Ranking up fast in cooperative games

[tarongot]

I don't see removing 2+ player ratings reducing the playerbase to 5%. And the ones lost would be tilted to those locked into 2+ player games to begin with, which wouldn't be the sort of players getting paired for 2 player games.

I thought you were reffering to cancel ELO for ALL games, not only 2+. Still considering that many games are not playable with 2, and those which are, are rarely balanced for 2. Player base will be much lower.

You say there's one games that "might" have it, so you're not even sure, but if you wanted to streamline things it would be just as easy to say why not eliminate the holdout. Easier even, given its removing work to deal with down the line rather than adding to it. Hard to say given I have no idea what game you have in mind, which might give important details.

Yes, and I also say why I can't be sure: because BGA is hiding ELO details for coop games. But behavior I observed give clues that system, or something close, is implemented there. Game is Pandemic, its not secret, you can check if that will give you details you need.

But cooperative games are very different from competitive ones. For one thing, competitive games have real opponents, ones that actually think, where cooperative games don't. This distinction alone is the cause of one of the boosting issues in cooperative games (everyone agreeing to cancel losing games) that had to be handled. Cooperative games are going to have their own problems and need their own solutions, what works for one group won't necessarily work for the other.

Part of agreeing to cancel was already handled, badly but general principle is ok. Just instead of all or nothing policy, set a border when you CAN cancel if all agree, like below 50% game in progress, or 33% or 25% or whatever value you feel like. That will be enough time to cancel games with people you can't agree on basics, and still far from abusing it for boost your ELO.

And that's the main problem, the believe that coop games are different. They aren't. Neither in this, nor in most of matters.

System could be same, in basics as one used for competitive games. With 1 major adjustment: not treating coop as something different. Treat coop games as competitive games where score depends only on winning with 1 player (many versus 1): player 0 (game). So games in witch only valid strategy is cooperate with others to do that.

PS: Thinking opponent is not needed for ELO calculation at all. That is not a factor if your opponent thinks or is some sort of automatic process.

[Ceaseless]

I want to know why you can select to play a whole campaign but then stop at 1 round and not have it count as a loss. My suggestion would be for settings that lock that sort of thing in "play to X", or if you want it to be flexible, have that as an option to "players vote to continue or not at the end of each round".

You actually really decide after each round if you want to play the next round or not. The whole campaign can be 50 missions long and one round can have multiple tries before you win it. I don’t mind if someone don’t want to play anymore. But this was a turn-based game. And the same person for example who only play 2-player games with missions 1, was one of the team members in my game, which started at a mission over 20. And then he chose to play only one round again. But he saw, that I wanted to play the whole mission, before joining. Others who do the same too will play only one mission and it doesn’t matter if they win or not. They don’t want to start again to try to win and they don’t need it actually, because even a loss is counted as a win.

So there's no way for players to decide otherwise? They're always forced to check round by round and can't all agree to play the whole thing out from the start. That's the part that I find bizarre.

But cooperative games are very different from competitive ones. For one thing, competitive games have real opponents, ones that actually think, where cooperative games don't. This distinction alone is the cause of one of the boosting issues in cooperative games (everyone agreeing to cancel losing games) that had to be handled. Cooperative games are going to have their own problems and need their own solutions, what works for one group won't necessarily work for the other.

I honestly don't understand what you're trying to say. Cooperative games also have real thinking people and are even sometimes heavier than competetive games. Because you have to find solutions together, some of them very difficult. If someone makes a fault, does not understand the game properly, or has little experience, you can no longer win the game or progress at all.

I didn't say cooperative games have no thinking people. I said they have no thinking opponents. That's why I gave the cancel example. In a cooperative game the team has complete control over their own outcome, with no opponent who can object to nonsense like cancelling when things get rough. In a competitive game, if one player is losing, someone else is winning. There's no longer a mutual benefit to stunts like that. Distinctions like these can affect how each group should be handled.

[Ceaseless]

I don't see removing 2+ player ratings reducing the playerbase to 5%. And the ones lost would be tilted to those locked into 2+ player games to begin with, which wouldn't be the sort of players getting paired for 2 player games.

I thought you were reffering to cancel ELO for ALL games, not only 2+. Still considering that many games are not playable with 2, and those which are, are rarely balanced for 2. Player base will be much lower.

You say there's one games that "might" have it, so you're not even sure, but if you wanted to streamline things it would be just as easy to say why not eliminate the holdout. Easier even, given its removing work to deal with down the line rather than adding to it. Hard to say given I have no idea what game you have in mind, which might give important details.

Yes, and I also say why I can't be sure: because BGA is hiding ELO details for coop games. But behavior I observed give clues that system, or something close, is implemented there. Game is Pandemic, its not secret, you can check if that will give you details you need.

