Freeciv-Web Wiki

This page is the "rulebook" for how to conduct games, and for interactions between players and GM.

This page assists new Gamemasters to learn how to GM games.

Game Setup[]

A .serv file defines all the game settings for the map and other rules, which you can request to see and work on with a FCW admin. The .serv file is carefully reviewed for correctness.

At the time the game is launched, the admin will copy the .serv file into the folder of on-going longturn games. The game will automatically launch.

Evaluating the Map[]

  • The random generator does a good job, but 10%-30% of the time, a player will have a starting position that's too strong or too weak.
  • For the GM to evaluate the map, a real player must first enter the game. The GM should only enter the game after a real player is in it.
  • At this point, the GM should carefully evaluate starting positions. If the map is imbalanced, request a regen. The admin must go into the server, delete the game, relaunch, then repeat. This is a lot of work—there are also other methods to adjust away the imbalances of an otherwise good map:

Hand of God rule[]

Generally, the GM should try as much as reasonable to see the game has fair starting positions for every player. Exact equal fairness is never possible due to randomness, and would be impossible to evaluate anyway. Randomness in the game is actually part of what makes it good--so that one has to find creative ways to take advantage of different starting conditions, make the best with what one has, and not always play the exact same way!! Still, if any positions are obviously worse, you should consider regenerating a new map until they're all reasonably similar in opportunities. Factors to look at:

  • Land area
  • Number of neighbors
  • Resources
  • Opportunity to quickly migrate to new areas via ocean or rivers.

No nation will be exactly equal in these things. But where worse in one thing, they should be better in another. If worse in multiple, and not compensated well enough in others, consider regenerating the map.

The Hand of God rule also permits a GM, in the first 7 turns only, to do micro-adjustments into starting positions and context. The earlier this is done, the better. This is frowned on but also understood to sometimes be necessary and better than the alternative (unfair starts or hours of regen where time and human opportunities are not available.) Nevertheless, it's always better to avoid it, and it should not be overdone.

Even so, it might be that after 4 map generations and much effort, a superior map exists for 24 out of 25 players and the 25th player could do OK with just a little adjustment. Getting a second opinion for such cases is highly advised in order to protect your actions later. Adjustments can be:

  • a little extra gold
  • a little extra bulbs
  • an extra unit like a Tribesmen, or Boat, or Goods
  • a "prophecy" -- the situation COULD be fair but only if the player knew some little detail that he can't see. Be careful with this one!

Except for prophecy, these adjustments require a non-player with GM access to do a command-line in-game LUA script command. These are difficult to do, easy to make an error on, and dangerous. One more reason to not be too liberal in the hand of God rule. Another reason is it's very hard to be 100% certain your adjustment is not an overcorrection. And a third reason is there is a high chance other players will be upset if there is info-leakage that someone got a special compensation (do your best to avoid this!). And a fourth reason is that as more players hear of it, more will be begging and arguing to get special awards in every game, or perhaps making accusations, debates, and dramas about fairness.

👉🏽 A silent player should have just as much chance of getting a balance adjustment as a "squeaky wheel" making long rhetorical arguments. Be heartlessly self-conscious to not make it so that whining and drama gets rewarded more than the silent player. Also, the same goes for upstanding well-respected players vs. potentially notorious or hated players.

👉🏽 Keep in mind, very few people know how well the game self-balances itself for randomness. Even though it's a game where one must use even the slightest advantage to leverage a win, it has multiple complex layers of self-balancing mathematics. Consider a player who has room for 12 cities and one who has room for 11. It seems to be an advantage for the former, but 12 cities creates 1 extra unhappiness in Monarchy. One player can have 11 size 4 cities with no temple (44), and the other can have 12 size 3 cities (36), with a total cost of 660 gold in temples to get to (48). This 660 gold can be spent in other ways by the player with 11 cities, such as in bulbs, workers, military units, etc. This is only one of hundreds of examples of how two or more rules form "self-balancing rule-couples" within the game.

