Possible ratings correction to stop FFA teaming

That sounds good. It really can change 3 players stage.

I think the first is better, but you can pick up the function = 0.5 for the +1850 rating (the conditional boundary is when people prefer team strategies to solo, but I could be wrong).

sorry i deleted again cause still in discussion with Bab.

what do you think the X should be for 1800, 2000, 2200 average?

i think it's shouldn't be so drastic. if it's not effective enough we can always make X smaller. But I suggest small steps, also to keep it more different from Solo

18050: 0.5

2000: 0.3

2200: 0.15

?

?

Yes, but I'm not sure that +1850 is exactly the border that I spoke about. What was the +1850 (+1550) border based on when the FFA games became games with a Solo rating? The maximum then seemed to be +2200 (+1900). (2200 + 1500) / 2 = 1850. Or am I wrong? Perhaps now you can even make x = 0.5 at +1900, since Cha_Cha reached +2300.

i agree that the teaming only starts around 1850. that said you can expect it to get lower over time as lower rated players catch on.

I think we should start with a mild version for two reasons: 1) not change things too drastically, first see what a mild version will accomplish, we can make it more drastic any time. 2) keep it more different from Solo. (The alternative would be to just get rid of Solo and make the curve much more drastic, and also maybe make it go towards 4 -1.3 -1.3 -1.3 rather than 3 0 0 -3. I don't see the point of having two modes that are essentially the same)

Contrary to me @BabYagun is in favor of a much more drastic curve.

What is your opinion (all of you)?

My mild suggestions would be something like this:

aka at 1900 average it's half the points for 2nd/3rd, at 2000 it's a third. For those fighting for the leaderboard the changes for 2nd and 3rd will be 5 to 10 times smaller than currently.

At 1800 we have 0.7.  most 1800s aren't teaming like crazy at all from what I can see, why change he game for them?

yes. to stop the continued teaming 3-way. 4 way opposite cooperation is pretty standard even in Solo. At least to some extent.

I think this will also limit the cooperation 4way, because you want to maintain chances for 1st and can't rely on opposite giving you 2nd even if you're "dead lost". And also: You really don't want to end up 4th in this game, so that will probably make play much more cautious/conservative overall.

> Contrary to me @BabYagun is in favor of a much more drastic curve.

Well, I proposed a straight line. But mathematically a line is just a variant of a curve, just ironed.

I like what you are doing @spacebar!  Justin fears people will be able to get inflated ratings with this set up.  I don't know, I have thought the same thing, but at least this ratings set up will try to address the bigger issue in the 3 player stage which is the concern of a lot more players.  @BabYagun's solution is still too dramatic for me!

@jb9656, right now we want to reduce/soften excessive teaming in games of higher rated players. "Excessive teaming" in this particular case (modern 4PC FFA) means that top FFA games have the same scenario in almost 100% cases:

Opposite players play like a team from the very beginning of a game to its end. They coordinate attacks against side players and don't capture pieces of a "teammate" (aka "loyal opposite") until they kick off 2 other players from the board. Then one of them gifts points to another one (if needed), and they finish as 1st and 2nd.

Free-For-All should have an "Everyone vs Everyone" meaning and spirit. There should be different scenarios. And "2 opposites team up" may still be one of those scenarios, but not the only possible/popular one.

> What's even the goal? ... Make it tempting to betray a teammate?

Your question confirms what I said above. "A teammate" is not a word from FFA universe. There are no teammates in "Everyone vs Everyone" by definition. If there are teammates then either the rules or the game name should be changed. There is 4PC Teams mode, we don't need to have 2 of them. So, we don't rename FFA to FFT (Free For Teams), we change the rules.

> @BabYagun's solution is still too dramatic for me!

Thank you for your feedback. This is the current FFA Leaderboard:

As you can see, today it starts from 2109. I believe that players on this list (FFA Top 20) should get there only if they earn rating for the 1st place in Top FFA Games. It should not be possible to get 2nd place in many games, slowly slowly earn rating this way and appear in Top 20.

If a player is in Top 20 he/she is our exemplar, beacon and favorite. He/she is an example to follow. Should we follow a guy who plays for 2nd place? Meh.

I vote for pure 3 0 0 -3 for 2000 (or 2050) players. Not 3 0.3 -0.3 -3. Just give 0 for 2nd place. Want to be in Top 20? Go get there by playing for the 1st place.

This is how I see it.

I believe that the 2 place should be encouraged, otherwise, in my opinion, it will be Solo with other rules in which the entire severity of the defeat falls on the 4th place, which in my opinion is bad, as it can make the game very passive at the beginning. Therefore, I like spacebar ideas more.

Elo hungry. You should be more worried about the team players coming into ffa and literally taking over. I don't even know if hest will survive this. He is ok at teaming but not as good as Kujtimi, icystun, valger, hamanda, ect.ect.ect. Stay focused on what is going on, then the FFA will grow fantastically. I've seen players in earn ratings that were very high in 2v2 chess. Simply by playing against noob players over and over. That's not real ratings. I think that's called "abusing", or something. I've done it, but then stopped, cause it don't really represent your skill anymore just shows how one can gain elo in a very pathetic way. NO. We must all be competitive and not worry about the elo but rather the game. Then we can see some verrrrry interesting games which we can ALL learn from. Some rely on deception, others laugh at it. I wonder why....

> I believe that the 2 place should be encouraged

Yes, but only in a real FFA (all vs all), not in a "2 vs 2, with points for captured pieces".

I still hope that one day we'll find a way to make it a real FFA.

I agree that -3 for 4th place only is a bad outcome., The game is unlikely to change at 4 player phase, you still need to be on a team to avoid getting the -3. The game needs a bigger shake up than this to actually change the game play at 4 player phase. My suggested ratings tweak at the start of this thread successfully achieves that.

It will be very good, it will improve the quality of the game in the first stage 2 vs 2.

yeah 2v2 games isn't easy, it's real chess. So many variables to take into consideration, that we have diduseethatcomeing laughing at the players who only play standard, and have little idea what to do against his tactics. Fundamental education is more valuable then learning theory. Learn the fundamentals, then one can glance at the theory with cunning eye's so to speak. Yeah however second place is what is really encouraging this kind of "teaming" that is going on. Hamanda just makes sure he gets second place then resigns...... That's pretty funny, and pathetic that people would want there to be a second place so much then complain about the teaming that is going on. hahahahaa. I love it but, I want the see something done about this. Everything that is going on in FFA is only natural. We must understand this.

Personally I want to see the game impact 4 player, not just 3 player. The person out first is normally a victim of a team, not necessarily the weakest player, so punishment should be less

That would be getting very close to Solo. Fwiw even in Solo opposites tend to cooperate 4-way, so I think to tackle the 4way teaming a greater change of concept is needed.

I've added a variant (see test-server) called "Play-4-Mate":

It's an old idea which I want to have people try out.

In its current form, there are total 105 points to be earned per game. The first to checkmate a player gets 60 points and will win. Rating changes will depend on how far above or below the average of 26.25 points (105/4) you end up.

Whoever gets the first checkmate is in the lead for sure. And if the other 2 gets the other checkmates, then the player who gets the first one is the winner. 45  if another player gets 2. The guy still wins. I don't think that works. Unless there is something I'm missing. Why not try a last man standing version. See how that works. Just as an experiment.