The optimal 2nd and 3rd scoring system in non-solo FFA

Sort:
liquid-sun

UPDATE: The following ideas are meant to generate a discussion and not to suggest an immediate change.

To my knowledge, most agree that the optimal FFA scoring system has a discrepancy between 2nd and 3rd, and 4th. If we are to follow through with this system, I think the next question is what the optimal values for 2nd and 3rd are.

I don't think that the answer to this question is obvious given multiple pros and cons. So far, I have seen the following suggestions:

2nd 3rd

+0 / +0

+0 / -1

+1 / +0

+1 / -1

-1 / -1 (UPDATE)

 

Potential cons of +1 / +0: players playing for second place when they could play for 1st

UPDATE: Will this cause players to team to the end? As I state in a later post, I wonder if giving 2nd a + may, ironically, prevent teaming by allowing not only the strategy to play for second, but also allow the strategy to punish one who plays for 2nd into 3rd (I sincerely do not know if this alternative is true and do not mean to suggest that it is. I just want to raise it as a consideration).

Potential pros of +1 / +0: the option to play for second place as a means of punishing others playing for 2nd

These cons and pros also exist for +0 / -1, but may further dissuade playing for second place. Still, one could choose to play for 2nd out of spite for or fear of 3rd in such a setup.

Potential cons of +0 / +0: constant shifts in 2 vs 1 alliances in the 3-player stage, which may result in duller play.

UPDATE: At the same time, +0 / +0 may also result in more interesting FFA play if players are crafty.

Potential pros of +0 / +0: the benefits of playing for 2nd place are nullified

Potential cons of +1 / -1: playing for 2nd place may become a viable strategy

Potential pros of +1 / -1: the viability of playing for 2nd place may encourage players to prepare for safeguarding against 3rd place during the 4-player stage (and may thus reduce "strong" teaming with sacrificial play against a player with a passive opposite)

 

Potential cons of -1 / -1: As stated, I think this leads to duller play and prevents the capacity to punish foolhardiness when one plays for 2nd.

Potential pros of -1 / -1: Others believe that this may help to prevent foolhardy teaming.

 

What are your thoughts?

ProfBlundermaster

Hi Sic Vic, I really wish you were one of the people part of the long-winded Admin debate.

I proposed the following reasonable options:

+2 -0.5 -0.5 -1

+2 -0.4 -0.4 -1.2

+3 -0.5 -0.5 -2

+3 -0.6 -0.6 -1.8

 

For all of the options above:

1) Notice how: 1st > (2nd = 3rd) > 4th
2) One needs to study 4PC FFA Opening Theory, and play it well, to survive the 4P Stage.

3) One needs to understand material balance in the 3P Stage... and playing for 2nd is not an option.

 

What are your thoughts?

Typewriter44
liquid-sun wrote:

Potential cons of +0 / +0: constant shifts in 2 vs 1 alliances in the 3-player stage, which may result in duller play

How does this result in duller play? 

spacebar

I think we should consider other means to improve FFA, and not only consider rating systems. For example, how would 10 point checkmate affect the game?

Would 3 player stages be less dull, if one has to be much more careful about not allowing point leads (rather than not allowing/blundering mates)?

At_d0sA_fNLt_Laris

Rating system is the most important thing to, how high level queues and farming works.  Checkmate+10 or Stalemate being +7 +7 +7 / +10 +10 or +3 +3 +3 / +5 +5 doesn't really matter. If you combine it with promotion 11th it only makes the game longer. Basically discouraging players to push pawns and shuffle for life, back to 100 moves maneuver. Also, if it's played in Solo format, it discourages attacking, 3 player stage cooperation and winning it, takes luck rather than skill, idk about that, but rating system affects this game the most imo because of those 2 things: 1) playing for 2nd  2) just playing not to get 4th, even with the cost of ending up 3rd every game 

Darksquareman

How would constant shifts in alliances make for duller play? I think that’s the most exciting part of 4pc.

LosChess
spacebar wrote:

I think we should consider other means to improve FFA, and not only consider rating systems. For example, how would 10 point checkmate affect the game?

Would 3 player stages be less dull, if one has to be much more careful about not allowing point leads (rather than not allowing/blundering mates)?

Please, under no circumstances make the King +10, the main issue we the players have is the scoring system, the game itself is optimal IMO.