  • Don't allow assumptions about imbalance to underestimate the influence of hundreds of self-balancing rule-couples,
  • Don't over-apply the Hand of God.


Commands[]

Server commands can be reviewed here: Server commands

Client commands for Gamemasters are listed below:

Supercow lock - ctrl-alt-shift-S: by default the lock is on. Turning it off will allow you to better inspect game elements, as if you were a player. For example, with supercow lock OFF, you can click on cities or unit stacks. When supercow lock is OFF, the game client has taken you out of OBSERVER MODE. This means other important observer and Gamemaster features will be lost while the lock is off. You can recover them by relogging or toggling the lock.

Debug Mode - ctrl-shift-E: exposes the server's log messages, which may be useful for helping developers debug a crash, etc.


Trading Cities[]

Freeciv-Web allows players to trade cities. However, the setting is OFF by default. (If it were ON, there would be too many temptations for cheating.)

Players may approach the GM and explain they want to trade cities. They must give a normal reason that's a win/win for the self-interest of both nations. A win/lose city trade that unilaterally favors only one nation (e.g., "I'm quitting") is not encouraged or automatically deemed acceptable. In some cases, the GM may decide that unilateral favor is acceptable: it may preserve game integrity or game balance, prevent giving a hollow victory to the competing alliance who would win simply from having an enemy quit the game, etc.)

Doing the trade[]

Inform both players that it's their duty to notify you after the trade is completed. Then do:

/set trading_city enabled

Be sure to turn it back off after they inform you that the trade is completed:

/set trading_city disabled


Turn length[]

Turn length should be increased when players need it. Here we tend to favor the rights of the minority who need extra time, over the majority who don't need it and want the game to go faster. It's a balance between preserving integrity by not forcing players to lose their turns or play improperly or stressed out, vs preserving fun by not making players bored and impatient or stressed out they can't finish their turns. The table below gives a vague approximation about what turn lengths are appropriate during which phases of the game.

Turn lengths should always be one off from a multiple of 12. This aligns the timer to the 24 hour earth clock, but off by 1 so that specific geographic regions don't get turn change always at the same times of day, which would give advantage to some time zones and not others. Acceptable lengths under this paradigm are: 11, 13, 23, 25, 35, 37, 47, 49. Ultra-longturn: 59, 61, 71, 73, 83, 85.

Starting with UR32 and R44, an automatic turn timer extension is built into the game. Turns become 25 hours at T60, 35 hours at T100, and 85 hours on T121. This potentially reduces the number of times the GM must change the timer to only once per game: on T121, where roughly half the games have already finished.

/set timeout is used to increase turn length. Note that the command does not add to the timer. It resets the length of the timer, in seconds. There are 3600 seconds to an hour.

Setting the turn length with console commands
command turn length appropriate
/set timeout 82800 23 hours T0 - T150
/set timeout 90000 25 hours T30 - T150
/set timeout 126000 35 hours T85 - T150
/set timeout 133200 37 hours T100 - T150
/set timeout 169200 47 hours T110 - T150 *

* 47h is appropriate when a nation becomes "superpower size" and requests it: approximately 80±15 cities.


Multi-account issues[]

Multi-accounting definition[]

Multi-accounting is when one player secretly controls two nations in the game. That is, not in a legal way through delegation. It is illegal and is considered cheating. It's a threat to the trust and stability of the community and, like all cheating, must be handled with maximum strictness.

  • Two players within the same household are also considered multi-accounting unless they have got permission from the GM before joining the game. The GM may evaluate their situation and make special rules and public announcements about them, if it is felt they can be included in a proper fair way. (Different starting continents, or known by all to be allied, or prohibited from alliance, etc.) This is a published rule.
    • 👉🏽 The reason for this rule should be rather clear. Any multi-accounter when caught could make some claim like "but it's my brother, we live in the same house and use the same computer." There is no way for an admin or GM to sort out the truth of such claims.

