> I think that all queens should be automatically full (9p) queens.
The issue is that when there are only 2 players left, they can farm points by trading Q's. Normally 2-4 pawns will be present for them, so they can get 18+ points.
I don't think we will need that many changes, but I agree that variants rapid & bullet can be merged into a rating. At least chesscom doesn't have sth like "Crazyhouse Bullet" and "Crazyhouse Blitz".
Another problem with 4pc is that when u stalemate someone I get 20 points and the players with more points than me get even more even tough it was me who stalemated him
> I think that all queens should be automatically full (9p) queens.
The issue is that when there are only 2 players left, they can farm points by trading Q's. Normally 2-4 pawns will be present for them, so they can get 18+ points.
Yes, but why is it an issue? I mean, the 2 surviving players could indeed be given this advantage to be able to become 1st and 2nd if they have enough pawns and not enough points to counterbalance the 2 dead players points.
Another problem with 4pc is that when u stalemate someone I get 20 points and the players with more points than me get even more even tough it was me who stalemated him
YES, definitely stalemates should be thought into AGAIN. Something is wrong there.
I like the fact that a player gets +20 if he manages to be stalemated; but let's recognise, in 99% of cases it's done by chance and not really deliberately; I mean, we try to get stalemated, but we manage to do it rather thanks to lack of attention of another player and not because we've been smart ourselves.
1p queens have a reason, they are pawns after all. And seriously, rewarding points according to rating looks nonsense to me. If the double/triple check rewards is abandoned, which would IMO be reasonable, then there will be fixed points every game. The simple solution to opposite teaming in ffa is reward double points for capturing opposite pieces (40p for king/checkmate, 18 for queen etc) as long as there are 4 live players. Once a player is eliminated, capturing opposite pieces would yield standard points.
As for ratings, I agree that having too many is rather pointless. I suggest go with THREE: teams, ffa & solo, and variants.
The simple solution to opposite teaming in ffa is reward double points for capturing opposite pieces (40p for king/checkmate, 18 for queen etc) as long as there are 4 live players. Once a player is eliminated, capturing opposite pieces would yield standard points.
I thought about something like that (a premium for the opposite player pieces). The problem is though that it's pretty unclear WHERE to put the balance (impossible to find... there will always be people who will criticise, saying that either teaming is not discouraged enough or that it's too advantageous to attack in front).
But what I definitely like, it's to distinguish as for points between PHASES of the game: 4, 3 or 2 players, it's not the same.
And anyway I am certain that we shouldn't be given the same points for just ANY checkmate.
a) to checkmate a 2200 guy cannot be worth the same as for a 1400 player,
b) to checkmate "alone" cannot be worth as with a solution where you gang with the guy in front.
d) we should be given a premium when eating pieces from a stronger player. I.e. I propose here firmly to remove the dogma of "any bishop is worth 5, any knight is worth 3...". Typically, the normal amounts (1, 3, 5, 9) should be multiplied by a ratio calculated as follows: LOG (rating of the player who loses the piece) / LOG (rating of the player who eats). Typically, to eat even a pawn from a 2150 guy shouldn't be the same as eating a pawn from a 1350!
When a player with a high rating plays with players with a low rating, he risks so much he will go to a big minus in the rating if he loses (and if successful, he will get a very small plus to the rating). Essentially, your proposal encourages players to attack a strong player from the start until they finish him off. I think Sumat or Justin or Spencer will not like it and they will stop playing with low ratings.
And another suggestion: the server definitely shouldn't accept a game where there are 2 participants having let's say more than 100 points of difference in rating. Because otherwise it's totally a lottery. I can be just 1750, but with an opposite of 2100 I can win points just because he is good or just because my (and his) neighbours will rather attack him together because they know that it's their only chance to be 1st and 2nd...
d) we should be given a premium when eating pieces from a stronger player. I.e. I propose here firmly to remove the dogma of "any bishop is worth 5, any knight is worth 3...". Typically, the normal amounts (1, 3, 5, 9) should be multiplied by a ratio calculated as follows: LOG (rating of the player who loses the piece) / LOG (rating of the player who eats). Typically, to eat even a pawn from a 2150 guy shouldn't be the same as eating a pawn from a 1350!
When a player with a high rating plays with players with a low rating, he risks so much he will go to a big minus in the rating if he loses (and if successful, he will get a very small plus to the rating). Essentially, your proposal encourages players to attack a strong player from the start until they finish him off. I think Sumat or Justin or Spencer will not like it and they will stop playing with low ratings.