Also, no 11th rank promotion please.

https://www.chess.com/clubs/forum/view/new-tournament-modern-ffa?quote_id=69839277&page=1#comment-69839277

HSCCCB

no longer applicable to op's point

The problem with this is: now is not the best time to make drastic changes to the scoring system. Is there any reason why we just can't wait for now, until the player base is settled?

LosChess
HSCCCalebBrown wrote:

The problem with this is: now is not the best time to make drastic changes to the scoring system. Is there any reason why we just can't wait for now, until the player base is settled?

Agreed, when I played the BSRTI Modern FFA arena, my initial reaction was "they are focusing on the wrong changes", there's plenty of threads that spell out what we want, it should no longer be a mystery.

1.  Get rid of SFA and bring back FFA.  Keep Solo separate like it was.

2.  Change starting position to Old Standard or BSRTI, get rid of Omatamix.

HSCCCB

So, for the question, my thoughts. If I misunderstand something, please tell me happy.png

---

Main problem: passivity in 3 player stage (mostly the shuffling found at the end) 

---

Two things to avoid

A. passivity

B. doomed players in the 3stage: Historically teaming till the end, but could also be the weaker player always doomed to be third. This was a major problem, and the consensus was that this is bad. That is still the consensus, correct? If not, that widly changes the direction going forward, if we are okay with opposite teaming through the 3stage.

C. Liquid states that changing aliances is bad. I disagree, but I am argueing that it is unwise to make decisions off this reason because this is not the consensus.

---

good things

A. Dynamic play (risks/activity)

B. It is a good outcome, in my opinion, if most or all of the game is in the four player stage, or we enter the 3stage imbalanced.

[The idea is, if it becomes that almost always that a certain player, for either having the most points or piece activity, will end the game in a draw and win (i.e kings worth no points/ ten..if I'm ahead), then the other players will just prolong the four player stage and eventually play for points. This may also happen if players improve enough so mistakes aren't at the very end.] 

[The other reason is, is that dynamic endings are more likely to happen out of dynamic plays, so a very long 3 stage is likely to have a slow ending because players have a long time to adjust to the imbalance]

---

---

So we want the last two outcomes. This is where it gets harder for me. I am not sure how much rating systems help

For good outcome A.

If we increase the points for 1st, like 5 -1.5 -1.5 -2, it is possible that will encourage players to take risks, but it's possible it will encourage them to be more passive. Right now, I'm just not sure a rating system will do that.

 

For good outcome B. Which I wonder is the best option

We would want to get to where the strongest person entering the 3stage will be the strongest one leaving it. I don't know how the rating system would change it. Reducing king points would do it. I don't think that would make the game passive, but rather accomplish option B. It is possible we'll get there anyway if players keep improving

 

---

For the suggested things

For everything with 2nd = +x, this will mean that people will just team with their opposites till the end. Why risk third when I have a sixty percent chance of getting a little win (+points) and a fourty percent chance of first (+ more points!) Not garenteed, but I think this is likely in a 2nd +0 as well.

Why would people not just team with their opposite in this situation?

For prof's ideas, I don't see a huge difference in play for those, both with themselves and what we have now. 1/3 are the same now. For two, I think that is more of a product of the starting position and the strength of players

---

I personally don't think 2nd/3rd need to be different. i should note, you (liquid) played more with sub-2400 players, and my point of bringing it up is that different ratings require different approaches. I think your suggestions would be fine for, say 2000s, but I'm not sure that the ideas you stated would work for 2500s, because they will just team to the end

---

Finally, two points that [edited to add are no longer applicable to the op's point]

A. as I said earlier, now is not the best time to change that. In my opinion we should as soon as we can change the rating system back to what it was before, with or without solo, and go from there. I might as well mention the starting position, I think we should change that as quickly as we can too, unless there is some knowledge we'll gain by waiting, which I don't see

B. I think these changes that are mentioned are very prone to unforseen consiquences. Not saying we shouldn't do it, but we need to be careful. Another reason we shouldn't change it now: right now, we need a damage control option, not the long term solution that requires more thought.

 

liquid-sun

To clarify, I am raising this discussion now for future decisions. I do not expect or hope for anything to change immediately. I agree that such changed require careful consideration, which is why I bring it up now and preface my ideas with the solution not being obvious.

 

Thank you all for your ideas. You have good points @HSCCCalebBrown.