How to discover multi-accounting[]

There is an art to discovering multi-accounting, and it comes down to the fact that multi-account play always seems to fall into the same suspicious patterns which give a red flag to something unusual. This form of cheating tries to leverage the power of two nations to gain an advantage. But managing two large nations in an alliance that avoids logging on at the same time of day is just not tenable for the course of an entire game. Sooner or later, a suspicious "tell" will reveal that multi-accounting is going on:

  • It is noticed by GM or players that two nations always log around the same time.
  • One nation acts against its own self-interest and "accidentally" in the interest of another.
    • armies behaving not as they should
    • undefended cities being settled in convenient locations for the other nation to conquer
    • capturable units for no self-interested reason putting themselves into situations advantaging the other nation
    • actions/reactions to all the above that could not happen with the vision the nations have
    • irrational transfers of gold, technology, or other diplomatic acts
    • too many coincidences and lazy mistakes taking place where one or more of the above are also in effect
  • Reports from the players should be taken seriously. Their constant attention to the game will notice even the slightest irregularity.
  • Once suspicion arises, flag the nation for magnified scrutiny.

Once a suspicion exists, it's possible to get admin help to use a number of tools to look further into the situation. And also possible to look much more closely at every move on every turn.

What to do after discovered[]

It is important to make an example out of it, but not in any way that reveals how you know. Our job is not to teach cheaters how to not get caught next time. Do the most strict and severe action possible that doesn't unfairly give some other player an advantage:

  • Setting both nations to AI
  • Using /take and setting all cities to make moderate defensive units
  • If fair and balanced, assigning a new player to take over a nation carved out of the former two nations, that is fair in size compared to the other nations currently playing. You can set city trading enabled and transfer the ownership of the cities to create such a nation.
  • Starve or destroy certain cities using the Destroy City command.
  • 👉🏽 If you're not sure how to resolve things fairly to the non-cheating participating players, you can use the GM channels on Discord to ask for help. Make sure to then invite others to a secure location: we don't want any GMs who are playing in that game to see screenshots or partake in such a discussion.


Cheating Suspicions[]

These will generally be reported from an in-game player. You should also try to catch them yourself by logging in often enough to have an idea what's unfolding in the game.

Not only is cheating strongly prohibited at FCW, but our policy is to be very public about taking it seriously in words, appearance, and actions. You may escalate any allegation or discovery of cheating into private GM channels to get support for your actions. This process also protects you from anyone claiming the type and degree of action taken was inappropriate.

  • FCW rules protect innocent players from any false public allegations, so make sure you disallow publicity of it while the truth of the allegation is still under question.


Delegation issues[]

This is an area that FCW is trying to evolve more defined rules upon, to avoid vulnerabilities to misapplication. However, it is difficult:

  • Control over a delegated player for a long period of time can effectively give a "multi-account advantage" to a player, which is bad.
  • Failing to allow a player to control a delegate for a longer period can make the player's alliance lose for no other reason than someone quit.
  • Forbidding a delegation can give a victory to another alliance.
  • Permitting a delegation or allowing its extension in time, can give a victory to an alliance.

For now, be familiar with the FCW rules on delegation and use your best judgment to preserve game balance, integrity, and enjoyment.

Efforts will continue to diminish idlers, make replacement players more easily recruited, and better define fair delegation rules.


Mediation Scenarios[]

No RTS requests[]

RTS theory[]

RTS is not ipso facto bad or discouraged. In pure game theory, RTS is more fair than allowing a player exclusive right to attack with a guaranteed privilege of sustaining no counterattack. So why are NO RTS requests part of the game? Because players are not online 24 hours per day, and this creates a chain of implications. How? Well, some players get NO RTS simply because the other is offline. Players who are online less would then incur the disadvantage of being attacked without getting to RTS. Meanwhile, players who camp online more often would get both NO RTS when it's to their advantage, and RTS when it's to their advantage. FCW balances this by having periods where RTS is allowed, and also periods where it is not allowed.