But cooperative games are very different from competitive ones. For one thing, competitive games have real opponents, ones that actually think, where cooperative games don't. This distinction alone is the cause of one of the boosting issues in cooperative games (everyone agreeing to cancel losing games) that had to be handled. Cooperative games are going to have their own problems and need their own solutions, what works for one group won't necessarily work for the other.

Part of agreeing to cancel was already handled, badly but general principle is ok. Just instead of all or nothing policy, set a border when you CAN cancel if all agree, like below 50% game in progress, or 33% or 25% or whatever value you feel like. That will be enough time to cancel games with people you can't agree on basics, and still far from abusing it for boost your ELO.

And that's the main problem, the believe that coop games are different. They aren't. Neither in this, nor in most of matters.

System could be same, in basics as one used for competitive games. With 1 major adjustment: not treating coop as something different. Treat coop games as competitive games where score depends only on winning with 1 player (many versus 1): player 0 (game). So games in witch only valid strategy is cooperate with others to do that.

PS: Thinking opponent is not needed for ELO calculation at all. That is not a factor if your opponent thinks or is some sort of automatic process.

I was referring only to what I quoted, which was your own post specifying over 2 players. I expect most of the impact to be felt in the places being cut out rather than echoing into the unrelated 2 player areas, so I'm not too worried about that idea.

I was under the impression Pandemic works like the other cooperative games; guess you can ask around there if you think it's special.

Game progression percentage is a joke, and even worse if you try to universalize it to multiple cooperative games. Even cancelling with things like rough starts can easily be a problem.

You talk about things like "can't agree on basics" and then say how "cooperative games aren't really different from competitive ones" in the same post. These "can't agree on basics" issues don't really play out like that in 2-player competitive games. If I'm playing say, a rated game of Ticket to Ride, and my opponent tries to tell me I'm not allowed to block them from their routes, it doesn't matter. I can laugh at their pathetic attempt to cheat, and if they try to hold the game hostage, I'll just kick them from the game, giving them a well-deserved reputation and elo penalty.

As for the thinking point, I've covered that here.

But cooperative games are very different from competitive ones. For one thing, competitive games have real opponents, ones that actually think, where cooperative games don't. This distinction alone is the cause of one of the boosting issues in cooperative games (everyone agreeing to cancel losing games) that had to be handled. Cooperative games are going to have their own problems and need their own solutions, what works for one group won't necessarily work for the other.

I honestly don't understand what you're trying to say. Cooperative games also have real thinking people and are even sometimes heavier than competetive games. Because you have to find solutions together, some of them very difficult. If someone makes a fault, does not understand the game properly, or has little experience, you can no longer win the game or progress at all.

I didn't say cooperative games have no thinking people. I said they have no thinking opponents. That's why I gave the cancel example. In a cooperative game the team has complete control over their own outcome, with no opponent who can object to nonsense like cancelling when things get rough. In a competitive game, if one player is losing, someone else is winning. There's no longer a mutual benefit to stunts like that. Distinctions like these can affect how each group should be handled.

[tarongot]

I was under the impression Pandemic works like the other cooperative games; guess you can ask around there if you think it's special.

And what is that ? How ELO works in cooperative games? Or how cooperative games works?

You talk about things like "can't agree on basics" and then say how "cooperative games aren't really different from competitive ones" in the same post. These "can't agree on basics" issues don't really play out like that in 2-player competitive games. If I'm playing say, a rated game of Ticket to Ride, and my opponent tries to tell me I'm not allowed to block them from their routes, it doesn't matter. I can laugh at their pathetic attempt to cheat, and if they try to hold the game hostage, I'll just kick them from the game, giving them a well-deserved reputation and elo penalty.

You missed part "in most matters" and what I think I said clearly, most I'm interested in ELO/ranking matter.
I feel, you don't want to address that and keep trying to expand on "feeling of game experience" matter. That is in fact different for coop. You can't laugh at your opponent face in coop games.

So i will write that again:
Treat coop games as competitive games where score depends only on winning with 1 player (many versus 1): player 0 (game).

[ChiefPointThief]

And same what you speak of is in competitive games with more then 2 players (kingmaking, additional accounts, poor player whose game favours opponent, etc.). Maybe delete ELO for those as well.

If BGA decided to do that, I won't be shedding any tears.

I think you will when your player base will reduce to 5%. Its already hard to find players for many games, except few most popular one. Imagine how often you will play if you will have not 10mln but 0.5mln players.

There isn’t anywhere close to 10m players. There have been 10m accounts created. Counting multi & inactive accounts. I agree that some players would be lost but 95% is a very drastic number.

