Proposal for Spy model

From FreeOrionWiki
Jump to: navigation, search

A Proposal for a Spy Model

This is a generic approach to the problem. The main idea is to make it flexible so differences in tech, race and other bonuses/penalties can translate directly into a value and don't affect the core mechanics at all.

Covert Missions

Spy activity is organized in missions. Each mission has 2 inherent values. First a percentage of completion witch states how much of the mission objective has been completed. Second, a percentage of awareness from the other faction of the existence and nature of the mission, when this value reaches 100% the mission is foiled and the player gets no more information (An auto-abort-mission option can be used to abort any mission that is past some level of awareness). Both values start with a base value of 0. There are several types of possible missions, here are some examples:

- Take tech (any / any in category / specific)
- Get information (empire / star / fleet)
- Incite revolt (empire / star / planet)
- Influence diplomatic relations with other empire
- Steal (resources / ships)
- Sabotage (resources / ships / infrastructure)
- Assassinate (emperor / leader / spies)
- Infiltrate

Note that some missions are more difficult than others. In general, a broad, unspecified target is easier to infiltrate and less probable to get caught. For this each kind of mission should have a "base difficulty" (see below for its use).

Spy Abilities

Spies have some abilities much like an RPG game. These abilities are shared to all the spies in the empire and depend on tech, training, etc. Individual stats for each spy are IMO unnecessary. Each ability can be more or less relevant to each mission type (a relevance factor from 0 to 1 should be measured for each ability/mission combination, all the factors in one mission row should be normalized so the sum of them is 1). This "relevance matrix" is fixed. For starters I would use 2 abilities: tactics and spying but a more complex net of abilities can be devised easily.

How things happen

Having designed an RPG game system myself, I can say that the following system works quite well: Each active spy make a roll each turn. You multiply each ability by its relevant factor and add it to the roll. Then you subtract the average value for the roll (if the roll can go between 1 and 100 a 50 is subtracted) and the base difficulty for the mission.

What you get is a number that represents the level of success (LoS) for the attempt. This allows us to know how good things went. For example, a get tech mission LoS measures how much info on a tech I got this turn (in %) and how much noise I made (added to the awareness for this mission). Note that a higher base difficulty affects the LoS so we get the behavior we need. If more than one spy is assigned to a single mission each spy rolls independently (thus generating awareness and results faster).

A LoS of 0 is the breakeven point between success and failure. Bigger LoS means more success and less awareness while a negative LoS means things didn't go well (more awareness, less success). Note that a critical failure (a BIG negative number) can foil a mission completely. Since the roll is random the auto-abort option won't work always but it can set how risky your mission can get in exchange for results.

Support and Internal Security

This aspect of spying is very similar to Moo2 mechanics. You put a number of spies on an empire and leave a number of spies at home to provide for internal security. The idea is that internal security increase the base difficulty of missions inside the empire by some factor. Optionally a normalized array of values between 0 and 1 for each mission type can establish different policies for internal security focus.

Before assigning spies for a mission your spies need to infiltrate the enemy empire. Infiltrating is a mission by itself but once you infiltrate your agents you can create new missions for them inside this empire. This is not really realistic but has good game mechanics and it abstract the need to infiltrate in each mission. When a spy finish a mission without being discovered it automatically return to the infiltrated pool.

Also, each infiltrated spy can work to weaken the internal security of the other empire. One of the missions mentioned before is useful for this. We use "Assassinate Spies" to decrease the defense network of the enemy, but the difficulty of this mission need to be balanced so It's at least equally possible for us to loose that spy. This also allows the combined action of several empires to destroy the security network of a single one.

The price of Failure

When a mission is foiled there are 4 possible outcomes (that can be equally probable or not): First, a spy gets caught and reveals some information to the enemy empire. Second, the spy suicides and don't reveal anything. Third the mission fails but the agent is not discovered and goes back to infiltrated status. And finally, the spy manages to escape and returns home (to the pool in security). Note that this means that depending on the difficulty of assassination missions we could eventually loose more agents than we kill. In case of multiple agents in a mission each one roll independently.