New multiplayer mode with looser synchronization

[Folket]

Koshling, you original plan was alot better.

"Unit not previously visible to player A becomes visible to player A - requires turn sync between the two players concerned" Does not solve anything.

Turn sync and turn order enforced should be only be during war. Perhaps have three configurable options.

Turns allowed ahead during peace. (integer default 20)
Turns allowed ahead during war. (integer default 0)
Turn order enforced during war. (boolean default true)

This way it is possible to create fast games with many human players. (20 5 false)
The goal most be to make the game playable and fun.

Declaring war is only allowed if you are ahead in turns. (i.e. you will continue you turn later)

 

[Koshling]

Koshling, you original plan was alot better.

"Unit not previously visible to player A becomes visible to player A - requires turn sync between the two players concerned" Does not solve anything.

Turn sync and turn order enforced should be only be during war. Perhaps have three configurable options.

Turns allowed ahead during peace. (integer default 20)
Turns allowed ahead during war. (integer default 0)
Turn order enforced during war. (boolean default true)

This way it is possible to create fast games with many human players. (20 5 false)
The goal most be to make the game playable and fun.

Declaring war is only allowed if you are ahead in turns. (i.e. you will continue you turn later)

I think it's reasonable for the set of conditions to be (somewhat) configurable. Having an enemy raiding party able to walk right past your posted lookouts and never get to see them due to them taking several turns IS a major issue, but the condition set that triggers synchronization requirements can be at varied levels of strictness (which would be game-option selectable). For the visibility case the options encoded might be:

• strict - newly visible unit forces synchronization
• medium - newly visible unit to a unit which is in sentry mode (ONLY) forces synchronization
• weak - no visibility checking

Personally I'd play on medium, but not weak - having an entire army walk past my lookouts, and wind up attacking me from the other side with no way to detect that would not be acceptable (to me).

 

[Thunderbrd]

I think it's reasonable for the set of conditions to be (somewhat) configurable. Having an enemy raiding party able to walk right past your posted lookouts and never get to see them due to them taking several turns IS a major issue, but the condition set that triggers synchronization requirements can be at varied levels of strictness (which would be game-option selectable). For the visibility case the options encoded might be:

• strict - newly visible unit forces synchronization
• medium - newly visible unit to a unit which is in sentry mode (ONLY) forces synchronization
• weak - no visibility checking

Personally I'd play on medium, but not weak - having an entire army walk past my lookouts, and wind up attacking me from the other side with no way to detect that would not be acceptable (to me).

Neither would it be acceptable to have an enemy wander past a vulnerable position to your unit simply because your unit is not on sentry.

To be honest, I don't think the whole thing would work well being able to go out of game turn (player turn sure... that works). Say I have a scout out wandering about and because you're 4 turns ahead, your scout moves into view of my scout. Does he 'see' my scout? My scout wouldn't even BE there 4 turns ahead of where I am now. What about what this means to a goody hut grab?

Personally, I don't mind if the system is set up to bypass the need for server and turn order within a turn, but imo trying to allow turns AHEAD to be played is going to cause problems no matter how innocent it may seem.

 

[Koshling]

Neither would it be acceptable to have an enemy wander past a vulnerable position to your unit simply because your unit is not on sentry.

To be honest, I don't think the whole thing would work well being able to go out of game turn (player turn sure... that works). Say I have a scout out wandering about and because you're 4 turns ahead, your scout moves into view of my scout. Does he 'see' my scout? My scout wouldn't even BE there 4 turns ahead of where I am now. What about what this means to a goody hut grab?

Personally, I don't mind if the system is set up to bypass the need for server and turn order within a turn, but imo trying to allow turns AHEAD to be played is going to cause problems no matter how innocent it may seem.

If your scout became visible to his unit that would force synchronisation and you'd be unable to play further until he catches up. Similarly, goodie hut popping was something I listed in my original set of sync-forcing conditions, if the ahead player attempts to pop a goodie hut visible to a behind player that force a sync and they are unable to do so until the other player catches up (which may mean the other player gets it first).

There is SOME advantage even with all the sync rules to playing ahead at first, since you get to explore a wider area, and possibly encounter goodie huts NOT visible to anyone else sooner, but provided the play ahead limit is set reasonably low that's really not much of an advantage at all.

It makes the difference between perhaps 4 turn a week as we are now, and 20-30. To me that's worth some compromise.

 

[Folket]

I think you can replay what has happened since your last turn, or was that some other game. I thought it was civ 4.

Also forcing syncronization does not help any in the case you have listed.

If someone 20 turns behind you move a unit so you see it, forcing him to play 20 turns will not help you notice him earlier.

 

[Koshling]

I think you can replay what has happened since your last turn, or was that some other game. I thought it was civ 4.

Also forcing syncronization does not help any in the case you have listed.

If someone 20 turns behind you move a unit so you see it, forcing him to play 20 turns will not help you notice him earlier.