[Ceaseless]

I was under the impression Pandemic works like the other cooperative games; guess you can ask around there if you think it's special.

And what is that ? How ELO works in cooperative games? Or how cooperative games works?

You talk about things like "can't agree on basics" and then say how "cooperative games aren't really different from competitive ones" in the same post. These "can't agree on basics" issues don't really play out like that in 2-player competitive games. If I'm playing say, a rated game of Ticket to Ride, and my opponent tries to tell me I'm not allowed to block them from their routes, it doesn't matter. I can laugh at their pathetic attempt to cheat, and if they try to hold the game hostage, I'll just kick them from the game, giving them a well-deserved reputation and elo penalty.

You missed part "in most matters" and what I think I said clearly, most I'm interested in ELO/ranking matter.
I feel, you don't want to address that and keep trying to expand on "feeling of game experience" matter. That is in fact different for coop. You can't laugh at your opponent face in coop games.

So i will write that again:
Treat coop games as competitive games where score depends only on winning with 1 player (many versus 1): player 0 (game).

How their elo system works. If I'm wrong about that, I'm wrong, but I didn't notice anything special when I took a quick look.

And in elo/ranking, yes, they're different. As stated in the matters of not having an actual opponent, leading to it having different needs than competitive games. Those sorts of issues are probably why BGA keeps having to add restrictions to cooperative games.

[tarongot]

There isn’t anywhere close to 10m players. There have been 10m accounts created. Counting multi & inactive accounts. I agree that some players would be lost but 95% is a very drastic number.

I'm not going to argue about number here. That was just imaginary number I took, to show logic of my adversary. Just because there, supposedly (author is not sure, if, what, and how it was), were some failed attempts to create ranking for coops before. Now system can't be made at all, forever.
Just because some problems were faced, coops have no ranking.
So to show that logic I made a point that current ELO for competitive games, also have problems, not addressed abuses, faults etc. . So lets cancel ranking for competitive games as well. No ranking, badges, tournaments, arena. Lets see how many people will want play for play only.

How their elo system works. If I'm wrong about that, I'm wrong, but I didn't notice anything special when I took a quick look.

And in elo/ranking, yes, they're different. As stated in the matters of not having an actual opponent, leading to it having different needs than competitive games. Those sorts of issues are probably why BGA keeps having to add restrictions to cooperative games.

Coop games HAVE opponent, it impacts game state: active/passive/both, that doesn't matter. Difficult of winning with it can be described with number. So there is number to compare with number that describes player skill. You can get points for win, and lose points for lost.
That's everything what is needed to create ranking system.

Differences, potential problems, different way of reading that ranking doesn't change that.
It is still a ranking. And I still see that it can use same algorithms and factors as ELO for competitive games in BGA uses.

[ChiefPointThief]

I was under the impression Pandemic works like the other cooperative games; guess you can ask around there if you think it's special.

And what is that ? How ELO works in cooperative games? Or how cooperative games works?

You talk about things like "can't agree on basics" and then say how "cooperative games aren't really different from competitive ones" in the same post. These "can't agree on basics" issues don't really play out like that in 2-player competitive games. If I'm playing say, a rated game of Ticket to Ride, and my opponent tries to tell me I'm not allowed to block them from their routes, it doesn't matter. I can laugh at their pathetic attempt to cheat, and if they try to hold the game hostage, I'll just kick them from the game, giving them a well-deserved reputation and elo penalty.

You missed part "in most matters" and what I think I said clearly, most I'm interested in ELO/ranking matter.
I feel, you don't want to address that and keep trying to expand on "feeling of game experience" matter. That is in fact different for coop. You can't laugh at your opponent face in coop games.

So i will write that again:
Treat coop games as competitive games where score depends only on winning with 1 player (many versus 1): player 0 (game).

How their elo system works. If I'm wrong about that, I'm wrong, but I didn't notice anything special when I took a quick look.

And in elo/ranking, yes, they're different. As stated in the matters of not having an actual opponent, leading to it having different needs than competitive games. Those sorts of issues are probably why BGA keeps having to add restrictions to cooperative games.

Is it different yes. But there could still be ranking systems applied to separate skill level is what this other person is arguing. The issue is devs/bga being too lazy to do so or just simply thinking that it isn’t a good idea. Not because it is impossible but for reasons like decreasing toxicity in coop games which there still is even with hidden elo. I know ufm has argued in other threads about the +1 model being good for coop games. Meaning that you just get 1 elo for the game no matter the outcome. I was just reading how this is problematic in the game the gang because their is no separation between players from 1-99. So you can’t even exclude poor players from matches w/ advanced settings because it would require 100 games to be labeled as “average”.