RTS legality chart

Mediating incidents[]

Players will often complain about a NO RTS request not being honored, during a time of the turn when RTS is allowed and encouraged. For reference, the times RTS is allowed can be found here. In such cases simply link them to the chart and the rules. The "Turn Done" button has been repurposed as an RTS legality indicator in longturn games. You can also make them aware of it and give explanations if appropriate.

Sometimes players will be emotional and accusatory on each other over RTS. At such times, you can create a mediation server or a 3-way chat to hear both sides. These chats are often very hot, emotional, and disorderly. But can be controlled with some methods. Before moderation starts, inform the players that they will speak one at a time when you call upon them to do so, and the party who is least civil and most out of order may have the decision go against them. This will usually establish order during the moderation.

Mexican standoffs[]

The current RTS rules have usually resulted in fair games with infrequent issues. However, in theory there are cases where whoever calls "NO RTS" first would get a game winning advantage. Obviously we don't want our games decided like a simple children's game where the first person to call something wins. These cases are called Mexican Standoff and the GM is expected to help explain these rules and link to https://www.freecivweb.org/rules. In such cases where nothing else works, it's likely best for the GM to be a supervising referee to an arrangement where, for 5-10 minutes, each side moves one unit, reports the move, then the other alternates. While in theory this is fair and should be used more often, it also encourages expectations for users to call "Mexican Standoff" all the time and start demanding it ever more frequently, with ever more bickering about when it's appropriate, and more frequent incidents created from it. Therefore, be judicious and careful in how often you allow Mexican standoffs.


Psycho-aggression complaints[]

The community always has a small minority of players who are "psycho-aggressive" and also a minority of players who are sensitive to it. Sometimes, it's the same player! So, from time to time, in every few games or so, you'll be the lucky GM facing an issue like this.

It is recommended to tell the complaining player to use their Discord "Block" command (right-click user >> Block) and the in-game command:

/ignore player_name

Being a nanny and moral judge over people of different nationality, culture, race, religion, politics, dietary faith, personality, "identity", and social caste, is difficult and likely to upset one or both parties. Thus encouraging them to block each other if they can't resolve it, is often the laziest but also best choice to handle these issues. However, sometimes you may have to go deeper into it and make judgments, warnings, threats, and penalties.

Even so, please notify the GM community of ALL such complaints WITHOUT EXCEPTION. (Players don't get much chance to notify the community, under the non-impugnment rules.) The greater GM community must therefore make itself always alert of players who engage in toxic behaviors, and how often they do so, across different games with different GMs. Reports in GM channels start developing a larger picture. The larger picture is needed to eventually take action on a player, or to understand that new rules are needed to regulate certain behaviors in the future.


Conduct between GM and Players[]

It needs to be minimal and factual and impersonal.

  • MINIMAL. The longer the contact, the more that interactions can subliminally leak things you know or imply you know things to players.
    • It is far easier to leak info than the average person supposes, and there are human skills in causing and hacking leaks, as well as advanced psychological and logical methods of inferring great amounts of information from microscopic and subtle statements or even lack of making a statement at all. Rather than take a course on this, which we can't require for GM certification, we can only say to keep the interactions utterly minimal, short, dry, and to the maximum degree possible, in no way referencing anything concrete about the game, its facts, lack of facts, actions, lack of actions, etc.
  • FACTUAL. The less factual the contact is, the more subjective it is, which can lead to human factors influencing different treatment of the players, and less impartial.
  • IMPERSONAL. The more personal it is, the more that social role, persona, ethical debates and expectations can be imposed on you, allowing you to be hacked by rhetoric and arguments.

