Mancala, Recycle Chess) implemented with rules making them less interesting. In comparison, Scrabble would be a good addition.

coan.net: The official list of words for French Scrabble can be downloaded here : http://perso.orange.fr/ledefi/ODS4.TXT

I would think that the Scrabbles in different language should better be defined as different games, with different rating lists.

Of course playing only in English is feasible, but not very fair. My French vocabulary should be around 50000 words, while my English one is probably around 2000 words. I can probably still have some fun while playing in English, but considerably less. And if I play both in English and in French for the same rating list, I will screw the ratings.

By the way, competition French Scrabble is almost never played by the original rules, but almost always in the duplicated form : both players use the same letters on the same grid, and are aiming for the all-best score for the current turn, without any strategical consideration. Then either the best word of the two is placed on the grid, either the all-best word according to the computer. This variant is supposed to get rid of the big luck factor that "family Scrabble" (the name of the original game) has.

I suppose that duplicate Scrabble would be feasible on Brainking, using the no-computer rule for choosing the word to place, and having players search for their word in turn, without seeing the choice of the opponent.

Thad:

pauloaguia: If you could find where he said that, it would give some hope to all auto-play fans !

pauloaguia: 100% agree, but those would need what we called "auto-play", automatic play when there is only one legal move. It is logically perfectly equivalent to auto-pass, but psychologically different - which is probably the reason why it is not implemented here :-(

mangue:

rod03801: Ok, I am sorry that I started it, I did not remember that there was a Mancala board...

AbigailII: My feeling is reinforced by the mention in the same article that Oware has been solved and that perfect play leads... to a draw !

AbigailII: Yes, that is exactly what I quoted, and you are right that it reads as if when I can't make my opponent play (and only in that case), I take all remaining seeds. But that does not seem to make sense with the two previous paragraphs on the subject, which explain that you must always ensure that your opponent can play. At first I saw it as a clumsy phrasing, now it looks more than a mistake.

Maybe this Wiki text is indeed where the error (if it is one) comes from. I admit that it would require someone more expert than me to judge this.

AbigailII: And apart from the Nokia phone, did you ever see one of those instances where the player without move lost all remaining seeds ?

AbigailII: Hmm, I stand corrected, when I reread the rules of Oware, they are indeed not the same as our "Mancala". Does the latter indeed come from the Nokia phones ? That would be a real pity. And I insist that being unbalanced is a major defect in a game.
Probably I am a bit of a nitpicker on this, but I would also like if the game rules stated a reference of where they come from.

AbigailII: I disagree. Whatever the name chosen, if you take an existing game (that is known as balanced) and change a rule (making it unbalanced), then the rules are wrong by any definition that I can think of. Mancala games have a very long tradition, and I am sure that there are higher authorities than Nokia about them.
Actually, naming the game Mancala was a clever try by Fencer to avoid all discussion about the game rules. Indeed I wouldn't have dared to protest if the present rules didn't seem to make the game badly unblanced.

On a more general note, I think that we need a defined procedure for changing the rules of a game. It has been done in the past with new games that had an obvious defect (e.g. Cloning Gammon), but since then Fencer has been very reluctant to make changes, even when there were strong arguments for them (promotions in Recycle Chess come to my mind, but also suggested changes in some unbalanced games).

I am aware that changing game rules doesn't go without short-term problems (mainly, some people won't remember whether they are playing under the old or the new rules), but I think that on the long term the quality of the games offered here is much more important. After all, aren't the games what makes us come here ?

That White seems to have an overwhelming advantage in "Mancala" could be that the "no moves" rule was set backwards. Indeed, I always heard that when one player had no stones left in his side of the board, HE captured all of the opponent's pieces. In other word, it is of the player's responsibility to ensure that the opponent can always move.

I wikied Mancala, and although it is a wide game family, the rules played here seem to be the rules of "Oware" (http://en.wikipedia.org/wiki/Oware) . Here is the part of the rules that deal about having no moves :

<quote>
The proscription against capturing all an opponent's seeds is related to a more general idea, that one ought to make a move that allows the opponent to continue playing. If an opponent's houses are all empty, the current player must make a move that gives the opponent seeds. If no such move is possible, the current player captures all seeds in his/her own territory, ending the game.
</quote>

Not the most enlightening way to write it, but I think it means that the remaining seeds are due to the player who finds himself without a move.

pauloaguia: Hmm you're right ! And now that you make me realize that, I hope that I will be able resist the temptation to create a 21-pointer Cloning Gammon tournament with a big french cheese as prize

Your suggestion is probably close to the ideal way. Some programmation work, but definitely a bigger incentive to create a prize tournament. Too bad for the cheese !

harmy: I think that it was part of the idea that prizes weren't formatted. Maybe I want to create a tournament where the prize is a big french cheese (sent by boat).
Now, maybe there could a new free-style field for the prize, that could be displayed in the tournament list. For the moment, all we see is that the tournament is in bold, whether the prize is a Brain Rook or a Brain Bishop subscription.

El Cid: BBW's proposal is already very fine with me, I don't think we need to make it more complicated to implement (hence less likely to be implemented at all).

Infidel: Yes it helps, thanks, although it would still be better with numbers in the holes. Just a question of comfort, it is playable that way but not a nice playing feeling.