I understand what you mean, but I still believe it doesn't make sense to give the same amount of points for having checkmated (ideally alone) a much stronger player, or having eaten his queen for instance.
In addition: to correct the effect that you pointed out, I precisely proposed to use LOG rather than absolute value! So in fact the impact won't be huge neither, don't worry.
BTW, I've just understood, playing a game where I could claim win, but preferred not to as the position was interesting, that if we implement one suggestion made above (the one consisting to give MORE OR LESS rating points according to the score we have at the end of a 4p game, that will encourage people a) to really continue to play (as they need more points), b) to eat some pieces or pawns just to get more points (and later to have a better return), i.e. they won't just aim at the 1st place like now. It's not bad, that could make more interesting games. Basically I hate blitzkrieg 4p chess games, and I think longer games are more interesting.
I have played a lot of FFA games and have only ever seen 1 triple check, they deserved their 20 points!
Yes maybe. I've seen 2, and never done myself unfortunately. Could do once, and saw only 1 move too late 
I like the idea that someone proposed recently, to allow players to propose to the others "Shared points", basically at any moment of the game, 2, 3 or 4 players can stop the game PROVIDED ALL OF THEM AGREE and share the pot like in poker.
And I definitely think that the rating points as I explained above shouldn't only reflect our PLACE and depend on the ratings of the other players, but also on the points we got during the game and on what we did during the game (to eat a piece of a strong player or to mate him shouldn't give the same amount...).
What I propose, that's to merge back Solo and FFA; to give to the first the main amount of the points; and as for 2nd, 3rd and 4th, to give them points/take points from them, depending on their points during the game. Typically, it is not fair that sometimes you have a 2nd player who has for instance 63 points and the 3rd, 62, and the 2nd was simply offered this position by the 1st who wanted reward his opposite for being loyal. If I am followed, the 2nd and the 3rd would have significantly the same amount in rating.
That would make FFA REAL FFA, i.e. there is ONE winner (whose rating is increased by X, using his rating vs. the average rating of the other 3 players), and the 3 losers progress or regress HAVING IN SUM the same amount X, according to the points they got, vs. the average of the points BETWEEN THEM.
Example: we have 113, 63, 62, 3 points for the 4 players; and the 1st, because his rating was for instance 0.94 of the average rating of the other 3 players, progresses a lot, by 31.45 points for example.
The average between the 3 other players is: 42.67 points. Two players are above, one is below.
Therefore we give to the 2nd 5.6 points, to the 3rd 5.4 points, and the 4th is taken 42.45 points.
It becomes a null sum game (i.e. the average rating must be 1500 for all players, and if we implement that, we shall first have to adjust all ratings by a ratio in order to get to 1500, or I'd rather suggest in fact 1000, that would make the things very clear: you are either above or below the average...).
I am a former top 10 individual player (haven't played in the last 6 months) and I don't see a problem with the teaming issue. Forming a team (unofficially) is part of the strategy of a 4-player game. The rating system can be adjusted for any desired outcome, whether it is that the 3rd player gets more rating points than the 4th player, or whether 2nd, 3rd, and 4th are all equal.
The problem is forming teams based on past relationships, or especially based on player ratings. Weak players will often team up against other weak players and never against the highest rating.
The solution is quite simple - use anonymous mode and 960. Anonymous mode prevents anyone from randomly attacking a player because of some previous encounter. And if someone does kamikaze against someone for no reason, it is going to be reflected in the ratings, and since players can choose the rating range for the game, they can screen out weak players. But the top players (and for each other), the other top players are very identifiable even with anonymous mode because of the way the openings are played. Hence the reason for 960, which also has the benefit of showing true talent and not memorized openings.
I also disagree that FFA intrinsically gets the opposite side to team up. Yes, that has become the norm in games, but I never followed it and got angry remarks from other ppl because they perceived that there was some automatic team formed before the game even began based on position. Fighting players will find a way to win without subscribing to such artificial notions. Anonymous mitigates this problem too.
1p/9p queens is just part of the strategy! And making everything 9p changes the game in a very materialistic and one-sided way. It's like making stalemate a win in regular chess.
And having no ratings for all "variants" makes no sense, 4p 960 is superior to what is considered standard, it should be become the norm, and not have a rating for the regular 4p with the fixed opening position.
4p openings theory? Enlighten me. I don't like 960 in chess, but I can agree it has some validation there. In 4p chess, there is none. There is some reason why pieces are placed where they are. The teaming issue is IMO a problem, which can be solved quite easily: please give the opposite pieces double point value until there are 4 players active.
