Building on the Drafts by amateurgamer, azum4roll and Stalker0 and on the discussion in those threads, I would like to discuss with you my own thoughts on the Espionage topic. All ideas and in particular all the numbers used are preliminary, feel free to comment and make suggestions.
First of all, some general considerations:
There are four possible ways to use a spy (in City-States, in other players' cities, as Diplomats, as Counter-Spies). Spies in City-States are obviously used mostly by players who focus on a diplo victory. The Spy/Diplomat dichotomy should correspond to a peaceful, cooperative vs. an aggressive approach against other civs. Offensive Spies should harm the targeted player and help the spy owner, while diplomats should give benefits to both parties. Counter-Spies focus on internal strength and on protection against foreign spies and foreign influence. In particular, I think it's not a good idea if offensive spies only copy yields and do no harm to the targeted player, as this would blur the distinction between Spies and Diplomats, and it would also make it hard to have useful Counter-Spies.
The draft of the rework:
In order to have more strategic and tactical opportunities with spies, the number of spies is increased to 2 spies per era when playing with 8 player/16 city-states. (How the number of spies is calculated when playing with other settings needs be reworked as well, the current scaling is not good.)
As in Stalker0's proposal, the time needed for traveling and for establishing surveillance is reduced and made more homogeneous.
Spy effects depending on cultural influence over the other civ is removed completely.
Spies in City-States:
Currently, sending a spy to a city-state is very much a gamble. If the spy fails to rig the election, they get nothing, and it's almost impossible to predict when this will happen. Proposal:
I would like to make the use of Offensive Spies impossible if the two players have signed a Declaration of Friendship, and to make the use of Diplomats impossible if the two players don't have shared embassies (so in particular if one of them has denounced the other). Like it is now, using a Diplomat will give a diplomacy bonus, having an Offensive Spy discovered will harm diplomatic relationships.
Diplomats:
Having a Diplomat in the Capital will give science and culture to both players each turn while the Diplomat is active. The yields are based on the average of the two players' empire-wide yields. The spy owner will get more yields than the other player (exact formula needs to be worked out). This means that the positive effects of using a Diplomat are higher if the targeted civ is ahead, but with the drawback of supporting them even further. The Diplomat will also give its owner a Tourism bonus against the other civ (maybe slightly higher than it is now to make up for the additional culture).
Offensive Spies and Counterspies:
While the idea of having a coup-like mechanic with advanced actions is tempting, I think such a system is rather complicated and it's hard to get it balanced. I would rather build up on the current siphon missions.
The problems with the current implementation:
Spy Levels and Experience:
First of all, some general considerations:
There are four possible ways to use a spy (in City-States, in other players' cities, as Diplomats, as Counter-Spies). Spies in City-States are obviously used mostly by players who focus on a diplo victory. The Spy/Diplomat dichotomy should correspond to a peaceful, cooperative vs. an aggressive approach against other civs. Offensive Spies should harm the targeted player and help the spy owner, while diplomats should give benefits to both parties. Counter-Spies focus on internal strength and on protection against foreign spies and foreign influence. In particular, I think it's not a good idea if offensive spies only copy yields and do no harm to the targeted player, as this would blur the distinction between Spies and Diplomats, and it would also make it hard to have useful Counter-Spies.
The draft of the rework:
In order to have more strategic and tactical opportunities with spies, the number of spies is increased to 2 spies per era when playing with 8 player/16 city-states. (How the number of spies is calculated when playing with other settings needs be reworked as well, the current scaling is not good.)
As in Stalker0's proposal, the time needed for traveling and for establishing surveillance is reduced and made more homogeneous.
Spy effects depending on cultural influence over the other civ is removed completely.
Spies in City-States:
Currently, sending a spy to a city-state is very much a gamble. If the spy fails to rig the election, they get nothing, and it's almost impossible to predict when this will happen. Proposal:
- Remove the concept of voting periods and rigging elections altogether.
- As long as a spy is in a city-state, the player will get a certain amount of influence each turn (scaling with spy level and era).
- As long as a spy is in a city-state, the chance to successfully stage a coup increases by 1.5/2/2.5 percent each turn.
- The presence of a spy of the current ally will reduce the coup chance increase per turn for all enemy spies by 0.5/1/1.5 percent.
- The coup chances themselves do not depend on spy level anymore. The values for the coup chances and for their increase per turn are always shown accurately in the UI (no more "assuming no counter-spy is present").
- As it is now, if a coup fails, the spy will be killed.
I would like to make the use of Offensive Spies impossible if the two players have signed a Declaration of Friendship, and to make the use of Diplomats impossible if the two players don't have shared embassies (so in particular if one of them has denounced the other). Like it is now, using a Diplomat will give a diplomacy bonus, having an Offensive Spy discovered will harm diplomatic relationships.
