Gameplay Proposals/Campaign
From UFO:AI
Contents |
The final campaign
This proposal describes the singleplayer campaign as it should behave in its final incarnation.
This system is now (largely) implemented, though exact formulas and values might differ. Use this article as a general guideline to the campaign game mechanic.
REVISED by --BTAxis 00:38, 13 May 2007 (CEST) after discussing the issue with Wanderer. The revised version contains many of his ideas.
REVISED by --Wanderer 05:35, 13 May 2007 (CEST) Included more detail into the process, did some formatting and general cleanup, included more information regarding map scripts.
REVISED BY --BTAxis 00:56, 14 May 2007 (CEST) after another brainstorm session with Wanderer.
Overview
The campaign should be much less rooted in fixed definitions and targets than is the case at the time of writing. Rather than having a set of scripted missions, missions should appear depending on alien activity. As such, the missions should not so much be part of the campaign, but a result of the campaign parameters.
During the campaign, the aliens will pursue their goals. They will try to attack cities, build bases, subvert the human population and governments, and so forth. These activities will spawn missions depending on what is happening. A landed harvester will generate a UFO mission, an attack on a town will generate a terror mission, discovered bases can be attacked. And of course the aliens will launch their own attacks against PHALANX installations.
Over time, alien activity will increase in intensity. More UFOs will visit Earth, and they will be larger and better armed. The soldiers brought planetside will be bringing increasingly dangerous equipment, and they will be harder to kill. This process should theoretically continue indefinitely until the player is defeated, or wins the game. This is important because it makes the campaign a race against time, rather than a race to the finish.
The difficulty setting should ideally influence the process described above. On higher difficulties, the aliens will gain in strength faster, and their numbers will be larger than on lower difficulty settings. In effect, this means that players will have less time for R&D in the strategic part of the game, and will have to fight harder in the tactical part (of course, other difficulty-related factors have an impact on that as well).
Implementation and definitions
Alien activity
Alien activity is determined by a framework of "interest" levels. This interest level starts at 10 and keeps increasing over time. We will assume an interest level of 1000 to be the point at which an average player will be overwhelmed by alien activity. Good players will be able to endure beyond this point, but everyone will have their limit. The speed of increase of interest levels, and the 'upper threshold' that a player will be able to survive, will be affected by the difficulty levels.
In addition to the overall interest level, the aliens will have interests in a number of categories. These categories include:
- Recon
- Harvesting
- Terror Attacks
- Base Attacks
- Base Building (this interest doubles as government subversion by non-XVI means)
- Base Supply
- XVI Propagation
The best way to think of interest levels is to think of them as overall goals the overmind has. The overmind is also distracted by the running of its ship, so at the beginning, Earth is nothing more then a footnote. The longer the planet is able to defend itself, the more interest the overmind takes in the conflict. The more interest, the more resources it's willing to use.
Interest will spawn events. Note I don't say missions. A mission that the player can affect should be about 90% of the events available, but the framework should be built to allow for non-player interactive events. It is also possible that an event is "null", meaning the event does not happen at all.
There are two kinds of interest. The first is categorical interest, such as interest in Recon or Harvesting. The other type is the 'overall' interest.
Choosing the Interest Events
First, we need to separate the process out to cycles (three days, for example, but this could be a function of difficulty setting) when a new alien "activity cycle" will start. The amount of alien missions to be generated will be calculated from the overall interest rate (as per a bell curve, such as #missions = (CurInt/10)^(0.6)), rounded down. This bell curve will ensure that the number of alien events is ramped up fairly early on in the game, while the late game mainly causes events to be harder to beat.