I must offer the correction that my strategies are intended for 2500+. The main reason why I am not 2800+ is that I keep losing rating when I play in 2200 - 2500 queues, but I then gain rating in 2500+ queues (mostly due to my ability to avoid 4th in 2500+), because I either get targeted or am a victim of foolhardiness in the 2200 - 2500 queues (sticking to 2500+ has led to my current rating of 2700). My ideas are thus oriented to 2500+ queues—I have found that it was much easier to "punish" a player playing for second in this range, which will be elaborated on shortly. As stated in my previous post, such punishment loses its viability when both 2nd and 3rd lose rating and makes more possible the "doomed player" scenario.

 

As @Typewriter44 also mentioned, shifting alliances may not be bad. I was initially brainstorming and will take the ideas mentioned in this thread and modify my OP to bring in new ideas. And yes, I think that passivity is not good.

 

One question I have (I sincerely don't know) is if allowing for a second place strategy may, ironically, help the "doomed player" situation. If second place is a strategy, then people will have to prepare to fight against it during the 4 player stage. Of course, this idea is entirely wrong if temporary teamers remain as such in a foolhardy manner (i.e., the willingness to "settle" for second place). I raise this question because I know that in the past, I was able to punish a foolhardy teamers into third place.

 

What are your thoughts? And is this the primary consideration moving forwards?

HSCCCB

I find it interesting that in 2500+ you often ran into people who played for second. I don't recall that happening, though, to be fair, I was out of the strong ffa curcuit for several months before the merge. Currently, I don't believe that's a problem. You wouldn't happen to have a specific game in mind you could use as an example?

So, 4pc longterm does not want a system for punishing players, I don't think. I don't like unwritten laws. I don't think that is healthy for the game. I personally don't like the entire system of policeing players. I mean, if you punish players, which is your right, what does that help? to some extent it works, but, like people punish (or used to) people who betray their opposites, while people don't do this as much, A.I don't think there's something wrong with it B. sometimes it's the best strategy! I don't think it is healthy for the game to encourage it.

So, allowing for second place. 

team with opposite

50% win lots of points

50% win a little points

don't team with opposite

33% win lots of points

33% win a little points

33% lose points

While that isn't strictly accurate..opposites are rarely at a material equalibrium, I would assume a lot of players are just going to take the safer option. Maybe if you made first a really high margin? But then only really high rated players would do it, and the status quo would be to team till the end.

As for punishing players who do this, this is not feasible because teaming till the end will become, I would think, sooner or later, the meta. 

Were you in the ffa sphere when we had the rating system like this? Quite simply, people teamed to the end then. I don't remember the rating system exactly, but I don't think there is a big enough difference in the established strategic order to prevent it from happening again.

 

Edited to add: all the others are either going to cause play for second or don't seem much different.

I think now that +0 +0 [could have promise], though I still worry about the teaming. eta: which we had, guess who forgot happy.png Ok, that makes me think that there might not be a difference in relation to passivity...Are people playing any differently? In three stage

HSCCCB

Is it still community consensus that teaming till the end is bad?

Typewriter44

https://www.chess.com/variants/4-player-chess/game/26042182/296/3

https://www.chess.com/variants/4-player-chess/game/25953009/142/1

https://www.chess.com/variants/4-player-chess/game/25554134/308/3

https://www.chess.com/variants/4-player-chess/game/25553115/313/2

Regardless of whether playing for second was intentional in these cases (I am confident that in the 4th game it was not), this shows how frequently a player gets 2nd/3rd because of the poor play of another player. This is why I will always support +3 +0 +0 -3. There are many more examples from before the merge when FFA was thriving, but I think this is enough to show that playing for 2nd still exists at the 2500-2800 level

 

liquid-sun
HSCCCalebBrown wrote:

I find it interesting that in 2500+ you often ran into people who played for second. I don't recall that happening, though, to be fair, I was out of the strong ffa curcuit for several months before the merge. Currently, I don't believe that's a problem. You wouldn't happen to have a specific game in mind you could use as an example?

So, 4pc longterm does not want a system for punishing players, I don't think. I don't like unwritten laws. I don't think that is healthy for the game. I personally don't like the entire system of policeing players. I mean, if you punish players, which is your right, what does that help? to some extent it works, but, like people punish (or used to) people who betray their opposites, while people don't do this as much, A.I don't think there's something wrong with it B. sometimes it's the best strategy! I don't think it is healthy for the game to encourage it.

So, allowing for second place. 

team with opposite

50% win lots of points

50% win a little points

don't team with opposite

33% win lots of points

33% win a little points

33% lose points

While that isn't strictly accurate..opposites are rarely at a material equalibrium, I would assume a lot of players are just going to take the safer option. Maybe if you made first a really high margin? But then only really high rated players would do it, and the status quo would be to team till the end.