No, but since he won't be able to attack you until he regains sync the advantage is in some sense 'contained'. However, I don't see 20 as ever being a sensible number to allow - I was thinking more like 5 at the start of the game, dropping to 1 or 2 later, so that the main advantage becomes just lack of strict ordering once interlocking contact is achieved. However, as observed earlier, even just lack of ordering requires synchronization rules to prevent someone having two turns in a row relative to their enemy during an attack.

 

[Thunderbrd]

No, but since he won't be able to attack you until he regains sync the advantage is in some sense 'contained'. However, I don't see 20 as ever being a sensible number to allow - I was thinking more like 5 at the start of the game, dropping to 1 or 2 later, so that the main advantage becomes just lack of strict ordering once interlocking contact is achieved. However, as observed earlier, even just lack of ordering requires synchronization rules to prevent someone having two turns in a row relative to their enemy during an attack.

I think if it worked like a simultaneous play game that didn't require any particular turn order, it'd be effective. You COULD establish a necessary turn order between particular players when those players are at war so that they couldn't take two turns before the other had a chance to take theirs but if you didn't make that necessary, if one player uses that advantage, the other is then set up to use that advantage as well. They'd be just as well to make an agreement between each other as to when they'll be taking their turns.

This business of enforcing synchronization strikes me as something that would begin taking place almost immediately among the players - especially on Start as Minors.

 

[Koshling]

I think if it worked like a simultaneous play game that didn't require any particular turn order, it'd be effective. You COULD establish a necessary turn order between particular players when those players are at war so that they couldn't take two turns before the other had a chance to take theirs but if you didn't make that necessary, if one player uses that advantage, the other is then set up to use that advantage as well. They'd be just as well to make an agreement between each other as to when they'll be taking their turns.

This business of enforcing synchronization strikes me as something that would begin taking place almost immediately among the players - especially on Start as Minors.

Right, but I **think* you're just arguing about what criteria should enforce synchronization now. I deliberately did NOT include 'being at WAR' exactly because of start as minors. I included DECLARING WAR (one time thing), attacks, and moving through another player's territory, because those are things you need to be able to react to. Simply being at war, but the opponent not taking any aggressive action in a particular turn IMO doesn't require synchronization (and that covers most turns of start-as-minor 'wars').

The argument that if you don't bother synchronizing allows the other play to take advantage the following turn doesn't really work IMO. It favors the attacker too much - if I have a large stack 2 away from a choice of 2 target cities and get 2 moves in a row, and you only have enough force to adequately defend one, you have to be able to react to my first move, or else your 'advantage' comes after you lose a city and I get all the city defense modifiers you should have had.

 

[Folket]

Koshling, you think that forcing syncronization give oppertunity to react. Which just is not true. It is only true if the party that forces syncronization is ahead.

Perhaps you me4an that the action would be unavailable until the player has reach the syncronization. But that would create very strange situation where you can sentry around a player who has played less turns to imprison them.

And yes we are arguing when to force synchronization. I would like to see a version that makes it possible to play a 2000 turn game with six humans. If that puts limitation on which options that are available I think that is just fine.

Also this kind of game would probably do better if AI players are not allowed. But I guess with a lot of work it is possible. Perhaps not for a first version.

 

[Koshling]

Not giving up on the very loose (allow multiple turns) model yet, but while I (and hopefully others) think about what that might mean a bit more, here is a simpler model that I think we could implement pretty fast:

• No detailed synchronization triggers
• Without any player-level synchronization requirements (see below) you may play your turn in any order (but all within the one turn)
• At any time any player may REQUEST synchronization with any other. While any such requests are in force (you can turn it off again on some later turn) the relative ordering of those two players must be maintained as whatever it is in the turn the request is made

This would allow us to play in any order until we start interacting, and then continue that way until someone feels the need to enforce strict synchronization with someone else (usually two players who are or are about to be at war). Even then, uninvolved players in the conflict retain the ability to play in any sequence.

It may be possible to extend this by-request model to greater-than-one-turn asynchronous too, but we can think about that later...

 

[Thunderbrd]

Not giving up on the very loose (allow multiple turns) model yet, but while I (and hopefully others) think about what that might mean a bit more, here is a simpler model that I think we could implement pretty fast:

• No detailed synchronization triggers
• Without any player-level synchronization requirements (see below) you may play your turn in any order (but all within the one turn)
• At any time any player may REQUEST synchronization with any other. While any such requests are in force (you can turn it off again on some later turn) the relative ordering of those two players must be maintained as whatever it is in the turn the request is made

This would allow us to play in any order until we start interacting, and then continue that way until someone feels the need to enforce strict synchronization with someone else (usually two players who are or are about to be at war). Even then, uninvolved players in the conflict retain the ability to play in any sequence.

It may be possible to extend this by-request model to greater-than-one-turn asynchronous too, but we can think about that later...

I'm personally a bit anal about wanting to keep things less manipulable and I think this will be a good solution that would make for an improvement that wouldn't lead to big potential unfairness issues. One neat trick would be if when the last player logs in his turn in a given game turn, it sends a notification email to all players with an email established. (But then again, you need the right kind of email server for that so not all would be capable - I myself ONLY use hotmail and other non-local email servers.)

Nevertheless, its my opinion that this suggestion would be optimal on all points.