For each mission to be spawned, the framework will select what kind of mission to spawn. The total value of all categories will equal 100%, and each category's value divided by the total will equal the chance it has of being selected. Once a category is selected, the next step is choosing a specific event from that category. For this we use event tables. They will most likely be shaped along the following lines:
| 0 | No Event | ||||
| 4 | Aerial Recon | Scout Craft | mintime 8, maxtime 20, direction changes 4 | Success: -50% Recon, +15% harvest, +15% BaseBuild | Failure: +20% Recon, +5% BaseAttack |
| 6 | No Event | ||||
| 7 | Aerial Recon | Scout Craft | min time 3, maxtime 8, direction changes 7 | Success: -50% Recon, +15% harvest, +15% BaseBuild | Failure: +20% Recon, +5% BaseAttack |
| 10 | No Event | ||||
| 12 | Ground Recon | Scout Craft | min time 2, maxtime 6, direction changes 3, number of landings 2 | Success: -50% Recon, +15% harvest, +15% BaseBuild | Failure: +20% Recon, +5% BaseAttack |
| 18 | No Event | ||||
| 24 | Aerial Recon | Escorted Scout | mintime 3, maxtime 8, direction changes 4 | Success: -50% Recon, +15% harvest, +15% BaseBuild | Failure: +20% Recon, +5% BaseAttack |
| 32 | No Event |
..... Continuing on until say, 1000.
The event with the highest number that is still smaller than ( (CI * 0.3) + ( ( CI * 0.7) * ( RAND() ^ 0.85) ) will be the event that is selected.
Increasing and decreasing interest
The overall current interest starts at 20 and rises at a steady pace. A good pace would be 1 per 28 hours. This would mean the interest would reach 1000 after a little more than three game years, which is a period of time that fits within the storyline.
The individual interests start at their own specific values. They will increase and decrease depending on what happens in the game. Referring to the above example event list, we see that a successful Recon mission as described by 24 decreases interest in Recon and increases interest in Harvest and Base Building. It is important to note here that the INCREASE percentage is taken from the overall interest value, and the DECREASE percentage is taken from the category's own interest value.
In addition, each interest category is weighted. The weight is a modifier that applies to all increases that are applied to that category interest, but not to the decreases. A second weight applies on top of this, but only if the category's interest value is greater than the overall interest value. In effect, this means that it IS possible for events that are more difficult than "average" to occur, but the greater the discrepancy, the smaller the chances of it ever happening.
Examples:
- Suppose overall interest is at 50, Harvesting is at 30 and Base Building is at 12. and Recon is at 35. Per the calculation given above, this corresponds to two events per cycle. Suppose one Recon mission is selected, and the other event is null. Suppose also that the recon mission is a success. That means the Recon category is halved to 17 (as we round down), Harvesting is increased as per 30 + 50 * 15% = 37 and Base Building is increased to 12 + 50 * 15% = 19.
- Suppose that same mission fails because the player intercepted it. Then Recon is increased as per 30 + 50 * 20% = 40.
- For the next example, suppose overall interest to be 50 again, with Recon also at 50. Should the mission fail this time, the increase in Recon is slower, because it gets weighted extra. This is because it now exceeds the overall interest. Supposing this extra weight is .5, the calculation would be 50 + 50 * 20% * 50% = 55.
Executing an event/mission
Once a list of missions is thus generated, the missions will be executed over the course of the activity cycle. Once executed, they will be carried out as per mission script files, in combination with team definition script files that help generate alien teams to be used during the events.
At this time, It is undetermined if the number of mission types should be broader, to allow for a second 'mission specs' file to be used, or if a smaller number of mission types with the parameters to be loaded included with the declaration would be the best effort. If a second mission script file is used, these script files provide parameters to the game engine that help determine what UFO to spawn, whether or not it should have escorts, where to spawn the UFO, where to conduct the mission, how long to spend on each action associated with the mission, and so forth.
Category Definitions
Here is a slightly more in-depth description of the mission categories and their behavior.
Recon
Recon can spawn either aerial recon missions (UFOs that fly across the geoscape without landing and disappear after a while) or ground recon missions (UFOs that fly around, land in one or more locations, then disappear).
Harvesting
Harvesting spawns UFOs that fly to an area, land, gather life-forms from Earth by means of Bloodspiders and soldiers with Kerrblades, then take off and return to the mothership.