As for punishing players who do this, this is not feasible because teaming till the end will become, I would think, sooner or later, the meta. 

Were you in the ffa sphere when we had the rating system like this? Quite simply, people teamed to the end then. I don't remember the rating system exactly, but I don't think there is a big enough difference in the established strategic order to prevent it from happening again.

 

Edited to add: all the others are either going to cause play for second or don't seem much different.

I think now that +0 +0 [could have promise], though I still worry about the teaming. eta: which we had, guess who forgot Ok, that makes me think that there might not be a difference in relation to passivity...Are people playing any differently? In three stage

 

When I use the word "punish," I do not mean to punish crossing arbitrary laws that those who play decide on. In fact, I think that such laws are what make FFA toxic at times. Once I betrayed an opposite who once regarded me as somewhat of a friend because his plays was slightly passive. He then proceeded to term me a "traitor" and since played spitefully against me. Arbitrary laws are still an issue in FFA, even at 2500+.

 

By "punish," I mean enact a practical strategy against a more dubious strategy, like how one "punishes" gambit lines in two-player chess with accurate and sharp defense.

 

I did not play enough FFA games where 2nd place was rewarded to know the extent to which people chose to team to the end. If what you say is correct (and if one tends to be powerless against such teamers), then I will support Typewriter's suggestion of +3 +0 +0 -3. Maybe people are just to foolhardy when it comes to playing for second / attacking a player for any other system to work.

 

From my limited experience, I just wonder the extent to which getting 3rd is inevitable in a +3 +0 -1 -3 if one actively preparing to "play for second" when faced with a 2v1 situation.

 

And yes, I think the community believes that teaming to the end is bad if one is powerless against it (which many appear to believe).

TheCatdragon221
liquid-sun wrote:

UPDATE: The following ideas are meant to generate a discussion

That's what Forums are for.

HSCCCB

Ok, I understand what you mean now. My apologies.

So, to recap, I don't think people typically go into the 3pstage planning to team to the end (nor do I think it will happen), so I don't think we need to make decisions for that.

So for weaker play. I'm not sure how much you can try to stop that in-game, nor am I sure you would want to. Most times it happens are more nuanced, I think, so you have a lot of false positives, and most of the times you can't stop it. Besides, the players lose anyway so I don't think that it will disaude them too much (leaving aside if this is healthy for the game, which I don't think so but its kindof irreleveant) I'm not sure a rating system would help. You could add, like, second place looses a bit more in 2500+ than third. That could be abused though, attack then resign, which might lead attacking players to resign, and sometimes resign incorrectly, which just increasesthe problem, if that makes sense.

 I think the best move going forward is to move solo upwards to like 2700+ In theory 2700s should be strong enough that everything is more nuanced.

 

liquid-sun
HSCCCalebBrown wrote:

Ok, I understand what you mean now. My apologies.

So, to recap, I don't think people typically go into the 3pstage planning to team to the end (nor do I think it will happen), so I don't think we need to make decisions for that.

So for weaker play. I'm not sure how much you can try to stop that in-game, nor am I sure you would want to. Most times it happens are more nuanced, I think, so you have a lot of false positives, and most of the times you can't stop it. Besides, the players lose anyway so I don't think that it will disaude them too much (leaving aside if this is healthy for the game, which I don't think so but its kindof irreleveant) I'm not sure a rating system would help. You could add, like, second place looses a bit more in 2500+ than third. That could be abused though, attack then resign, which might lead attacking players to resign, and sometimes resign incorrectly, which just increasesthe problem, if that makes sense.

 I think the best move going forward is to move solo upwards to like 2700+ In theory 2700s should be strong enough that everything is more nuanced.

 

 

Nothing to apologize for. happy.png

While in theory, 2700s should be strong enough to play solo well, in my experience, even 2700s often do not. I will repeat what experience has taught me to believe: FFA will always be a more fair, balanced, and nuanced game than SOLO for all ratings simply because FFA gives some value to playing for 2nd or 2nd and 3rd. This allows one to punish foolhardiness or the game to punish foolhardiness (playing for 2nd), which my experience has taught me will always be a potential issue.

We may simply disagree on this matter, which is totally fine. I think that the main argument against my view is that a SOLO format controls for opening teaming, which may be unfair at times? (From my experience, with careful team play or opting to betray a passive opposite, the temporary teams stage is not so unfair)