As for rating calculation, I would combine Solo and FFA in the following way: 1st gets most, then 2nd and 3rd both equally zero, and 4th loses what the 1st wins. That way donating points would be much lesser issue, as most often it just pumps 3rd to 2nd.
Riwwer, you cannot give/remove the same amount of points. Typically, if you have a very strong player A, 2 medium B and C and 1 bad D, the average rating can still be high, and if the guy B wins, whereas A is second, and D predictably 4th, you should still give A LOT to A (he managed to win against 3, of whom 1 was very strong), and you cannot remove that from D.
I mean, you are obliged to take ratings into account like now, it cannot be just the place.
Hi all,
After having read a lot of discussions and proposals, I decided to sum up what I think should be done in order to make 4p chess (other than Teams) more enjoyable and to correct some issues.
I think all of us agree that FFA pushes the opposite to team up, intrinsically. That will maybe be corrected partly with the new calculation system, but I'm not sure.
I thought a lot how we could address the issue and here are some proposals.
1) I think that all queens should be automatically full (9p) queens. Otherwise it pushes us to consider the 1p queens as rubbish that we can easily trade for a 9p queen, and it's abnormal and unfair. I mean, if someone managed to make a new queen, he shouldn't be able to say to himself, well, I don't care to throw it away and to weaken the neighbour, because he won't have any queen and will only get 1 point, whereas I shall still have mine and +9p. Unfair.
2) I think the surviving player (after all other have been checkmated or have resigned) should be given a clear avantage. It is not normal that he may (rather often) finish 2nd or 3rd just because the other players traded a lot. I suggest he gets an additional +20 like with a checkmate.
3) I agree that double and triple checks should be given a premium, but +5 and +20 were too much. Now it's not the case anymore with double/triple checks by a queen, but in fact I believe it should be let's say +3 and +10 for ANY case.
4) I see that since we have 1001 variants (and customisable games), there are less games and in addition I do not like the fact we have X different ratings. I propose a radical change:
a) we all have just TWO ratings (Individual and Team),
b) all Individual games, like previously, are FFA per construction, and become Solo after a certain level automatically (it used to be 1550 - i.e. 1850 as of today - and it was not bad at all), like that we shall again have more games and less particular rooms.
c) I am not against variants, but these shouldn't contribute to any ratings and should be here for fun. Like that we shall be able to say, "I am currently the 134th world 4p chess player in Individual", rather than having to say, "Well, it's just in FFA or Solo or XYZ variant"... Something clear, unique, public, and to be proud of!
5) as for calculations. Why is teaming so important? Because (smart) people know that it's pretty counterproductive to attack in front before at least one of the neighbours died and one should help the opposite, not weaken him, never ever. People are pushed into that by the fact that they must finish 1st or 2nd, and the only way to do it (when you play with people 1900+) is to team up.
To solve it, I propose the following (in addition to the suggestions above, I mean it's not a separate proposal, but a coherent system):
a) we decide once for all that to checkmate is more important than to eat pieces (so we give +30 for the checkmate);
b) we create an incentive to attack in front: a checkmate in front IF both (common) neighbours are still alive, would give +50;
c) a checkmate of a neighbour IF the opposite is still alive and IF it's done with a double check from both (i.e. typically: I check to my right with my queen which is hanged and can be eaten with a pawn, then the neighbour at my left does nothing and the opposite then checks with his own queen....) only gives +10, i.e. not 30 nor 50!
d) we should be given a premium when eating pieces from a stronger player. I.e. I propose here firmly to remove the dogma of "any bishop is worth 5, any knight is worth 3...". Typically, the normal amounts (1, 3, 5, 9) should be multiplied by a ratio calculated as follows: LOG (rating of the player who loses the piece) / LOG (rating of the player who eats). Typically, to eat even a pawn from a 2150 guy shouldn't be the same as eating a pawn from a 1350!
e) as for ratings, the way these evolve should depend on: the points we have at the end of the game and the average ratings involved.
Typically to win with 2 points of delta is not the same as with 35 !
So: games with all players above 1850 (as said above) should give the 1st player all the points, nothing to the 2nd (but he doesn't descend, or very slightly, if he was by far the strongest in rating beforehand), minus a few points for the 3rd and minus a lot for the 4th, weighting the ratings involved like now; games under 1850 being FFA way, with 2nd place meaning something (but more than now!).
f) to finish with, MAYBE we should also take into account when giving points for captured pieces, whether there were 4, 3 or 2 players. It should be more points in the first case I believe... (But I'm not sure; it's here for discussion).