A clear example of why the FFA rating system should change

Please observe this game: https://www.chess.com/variants/4-player-chess/game/27661510/186/1.

In my opinion, Blue played well this game and found some key moves. For example, consider this position: What would you play as Blue and why? I think there is only one "best" move.

Blue find Rxl6! if Yellow cooperates, then either Green loses the exchange while creating a vulnerable knight or his pawn structure becomes destroyed.

In fact, Yellow does find this idea.

This signals to Blue that Yellow is at least aware of tactics and position to some degree.

Now, consider the following position in this game. What would you play as Green and why? Again, I think there is only one best move. Furthermore, I think there is a best continuation for Blue.

Green has Qd12+, which then allows Blue to take Yellow's 9-point queen.

Blue doesn't play the hasty Ne6, which only trades a knight for a 1-point queen, but rather finds the best continuation.

Although Yellow's position has been greatly weakened, he still has fighting chances in the three-player stage.

Can you guess what Yellow plays next? He resigns, which more or less hands the win to Green.

Maybe Blue could have given up the knight to take the 1-point queen because he didn't want Yellow to resign in this foolish way. But that would result in less beautiful and sharp play. Blue played accurately and provided entertainment for his audience by finding the sharp move. And in return, Yellow (perhaps out of spite or the unwillingness to accept his reality) simply resigned. In other words, the current rating system allows weakness to have power over strength because Yellow gets to decide in this position if he wants to take Blue down with him by giving up. This is not an isolated example, but one example of MANY games. Too many players are simply too weak to allow the current rating system to be effective because they do not appreciate the beautiful complexity of the three player stage but simply opt to make easy decisions like Yellow did this game.

I completely agree with your sentiment that is was weak for yellow to resign in the last position. The problem is that blue's play is just completely disregarding the rules psychology has in this game. In a championship game their play is perfectly acceptable but in a regular game with average players they are playing in an egregious manner because I fail to see yellow in any rating range under 2700 take their play as neutral. Yes, we like to think that players will play for the integrity of the game and play on, but it's just simply not so in most cases unfortunately. 

You've just described why I no longer play Rapid "FFA" anymore, almost all my games have gone down like this, where one player throws for no reason, or 2 players attack me for being higher rated than them, instead of keeping the balance.

I've had games where opposites are discussing who to attack, and even speaking other languages during the game discussing moves.

I also find that the quality and level of play has drastically dropped since the merge, since so many good players no longer play Rapid FFA.

When we're stuck in a game with players that have no idea or plan to keep the balance, we at least didn't lose rating by avoiding 4th.

Why can't we go back to the FFA system that worked and we all loved?  Get rid of SFA and watch the player base return.

My point is that Blue was punished for playing well because the rating system allows one to punish others. Beautiful chess play should not be hindered by psychological weakness. If Yellow wants to resign, then let him do so at his own detriment. This is a reason why I am considering that not only +3 +0 +0 -3 is a good option, but maybe even +3 +0 -1 -3.

I don't think that players are helpless to opposite defection. While it's not easy to compete against this, I believe that knowing when to betray is an art in itself for FFA with a 1 or 2 loser system.

blablablablabla..............

Always contesting rating story...

⅔ of publications concern change rating, error of points, ...

If someone should win 1 or 2 points less than he should win, this cause a scandal 🙄. Is not drammatic!😉

Simple idea: switch from the WTA system to the FFA system as you get higher rated, rather than FFA to WTA. I have a feeling this might make lower-rated players happy since they realize they are playing for first and so teaming may actually decrease their chances of winning. After a certain point, say 2200+ (where I've seen people actually understand the idea of opposite cooperation well), we can make the slow switch to FFA. However, I'd probably make 2nd gain very little and 3rd lose a lot at this level since it's not yet at the high level where people are expected to play almost perfectly. 4th will definitely lose the most points, but this might be a bit unfair especially if they get teamed on and their opposite does nothing to help them, rather, plays passively. So maybe we might want to make the rating changes more lenient toward 4th and more aggressive toward 3rd. In fact, I'm thinking that in most cases, someone gets 3rd because they simply played badly. There is no real luck factor involved (unlike in the 4p stage, where you could get teamed on without any help from your opposite). So we might even consider making 3rd lose the most points rather than 4th. Not very natural, but to me it makes sense. Obviously we cannot compensate for all the other scenarios such as where someone got 4th simply because they played badly, but this will probably be made up for (if only a little) during the switch from the WTA system to the FFA system. Again, it's just an idea, and maybe it might work out better for everyone.

@Grathieben
> I completely agree with your sentiment that is was weak for yellow to resign in the last position. The problem is that blue's play is just completely disregarding the rules psychology has in this game. / absolutely correct, It is what I have mentioned in several posts, the psychological factor is an existing reality in the 4pc, that the best players in the world are aware of, The temperament or state of mind of the player can be highly variable during the battle, but this data , is not measurable mathematically. We can only distinguish and/or perceive it during the game.

@liquid-sun my proposal is very similar to what you indicate here: +3 +0 +0 -3 is a good option, but maybe even +3 +0 -1 -3.

