To betray too early means 3rd place in 95% of cases

But these betraying guys throw me to fourth place and occupy third place himself.

Who cares about the 3rd place? That was precisely the point: to weaken/attack/kill the opp is totally stupid, you do not win like that, you're mostly 3rd or 2nd.

Yes, makes sense for some players, I know. 

BUT I also know that players who play to-avoid-to-be-4th do not win often. And therefore their rating tends to stagnate as they mainly finish 2nd or 3rd.

I like to betray to see people get 4th

Thank you for your sincerity Slayer. It's obviously the case of many other players as well, and it makes part of the human nature. My thread was rather for those who THINK that they are trying to win when they do that, whereas it's counterproductive in the 1st stage FFA.

Hey, I have a question

I think that people overestimate helping the opposite player. I think it is good to coordinate attacks, like one attacks one thing, other second thing, then some captures etc. But I completely don't understand why a lot of people plays to kill player on the side at any cost, like sacrificing pieces only to destroy one player. I am about 2350 and I always play the safest way possible in order to win in lategame and it often works, but sometimes players have problems that I didn't helped (I help only if it is profitable to me, if I have to lose too much material then I just prefer to save it for later). And I would not leave hanging pieces so my opposite player could capture it, you never know when someone "betrays". I think people should play this game defensively and not to kill player on the side at any cost.

The question is - what is the way to play, like plan for the game? I learned by myself.

That's what I properly HATE about games under 2400/2500. Have just had a game with an idiot with 2355 FFA (and very proud of this low rating), who betrayed in the 1st stage killing the opp (me) and sacked a very winning position, finishing of course 2nd and offering an easy win to a side. 

What I hate about that? Not being 4th, who cares. But I'm bitter to have lost 25-30 min. of my life because of a pure incompetence of someone, and to have had a boring game instead of something brilliant. 

Once I'm at 2400/2450, that tends to be extremely rare (if ever). But if I have a series of 3-4 losses and drop to 2300, then it's terrible because lower rated games I have to play, more numerous incompetent players are, who simply ignore the basics like not to weaken the opp in the beginning. (It's not that I rely on them for winning, god no, and the opp cooperation - natural at higher ratings - is not an absolute necessity; but at least one shouldn't attack the opp during the 1st stage). And more I drop, more common are cases again when I finish 4th because of that. I lost more than 250 points like that within some weeks, as I am 2239 FFA currently which is ridiculous.

(I should oblige myself to never play a game lower than 50-100 points under my rating of the moment, and I should play with any player under 2400 as if he were to betray at any moment... It's a pity because it deprives us from many many interesting moves but that are risky with an imbеcile in front if he may eat your queen or checkmate you).

Some ideas are here for instance: https://www.chess.com/clubs/forum/view/basic-ffa-aspects

Because that is the best strategy (empirically proven for more than 3 years now).

Amigo, it's 100% your right to have your own ideas on what the strategy should be, etc.

But believe me, most high-rated players (2500+) won't ever trust you anymore later if like you did, you kill them and make them 4th (finishing yourself 2nd and throwing the game to a side) after they've sacrificed their queen to save you. In fact 80% of them would simply ban/block you immediately then, in order to avoid having you again at the board, and that's why some players are condemned to play anon. games only, as too many players have blocked them here. One doesn't have to help necessarily, etc. but to kill/attack/weaken/fail to save the opposite is seen by most top players as sign of stupidity/incompetence.

1. Many come here for the sense of doing well or seeming good that they can't get in normal chess. Everyone thinking for themselves would increase the odds of the best player winning too much. You will find in most top 20 lists that only about 5 of the players has any kind of decent normal chess ratings.

2. It is a confusing game. Starting with having three enemies are quite overwhelming and it would be too paralyzing to really think that way, it wouldn't just be defensive play, it would be nobody doing anything at all.

3. People get emotional, you might take a pawn or just aim a piece towards someone and they take it as an act of full war. In a sense it good to expect the flanks to mean you harm while hope the opposite will not.

4. It balances personal feuds a little. If everyone so passive nothing happens and the players get to pick any target, the odds are they go for someone they don't like or base it on rating. No matter if it swings towards hope to kill the lowest rated easy, or make sure the highest rated player has a hard game, it not very good or fair either.

With that said I agree, people play much to much by a script. Ruin their own game and then blackmailing the remaining two players into letting them rebuild, because the 2nd strongest player not just going to give the win away.