For me it was fairly easy to pick up the game because I know how to play poker (it is just coop Texas holdem) so I would say my ranking is “strong” before even playing the game. But there is no separation in my ranking compared to some of my friends who don’t have the understanding of poker that I have and there never will be with such a ranking system.

[tarongot]

Is it different yes. But there could still be ranking systems applied to separate skill level is what this other person is arguing. The issue is devs/bga being too lazy to do so or just simply thinking that it isn’t a good idea. Not because it is impossible but for reasons like decreasing toxicity in coop games which there still is even with hidden elo. I know ufm has argued in other threads about the +1 model being good for coop games. Meaning that you just get 1 elo for the game no matter the outcome. I was just reading how this is problematic in the game the gang because their is no separation between players from 1-99. So you can’t even exclude poor players from matches w/ advanced settings because it would require 100 games to be labeled as “average”.

For me it was fairly easy to pick up the game because I know how to play poker (it is just coop Texas holdem) so I would say my ranking is “strong” before even playing the game. But there is no separation in my ranking compared to some of my friends who don’t have the understanding of poker that I have and there never will be with such a ranking system.

Thanks for that info, ufm was totally wrong and short-sighted. This feature didn't solve anything, and created new problems.
Also it is not +1 for a game, it is +1 for WIN, and a proper win in ranked mode. Not something count as win because someone left table. That case doesn't give 1 ELO, maybe gives fraction maybe 0.

The Gang is very popular game so many people see that problem. But problem there is smaller because game is popular and rather short. Now Zombicide has it but with bigger impact, because its less popular game and much longer. And imagine how it is in game that after 2,5 YEAR only 3 people are above beginner level.

[Ceaseless]

And what is that ? How ELO works in cooperative games? Or how cooperative games works?



You missed part "in most matters" and what I think I said clearly, most I'm interested in ELO/ranking matter.
I feel, you don't want to address that and keep trying to expand on "feeling of game experience" matter. That is in fact different for coop. You can't laugh at your opponent face in coop games.

So i will write that again:
Treat coop games as competitive games where score depends only on winning with 1 player (many versus 1): player 0 (game).

How their elo system works. If I'm wrong about that, I'm wrong, but I didn't notice anything special when I took a quick look.

And in elo/ranking, yes, they're different. As stated in the matters of not having an actual opponent, leading to it having different needs than competitive games. Those sorts of issues are probably why BGA keeps having to add restrictions to cooperative games.

Is it different yes. But there could still be ranking systems applied to separate skill level is what this other person is arguing. The issue is devs/bga being too lazy to do so or just simply thinking that it isn’t a good idea. Not because it is impossible but for reasons like decreasing toxicity in coop games which there still is even with hidden elo. I know ufm has argued in other threads about the +1 model being good for coop games. Meaning that you just get 1 elo for the game no matter the outcome. I was just reading how this is problematic in the game the gang because their is no separation between players from 1-99. So you can’t even exclude poor players from matches w/ advanced settings because it would require 100 games to be labeled as “average”.

For me it was fairly easy to pick up the game because I know how to play poker (it is just coop Texas holdem) so I would say my ranking is “strong” before even playing the game. But there is no separation in my ranking compared to some of my friends who don’t have the understanding of poker that I have and there never will be with such a ranking system.

And how would that be done? Each coop game is going to present its own headache to solve. The people who had to actually deal with the effects of those systems and play it out didn't seem to get much traction with it, and when it was attempted, there were issues getting it shut down. Toxicity was one, boosting was another, you'd see these absurd ratings that didn't really match reality.

Do you have a link to Ufm's description of the system they had in place? I find the idea of only 1 point per game a bit odd, though it's not like I have any great ideas where cooperative games could really go myself.

How their elo system works. If I'm wrong about that, I'm wrong, but I didn't notice anything special when I took a quick look.

And in elo/ranking, yes, they're different. As stated in the matters of not having an actual opponent, leading to it having different needs than competitive games. Those sorts of issues are probably why BGA keeps having to add restrictions to cooperative games.

Coop games HAVE opponent, it impacts game state: active/passive/both, that doesn't matter. Difficult of winning with it can be described with number. So there is number to compare with number that describes player skill. You can get points for win, and lose points for lost.
That's everything what is needed to create ranking system.

Differences, potential problems, different way of reading that ranking doesn't change that.
It is still a ranking. And I still see that it can use same algorithms and factors as ELO for competitive games in BGA uses.

It can't think or advocate for itself at all. As I mentioned several times, that creates problems. Boosting was one obvious one. Another would be even sorting out what beating it is worth in the first place. I know Chess sites had enough issues trying to properly rank Chess bots created to face players, failing miserably, and that was for a competitive game where the bot did interact with your moves.