The necessity of these 3 requirements can't be understated. Experience has shown that a loose lid on any of these three principles will eventually cause an escalation of issues and loss of contentment. Even if you think no issue occurred between you and the specific player, when other players discover the lid was "loose" in your interaction with that player, they will make comparisons, justifications, arguments, judgments, and so on. A player who has no contact with the GM at all should feel like no matter how much contact the other players have with you, they are not getting even a 0.1% advantage in information or meta-information, over the player who is dead silent.

If, on the other hand, that impression does occur, it opens the idea that political favoritism and rhetoric become a meta-method of gaining in-game advantages. How you squash this will be up to your own personal style, but however you do it, if you do not insure minimal, factual, and impersonal contact, you're going to get into issues. Try not to find out the hard way from experience. Think ahead before responding to each contact from a player how it's already going to end in a minimal, factual, impersonal manner.

  • Players can ask about general rules but not concrete facts or things in-game. Allowing even the slightest concrete discussion of anything for any reason creates an avalanche of mistrust about to what degree you go down this slippery slope and how hackable and imperfect your judgement is in doing so.

Info-Leakage[]

As stated above, the risk of this is extremely high and extremely underrated. Please don't find out the hard way the truth of this warning! An abstract example can help clarify what you're up against:

👉🏽 Let's say the player Johnny has, from in-game intel, logic, knowledge of settings, and other methods, figured out that, in the game, either A is the case or B is the case. If the GM tells Johnny that A is the case, it does not seem to leak inappropriate intel to Johnny. However, if the GM tells Johnny if B is the case, it's an obvious illegal leak of intel to the player. Johnny can then ask the GM whether A is the case, in order to privately figure out whether B is the case. The GM is completely unaware that by harmlessly telling Johnny not to worry and that A is the case, he has leaked intel to Johnny that B is not the case.

This should make it obvious why all communication with players should never fail to be minimal. Even slight delays in how you word your answer to the question above, or refuse to answer it, will create all manner of speculations and mental triggers for the player, who could then rhetorically argue your obligations to answer, perhaps subconsciously looking for any eureka in how you answer to confirm some suspicion about intel to which he should not have access. Your obligation is to not answer and you can link the player to this paragraph to hold steadfast.

👉🏽 The web of interconnected information about game states, game setups, GM actions, non-actions, personality tendencies in how one answers or refuses to answer, make a vast puzzle that can be cracked and even hacked to leak intel. Therefore, any question even vaguely related by six degrees of separation to a game state, should be "non-answered" in the most minimal non-verbose non-self-justifying way possible.

Although their question might not be illegal, it may have enough similarity to some other possible future game situation where an answer would be illegal. Then, if you revealed their question was illegal, you just leaked intel about that situation and this situation. That is, the legality or illegality of a question itself reveals whether there is a corresponding game-state underneath it. Categorically refusing to answer a question and refusing to give reasons for why, should be done liberally. If a player has a problem with that, point them here.

👉🏽 You must refuse to answer harmless questions in order to avoid vulnerability to "photonegative logic"—that is, the mere fact you are known to answer some harmless questions, but don't answer some other question, photonegatively reveals that the question was not a harmless question. This then leaks intel that B was the case about an in-game state. Be excessively hypercautious. Build a reputation for being a brick wall, and you will later not only be respected for it, but also experience a decreased number of tedious and exhausting info-hacking attempts by the players, upon yourself.


Other Rules[]

  1. For obvious settings errors, a Gamemaster (GM) can fix any such settings before T10 provided they do not affect any player and provided the GM fully announces such changes, giving at least 1 turn notice before making the change. Exception: announcing creates an exploit, in which case the GM may elect not to give notice (but must inform those members of the admin council who are non-players in the game.)
  2. In the first 7 turns, the GM can make adjustments to players' starting conditions if and only if it improves an imbalance in fair starting conditions.
  3. Regarding multiple players in the same household OR sharing the same device: The GM can approve such cases if asked privately in advance of you joining the game; otherwise it is forbidden. Gamemaster decision will be final. GM may make special rules for allowing the case, like: making all players aware of it, or forbidding alliance between the two players.