Terror Attacks
Terror Attacks are missions to human population centres for various purposes, including harvesting, XVI propagation, and increasing unhappiness with PHALANX. The UFO is spawned some way away from its destination, then proceeds to fly there and generates a terror mission.
Base Attacks
Base Attacks are assault missions on PHALANX bases. A UFO will be spawned on the map, fly to a PHALANX base and, should it survive, initiate a base attack mission.
Base Building
A Base Building mission is a mission to establish a new base on the geoscape. At lower values (say under 400), these missions will never cause a new base to be built, but will instead contribute to the subversion of governments by non-XVI means. When a Base Building mission is successful, a new alien base will appear at the location the UFO landed, but stays in "under construction" mode for 30 days before going online.
Base Supply
Base Supply missions consist of supply transport UFOs that are spawned on the geoscape, fly to the nearest alien base, stay there for a while, then take off and return to orbit. Interest in Base Supply increases over time by a fixed value multiplied by the amount of bases on the geoscape, both operational and under construction.
XVI Propagation
XVI Propagation missions consist of missions where the goal of the aliens is to infect more humans by several possible methods. Alien syringes, Breeder units, and terror attacks on hospitals/blood banks are all methods by which XVI can spread rapidly. The mission will be carried out by a UFO and its crew, optionally aided by infected humans. Propagation missions can be initiated by UFOs or spontaneously in infected regions.
XVI Propagation missions must never occur in an area that isn't connected to at least one operational alien base by the XVI infection spread. PHALANX bases in regions with high XVI saturation should suffer fairly frequent attacks from infected humans. Once a region reaches 100% saturation, any PHALANX bases in the area should be automatically destroyed.
The XVI propagation category poses an interesting problem with the concept of 'saturation'. Once a region reaches 100% XVI saturation, no more Propagation missions should occur there. This category will have to monitor the region infection levels from the hard code to determine its values. However, the higher infection levels there are in regions with established alien bases, the higher BaseBuilding should increase so that the overmind can saturate more landmass with XVI.
Maps
As ALL missions are generated randomly (excepting a few story-related fixed ones, but even these are spawned depending on factors outside the coders' control), maps cannot be associated with missions directly. Instead, the Earth's terrain should be divided into a set of "zones". Each zone "layer" should focus on one criterion. Examples of criterea are terrain and architectural style.
To determine the correct map for a region, especially when dealing with the random maps, four criterion must be reviewed. 1) Terrain or zone in question. This can be anything from as simple as 'desert' to 'New York City'. 2) The type of mission occurring. A Terror Mission won't occur on a farm, and a ground recon wouldn't occur in the middle of Bangkok. 3) The alien ship in question, to determine the right build tile. 4) The player's ship in question, to determine the right build tile and to make sure there's enough spawn points on the map.
Implementation
Currently the thinking is the following.
First, a map.ufo file. This file would contain the parameters explaining how a map can be used. For example, the Antarctican City of FrozenButt would have definitions stating 'ice', 'city'. Missions would be 'Terror', 'Harvest'. Because the map was built as a solid build, instead of a random map build, it would have a single ship type on it at this time, so you would list 'Alienship: Harvester' and 'PlayerShip: Dropship1'.
A random file might have say, 4 listings, one for each of the Alienships that could land there, so it could choose the right build method.
The second would be the ZoneDefinition.ufo file.
This file would simply hold information that would attach to how the programmers decide to divide up the map, with logical information like the following:
Antarctica1 (The rim of Antarctica): Ice, Tundra, Water
Antarctica2 (The interior): Ice, Tundra
India: Tundra (for the mountains), Rocky, Plains, Farm, Village
...
You could theoretically include a choice for specific missions that had to select a location (instead of getting shotdown over it), such as Terror: City.
UPDATE:
See Proposals/Map_Descriptors and Mapping/Terrain_Types for more details on the implementation of maps.
