Hello Ever''body, I have an idea for a new type of branching missions in RTSes. Instead of having to meet all these victory conditions, why not be able to go on to a different mission in which you must compensate for your shortcomings, rather than playing any given mission all over again? I think that this would add a lot of replayability, and make the game more fun. It would also add a touch of nonlinearity to an often linear genre. Granted, it would be very difficult, especially in missions where you have several objectives, but I think you could get around that. That''s what I''m trying to figure out: How does one accomplish such a thing, while at the same time, retaining lots of mission diversity? I''d like to thank you all ahead of time for (1)reading this, and (2)any feedback you give. Thanks and God Bless, Jim

Well, you could design a series of levels that were keypoints to the non-linear play.
IIRC Warcraft 2 had different AI types you could assign - sea battles, air battles, ground battles, etc. You could design an observer AI to the user to keep track of how much time he spends working on his town as opposed to how much time he keeps track of warriors, and punish him / play to him for it.
If a player ( like me ) specializes in hydralisks, the AI would start working on seige cannons, the game would give me a bunch of island maps or aerial battles, and there might be a custom map where hydras were disabled all together.

I''m not sure though.... would you want to punish the player for playing only one unit, or design the levels for him. With my SC example, the game could either force me to vary my tactics by fighitng hydras, or the game could try and generate a series of specialty hydra missions. ( take on a battle cruiser ( hard but possible )) go on a gorefest against mutalisks, etc.

Well, I''m not sure if SoakinKittens'' reply was intended for this thread.. maybe I''m just not understanding the problem =)

What you would probably need (apart from designing 1000 missions to compensate for different conditions) is a good random storyline thing.

Initially, the computer sets up a list of Main Objectives for the entire campaign:

- Destroy base at Certain Town
- Eliminate radar facilities at Some Other Town
- Rescue civilians from a Final Town


Each main objective would have various completion possibilities, and sub objectives that need to be completed (each with various completion possibilities). Then, the method deemed "easiest" is presented to the player, and if he succeeds, well done. If not, another option is presented, and the player tries another tack. If the player runs out of options, it''s game over.

Note that this would be best for small campaigns - maybe a long list of smaller campaigns (4 missions each?) could work well?

On posible way design branching style RTS would to get rid of the consept of levels. Have large level which have chunks streamed from disk (like dungeon siege does). This gives you a large arena to play in. This blurs the difference between level to level and mission objective to mission objective.

One of the key problems in RTS, is when you pack too many mission objective into a small area, it becomes cramped(well in my opinion).

Hey there again...
Thanks for all these good ideas! As for the ''elimination of level design,'' I don''t know how well that would work in what I''m trying to do, but it could likely stand on its own very well. What I was thinking is that you design the map with very few objectives (i.e. no ''This Guy must survive'' stuff), and keep them generally open-ended. This would make it easy to construct a dynamic campaign by what objectives were completed or failed. How''s that sound?
Thanks and God Bless,
Jim

I think it''s a good idea. I pet issue of mine is that games should make "failure an option". I think changing the narrative if the player does poorly is much more compelling that making them repeat a task until they get it right.

The main challenge you face is keeping the amount of content that needs to be created from getting out of control. If every mission leads to 1 of 2 missions (one mission for if you succeeded in the previous mission, and one for if you failed) then the number of missions that the designer must create grows exponentially.

One solution is only branch into different missions occasionally, if at all. Instead of branching, you could alter missions to be slightly but significantly different based on the player''s previous performance.

Completing bonus objectives might offer resources for the next mission that wouldn''t be there otherwise.

Example: Destroying a heavily defended airbase on one mission may reduce aerial attacks on the next

Another: Locating and holding a supply depot for a period of time offers upgraded weapons for the next mission.

etc. etc.

Failed missions might make the next mission easier with little gain or harder to break even in the strategy and restore the player to a more equal ground with the opposition.

I do agree that making a boatload of missions to cover all possibilities would be discouraging for the designers, especially if it isn''t all that likely that certain missions will ever be played.

Now, if the game could dynamically generate missions, maybe with some specific scripted occurrences placed to drive the story... hmm...

Instead of designing missions one by one or by sets, instead make large wargame-like maps and let things develop, you wouldn''t even have to do anything approaching a-life, perhaps read some military history books and watch how the outcomes of battles affected things.

I think a random story/mission generator would begin to feel boring to the player. Add it as a different single player feature, like deathmatch, when the player just wants to play some new level.

I definitely like the what you do affects the future later on idea. (see Waverider''s post).