I would like to see how this system works in practice. I know very well that in the past when the second place won points, it did not work well and I agree.

1st: +2 wins | 2nd: 0 draw | 3rd: -1 losses | 4th: -1 losses

- But in this case, the second place does not win anything, and I insist, players from 2500 upwards, do not play to be second place, I do not know any, everyone seeks to be first with their best strength and knowledge. I have a list, and all or most of them are serious players.

- The third and the fourth lose equally, therefore we will not see passive players in the 4-player stage, they will seek to be a strategic alliance (not TEAM). From the beginning, and this is the most attractive and exciting of all.

- The winner gets a +2 and we avoid inflation.

- Finally applying the leveling improvement on ELO losses that the developers made, for the risk issue for high rated players, It will be perfect for third and fourth place.

By the way, this Approach was very well received in the polls, people liked this ranking system model: 1st: +2 wins | 2nd: 0 draw | 3rd: -1 losses | 4th: -1 losses

With this formula we manage to deliver the solutions and the audience will notice the change.
if you want to scale it to 1st: +4 wins | 2nd: 0 draw | 3rd: -2 losses | 4th: -2 losses 

its ok, They are all FFA models in essence, and we all played it with great pleasure.

If I must also say that the last improvement made to the SOLO rating system was very positive for high ELO players, since today a 2900 player playing in a 2500 queue does not lose as much ELO and with that we eliminate the risk, and today it can be seen that all the losers lose a very similar amount between them, and we must emphasize that as something positive.

Still, if you ask me, I still prefer the FFA rating system in essence, as it is graphed by @liquid-sun in this game.

+3 +1 -1 -3 for under 2000 average. 2000-2600 +3 +0 -0 -3. Over 2600 +3 -1 -1 -1

In my experience, even 2600+ will ruin games. I think it has less to do with rating and more to do with principle / theory.

There are many more games when players with this rating play teams with his oppo until end no meter what ( read revenge or loyalty ) in 3 players stage or with no idea what to do except ruin somebody game etc. - knowing that 2. or 3. place is 0 points ( like before ). And you want to return to this madness? Plus inflation of points which led us to abnormal ratings over 3000! Pure and clear proof that the ratings in the old ffa were unrealistic is that there are players with a rating of 2500 who have no idea what to do in the 3 players stage. 

Before the merge, I only played solo and together with a few other strong players we fought for the first place - the result is that, besides, we've played a lot of games, we were all around 2700 rating. After so many years, SOLO only reached 2700! And FFA after only a year and a half came to 3300! In another year, they would reach 4,000, maybe 4,500 ratings! Is that normal for you?

I know what you're going to say - it's because SOLO was played less! Absolutely incorrect! It was played the same as before, but with one important condition - no lower limit! Which, in particular, led me to have a series of 13 wins and ONE defeat and to advance on the leaderbord by only about 40 points!

And how did the FFA players get their rating? Because the MAIN OBJECTIVE of the FFA is not to be 4. !!! 

When that goal is fulfilled then relaxed - if you are the first ok, if you get bored of the game - you don't care, you don't lose anything (only the one who wants to really fight wastes time, but you are selfish enough and you give a s...t for him). To have a good rating in the old system, it is enough to be a good team player and then - it doesn't matter! 

Sorry guys but this is SOLO and not Teams!

Let me ask you something, you who want to return to the old system: how many times have you played a game where in 3 players stage someone goes after you like a kamikaze just because you ate his bishop? Or because he dont like your stile or something more stupid? He can do it without any consequences! And do you know why? Because he loses nothing if he is second or third and loves to spoil the game for others! And you want to make even more such players!?! 

Well done! Great suggestion (if you want to destroy this game completely)!

One more thing in the end - I think it is completely unfair that, after the merge, the FFA rating is recognized as an official and not the previous SOLO rating. Why? Because, we can all agree on this, this FFA is SOLO now, ( rules ). When I found out that there would be merge and that the FFA scoring mode would be abolished, I intentionally did not want to play FFA because, in my opinion, it would be unfair to brag about the rating, which was won under other game conditions! I am sure that my rating would have been 200, 300 points higher if I had played only FFA intensively in those months. ( Of course, because of points inflation ). But let me repeat, it is simply not fair!

( the other reason is that I haven't started from the "bottom" of the list for a long time, and tried to climb to the top. This was an ideal opportunity to finally win twice as much for win as when I lost and not just 4 or 5 points as before! )

Anyway, very soon everyone on the leaderboard will be where they deserve so I don’t complain much about it. That is why I will not mention this unfair situation again.

Your Neo, SOLO warrior forever!

This is exactly why I want the old system. Players kamikaze or blunder regardless of whether they lose rating points for it. Look at rated games of any level. Players throw games. 2600 rated players throw on purpose. 2900 players throw accidentally. Maybe at 3100 Solo works. But by that level, players wouldn't throw in FFA either.

To echo this thought, if you read the logic in the original post, you will see that I provide a clear example of this very thing. Although Yellow wasn't *entirely* mindless, he did appear to resign out of spite.