That it more about the players agreeing on a way to play than that their would be a right way is also the reason you almost only see references to "top players" or a rating like "2500+" instead of names off players.

Most players have huge swings in their rating, if we had players we knew where good, we wouldn't care what rating they had. The best players would still be the best players and people would agree on whom they are. But again, instead we have the reference to the ranking list or a rating interval with the suggestion they agree with each other and such.

I agree with some of the assertions above.

But here I would like only to react on one aspect right now: I mentioned 2500+ just like a simple benchmark, because I notice that above most players won't play bullsh1t. AND contrary to what you say, despite of the swings all of us have (for instance, over the last 12 months in normal Rapid FFA I've been between 2200 and 2550...), in the long run and in average, strong players still are always in Top 50 for instance, and those of the Top 20 tend to remain there for years. Typically, Arseny, Hest, ChaCha, Spaksi, SirMullih, Neo, Riba, Jbolea, Rojitto, and sorry to those whom I forget here, are there whatever happens, even if their ratings as well vary from time to time, sometimes significantly.

And yet another case. A guy with 2389 who kills me in front (and of course finishes 2nd; in most cases it's impossible to win like that). Normally with 2389 one knows that to kill in front in the 1st stage FFA is counterproductive. But in this particular case I know that it was more important for him to make me 4th rather than to win. It's sad some people react like that, and it makes sometimes games toxic.

Just stealing the checkmate is quite common.

You haven't hurt your opponent, but when they about to checkmated you take the checkmate.

Another way to cash in on your opposite without hurting them yourself is to defend their pieces. Instead of attacking or sacrifice on the flanks to help your opposite you just defend the opposites pieces. Since this is a slow kind of help that will not really stop the death of the opposite would the flanks decide to team on them, they will probably be unhappy with you. But you did help and lost nothing yourself.

However, with your opposite dead, you with most pieces and perhaps even points, the odds are high the flanks keep teaming to bring you down, so it very poor diplomatically.

The highest rated player with "poor team-play" is probably "Illingworth" which tend to spend time on create a pawn-shield with all pieces in the background, making it hard to target anything else than the pawns. And in so doing, give the flanks a very slow and boring structure to attack. That then leads to many flanks deciding to race their pawns to promotion instead of attacking. When both opponents begin a pawn-race it tend to create a small war, where Illingworth can move the pawn-shield a square or two forward while the flanks try to promote or stop the other flank from promote pawns.

In the sense of letting your opposite die and then expect to win the 3-player stage, that seems to be the method of best score. But I do believe you need to be a good 1-1 player to pull it off.

Sincerely, after 11000+ games and "some" thought on that, for me there is no place for polemics here, the matter is utterly clear:

- in FFA, in the 1st stage to weaken/attack/kill your opp, and to fail to save him when you can, and to fail to release pressure from him when you can, is simply stupid, counterproductive,

- in the 2nd stage it's totally over, there are no friends anymore,

- to kill your opp in the 1st stage is only acceptable and rational and normal if a) he's under strong joint attack from sides, b) you can't save him anymore, really, even sacrificing your Q if necessary, c) he is sure to be mated IMMEDIATELY, NEXT MOVE, BY A SIDE. In this case to take 20 (rather than leaving it to sides) is pretty normal. But that's the only case...

As already said above, when you see that the opp doesn't know anything about the NORMAL structure of a FFA game, plays nonsense, weakens you (...and him), attacks you, etc., to attack him (trying to avoid at least being 4th) is quite acceptable as the final outcome may well be 1st, 2nd, 3rd, 4th, whereas doing nothing or continuing naively to cooperate with an idiot usually means 4th place, so the expectancy is for the 1st scenario clearly.

Currently the record rating I've seen for someone to attack in front from start and to finish (as in 95% of cases) 3rd, it's 2342.

https://www.chess.com/4-player-chess?g=10471130

Let's see who will break this record.

(2 years ago, it was inimaginable to see it from a 2100+ for instance... The rating inflation also has this consequence: 2300 people playing like 1900 were 3 years ago).

Terrible stats. 3 years ago 1900 was top 20. Todays 2300 would be more like 1600 year 2018.

Yeah.

Part of the explanation is that people come, do 2-3 games quite often (or even just one), offer some points, and then never come back. I.e. the overall system is fed with extra points, continuously.

But still, I don't understand why we don't see such inflation in 2p chess.

Here the Top 20 players will soon be 4000.

The top 20 players will never be 4000+, also 2pc uses a different rating system so comparing the two ratings as if they are calculated the same isnt really fair.