Diplomats:
Having a Diplomat in the Capital will give science and culture to both players each turn while the Diplomat is active. The yields are based on the average of the two players' empire-wide yields. The spy owner will get more yields than the other player (exact formula needs to be worked out). This means that the positive effects of using a Diplomat are higher if the targeted civ is ahead, but with the drawback of supporting them even further. The Diplomat will also give its owner a Tourism bonus against the other civ (maybe slightly higher than it is now to make up for the additional culture).
Offensive Spies and Counterspies:
While the idea of having a coup-like mechanic with advanced actions is tempting, I think such a system is rather complicated and it's hard to get it balanced. I would rather build up on the current siphon missions.
The problems with the current implementation:
- It is hard to predict the amount of yields that can be stolen from a city.
- It is hard to predict which missions will be available.
- Spies are very strong against tall civs, and very weak against wide civs. The general value of spies is low if there are many cities per player.
- Counterspies are too random. Killing an enemy spy is too strong, but there's also a chance that a counter-spy has no effect at all.
Proposal:
Siphon Missions:
- There are siphon missions for gold, faith, tourism, science, and culture. The missions to steal a tech or a GW are removed.
- All missions are available for all spy levels, higher spy levels increase their effectiveness.
- Siphon missions will transfer yields each turn while they are active (not afterwards). Just like sending a spy to a CS, sending a spy to a major civ will have an immediate effect.
- The targeted player is notified that a spy is stealing from them, but they are not told in which city.
- The formula to calculate the amount of yields is reworked to address the wide/tall issue. The yields of other cities are taken into account as well.
- The formula details don't need to be known by the player, as in the Espionage UI, a new tab is added that shows for every city and every mission an estimate of the yields that can be stolen. This estimate is the correct value plus a random estimation inaccuracy of +-25%.
- Example formula: 25% of gold in this city, plus 5% of the gold produced by all other cities. Values increased by 33% for each spy level above the first one.
- Maybe: Yields are increased if the spy owner is technologically/culturally behind, and decreased otherwise (Example: Yields modified by +-3% for every tech/policy the enemy has more/less than the spy owner, capped at 30%). I don't know if this is really necessary, a player who's ahead will in general produce more yields anyway.
- For each player, there is a cap for the maximum amount of yields that can be stolen by enemy spies. For example: 25% of the total gold income from all cities.
- If the cap is exceeded, the effectiveness of all spies performing this type of mission will be reduced proportionally.
- Example: The cap for stolen yields is 200 gold per turn for a certain player. There are 3 spies stealing gold from their cities, they would normally steal 150, 100 and 50 gold per turn. The sum of these amounts is 300 which is 1.5 times the cap, so all values are divided by 1.5 and the spies will steal 100, 67 and 33 gold, respectively.
Counterspies and Buildings:
- Having a counterspy in a city allows the player to always see how many spies are conducting which missions in the city.
- The amount of yields that are stolen by enemy spies are reduced by 15/20/25%.
- Each turn, for every enemy spy in the city, there is a chance of 3/4/5% to catch and identify them. The mission is then aborted, the spy is send back, and there is a 10 turn cooldown during which the spy owner cannot send another spy to the same city.
- Constabularies and Policy stations reduce the amouts of yields that are stolen from the city by 15% each, and increase the chance to catch an enemy spy by 2% each (catching spies can happen only when a counter-spy is present).
- If a Constabulary is built, it will be shown in the Espionage UI if any enemy spies are currently conducting a mission in the city. A Police Stations shows in addition how many spies are conducting which missions (like a counter-spy).
- Counterspies have passive benefits for the player who uses them: +1 Happiness in the City scaling with Era, +6/8/10% City Strength, +6/8/10% Science and Culture in the City.
The counterspy system can maybe be extended later to include different options to choose from when a spy is captured, as in the other proposals. For now, I think such a system would be too complicated given how rare it is to catch an enemy spy, and I'd like to focus on the other parts of the rework first.
Tactical Missions:
- The missions to pillage tiles and to spawn barbs are removed. There will be one tactical mission which is useful in a war:
- This mission works like the city-state coup missions. It has a base success chance, and the success chance is increased by 1.5/2/2.5% each turn while the spy is doing it, a counterspy reduces these values. The "coup" can be attempted at any time, if it fails, the spy is killed.
- If the mission is successful, city defenses and healing rate are reduced, and the city gets +5 Unhappiness for a certain number of turns. The spy returns home.
- In addition, extended visibility is given while the spy is preparing the mission and, if successful, while the mission effects last.
Spy Levels and Experience:
- Killing spies is removed (except for coups and the tactical mission, and the player can control the risk there). Level 3 spies will get experience like other spies. When a Level 3 spy has reached 100 XP, the spy retires and a new Level 1 spy is immediately recruited.
- As long as a spy is on a mission, they will get 2 XP per turn. Spies in City-States and on a tactical mission will also get (100-success chance) XP if they successfully attempt a coup.
for each tech you have that they don't, (and +1 
for policy count differences? not sure on that one). So at least on the tech side, you only both get science benefits when you have different overall tech paths; if you're just behind then they only get the gold benefit.