The VERY FIRST game I played in the new rating system, an idiot targeted me when the third player was clearly the strongest. Same with the very second game I played in the new rating system. From 2720 + to 2680 + in only TWO GAMES due to suicidal play that was outside of my control. The only two games I will ever have played in this system.

@neoserbian You will never convince me until you can counter the evidence AND logic (i.e., what follows from Typewriter44's premise that people kamikaze despite potentially losing rating) I have presented to show that suicidal play by someone who throws the game is best handled in a non-SOLO rating system.

You forget one very important thing (which is why SOLO works) - players who play that way (like kamikazes) will very quickly be below 2600 ratings! SOLO sanctions such play! The FFA promotes that idea and support players to do that because there is no penalty for them and they are still at 2600! 

There is a big difference between players who play to win, learn tactics on how to reach balance and win in the end and those who play only for fun, they don't want to learn anything, all they care about is not losing their rating!

The first group will progress quickly and develop in the SOLO rating system. The second group will very quickly descend to their level - amateurs. And it will no longer be a problem for players who take this game seriously!

In the FFA rating system, such separation (serious from amateurs) can never happen because players will never be punished for their bad play.

That is the biggest difference and the biggest advantage of the SOLO rating system - 4 players chess has no purpose or future if everyone does not fight for the first place!

If there are still chess players who are in the top 50 and play only to avoid the last place (main goal) then this game becomes a mockery and will never develop to the level it should be!

Neo, don't kid yourself in thinking you'd be 2900-3000 if you had played FFA instead of solo. You have major gaps and flaws in your game that need to be sorted before you get to that level.

Radon, do you want to bet? For your information, I have been in the top 20 solo players in the world for 3 and a half years constantly, and most of the time I am among the top five (several times the number 1)(Of course I don't count the current Leaderboard which is a ffa Lead.).. I am one of the three players who were the only ones to participate in all 4 solo championships (I was second in one). And you claim that I could not reach that rating in a game that is much easier than solo?

Ask Michael how easy it is to reach that "top level" - the man is a solo player like me and, for joke, he started playing ffa and in a couple of months he improved his rating by 200 points! Just ask him!

a) I like the current system for many reasons (encourages people to play for 1st; discourages perverse behaviours like playing for 2nd or playing-to-avoid-to-be-4th, etc.). 

b) something should be done CLEARLY as for players who resign (I believe that one should never resign in 4p chess as it affects the result for other players, and so many times I've been deprived of a hardly obtained victory just because someone resigns before CERTAIN +20 or +9 pts for me. So in order to punish that, we should for instance decide that a player who resigns, automatically becomes 4th regardless of his points. BUT: the problem with that approach is that players would simply let their time finish. So it's not easy.

a) 0 points for 2nd is not good: still encourages playing for 2nd in many positions (once someone understand he wouldn't win eventually)

b) your both ideas punish the 4th far too much, and it encourages people to play-to-avoid-being-4th, that's NOT GOOD and lowers the quality of the games. Whereas you may be 4th sometimes purely because you have a passive player in front or even an idiot who attacks you in the 1st stage FFA. Not fair to lose that much points in these cases.

c) your second idea doesn't work; would lower the overall ratings in the long run (because it makes -1/4 every time as for math. expectancy). But why not after all, in the same time, as we know that the ratings here tend to increase because of the "fresh meat" that comes, loses points to the common capital and never turns back... Maybe your proposal could help to correct this effect after all, I don't know. The points' inflation is a big problem definitely.

The 2nd shouldn't be neutral, it has to be negative. The 2nd player is in fact the MAIN loser usually. The 4th shouldn't be punished a lot. The difference between 2nd and 3rd and the one between 3rd and 4th shouldn't be huge.

I like the current solution, rather a good compromise. Otherwise, we could implement something like +4,5  -1  -1,5  -2, or +3 -0,5  -1  -1,5.

Yes I saw; I explained above why for the 2nd it shouldn't be 0, but something at least slightly negative. As for -1 -1, I definitely like it.

Maybe we could do +4, -1, -1,5, -1,5. Not as abrupt as 3 -1 -1 -1 that many don't like; but still a good compromise allowing to avoid many of the potential issues of many rating schemes.

I think you do have a point that in the old system, there were weak FFA players with inflated ratings; although, this is another topic. This is why I think that +3 0 -1 -3 may be an option: -1 for third place could punish players who play "for fun" as you say. Although this could result in a player playing "for second" by continuing to team, if the other player is sharp, then this player could be forced into third place. I have used this particular strategy to deal with foolhardy teaming effectively in most cases. Maybe even +3 +1 -1 -3 could work if the not-fourth player remains sharp.

I think your logic does not address, however, how the SOLO system dissuades sharp and accurate play, as clearly observed in the game shared in this thread. Clearly, the tactics of Blue and Green here were not just "for fun," but based on clear strategic logic to seek to attain first. As observed in this game, such sharp lines rarely work in the current SOLO rating system because players like Yellow do not have the discipline to move forwards and accept their compromised positions, which thus make the game much duller, lest those who try such sharp and accurate play are punished for playing the game well. This specific limitation of the SOLO system is what I'm addressing and is a slightly different topic from your ideas.