Modding tools!

I think it's using a system of govenors, ie algorithms that can manage stuff for you.I would guess they will be similar, albiet more useful, to the governors found in empire total war. Set it to automanage contruction and taxes, letting you deal strictly with military. They won't be quite as good at the tasks as you will though. Thats my 2 cents worth anyway.

Simply put, it's impossible- all SP games start to dull when you play them enough.

MP will help Elemental out a lot, esp if combined with modding tools and mini-expansions.

"Heh, I was thinking more about repetition within a single game itself and not the repetition of multiple games being similiar."

Yeah. It's a hard balance to maintain. Either there's not enough to do early in the empire building phase, or at some point the empire becomes too big to handle. And at some point in-between, you are doing the same thing over and over again.

Random events, quests, etc. can help give you a change of pace throughout the game.

[quote who="Daikaze"]Civilizations was/is a great series, but eventually games became so tedious that you'd automate most of your stuff, go through boring repetition, or simply quit. It wasn't all that enjoyable to sit there and have to manually control every little thing in the game.
So, how will Elemental avoid such tedious repetition? [/quote]

  I recommend the following:

   1) A random map generator with optional settings.

   2) Modding Tools for terrain, races, technologies, magic, map structures, battlefield maps, AI Personalities, items, resources, etc., .

   3) An easy to use map editor such as the ones provided by AOW:SM and Heroes_3.   Painful map editors would be like from Dominions_3.

   4) Allow the game to include the Overseer(DM/GM) player who could provide unique and new changes within the game.   There's currently a topic.

Yeah, as far as I remember the devs haven't really elaborated much on how they're going to deal with micro-ey stuff.. I mean I think a lot of people have assumed governors, but I don't remember that specifically being said by the devs (although I do think they said that we'll only have to deal with as much or as little as we want to...which does heavily imply some sorta governor system). I think your idea sounds great GW, although I can imagine it being a bit of a resource hog for them to get right.

A slightly simpler way to allow players to speed thing along when they get accustomed to the game might be saveable custom build lists. It may turn out this isn't that useful (for instance if there's THAT much difference in what a city should build depending on circumstance/surrounding tiles etc that there's never an optimum path you'd like to follow).. but I like the idea of founding a town that I want to specialise in research, then just clicking my previously created research build list..and it'll build everything I've decided I want in a reasonable order. I know other games have done this (like Alpha Centauri) by allowing you to have a governor concentrate on a particular specialisation for a city.. .but sometimes they make stupid mistakes and I'd like to be able to set the order myself. I mean say there's 3 tiers of research building; library, school and university.. now an AI governor that's specialising in research may trot these out in a fairly wrote way, as obviously the devs thought they'd be useful or they wouldn't be in the game, but maybe >I< think that the school doesn't give a big enough bonus to research for the extra building cost over a library so I want to spam libraries everywhere or something... well I'd like the opportunity to have a list where I could go from game to game being able to engage that once, and have that happen at my research cities.

I think that if Elemental manages to be so replayable as the Civilization series many many people will be very pleased. I know I will. If I start up a Civilization game today then I know that despite it having the same controls/elements it will be a whole other game/story that unfolds.

With the random map options and the modding tools I've good hope Elemental will capture that same "magic" Still it will be a turnbased fantasy strategy game every time you play it...

Simply put, incorrect. Dungeon Crawl Stone Soup achieves high replayability through heavy use of randomisation, avoidance of no-brainers and grinding. I have slightly over 16 wins in that game, and want more. Levels are randomly generated, monsters, items and traps are placed randomly. Entrances to other branches can appear on random levels (and in development version, some branches may even not appear at all in a game !). You are not guarranted to find item X or spell Y in any given game. And there's no reloading, so you either succeed or die trying.

http://crawl-ref.sourceforge.net/

If you're offended by ASCII graphics or 2D tiles, it's better to stay away.

Other heavily randomised game: Spelunky, a 2D platformer.

http://www.youtube.com/watch?v=s5IRc9uPANo

-------------

In general, I think that replayability has a lot to do with heavy randomisation of a game (level generation, random events, object placement). Additionally, it's good to make sure that there are no no-brainers. This way all choices remain viable and are not only nominally there (HOMM3 has units no one bothers to recruit like Peasants). It is wise to avoid grinding.

Additionaly, I think using save/continue instead of save/load helps a lot, because it makes it impossible to replay the same situation over and over.

- b0rsuk

GW Swicord:

I'm not entirely sure what you mean by 'rich' randomisation. What the distinction is. If you're concerned that merely random levels or maps may look bland, that's not quite the case. Random levels can look quite interesting :

http://www.freefilehostingnow.com/image.aspx?code=9d526f53a112aa434baacd8a9c7fa5e48c16&large=1

http://groups.google.com/group/rec.games.roguelike.development/browse_thread/thread/4c56271970c253bf#

(I'm currently working on implementing the algorithm in Python)

But what I really wanted to say is that random level generation is not mutually exclusive with hand-crafted content. For example, while Crawl's dungeons are mostly random, 'vaults' - hand-made sections with specific or semi-random items and monsters are inserted here and there. In fact, when you start the game, one of randomly chosen 'entry vaults' is placed where you start. There's a big variety of vaults of various kinds, and players submit more (they're moderated so not everything will make it through).

Off-topic: I'm concerned that Diablo3 may be not very replayable. We already know that outdoors are going to be mostly static, unlike in Diablo2. Dungeons will remain random, although if the first gameplay video is of any indication, they're likely to be less random than before. More like assembled from bigger pre-fabricated blocks. Such is the price you pay for detailed and polished graphics using current 3D technology.

Oh, I see. This is a real concern, because I have yet to see a randomly generated story that is even remotely interesting. People have been granted randomly generated patents, though.

I think good story and ramdomisation don't go well together. For an interesting story you'd have to make sure that all events happen more less like they're described, which means mostly static map. Hopefully Elemental's developers can provide both good hand-made story/campaign and solid random level generation. For story, I generally prefer a good book.

Crawl intentionally has very light story. More like lore/setting. A good story wouldn't work in a game that's designed to be as replayable as possible.

Cloth map is doable without looking very ugly (see Dominions 3), and I'd argue that good AI demands that it's generalised and can handle almost any situation it's thrown into. Heroes of Might and Magic 4-5 is heavily scripted (AI is custom designed for each particular scenario) and they're criticized for poor AI.

Repetition is an inescapable part of existence and that can’t be changed. Of course, repetition also translates into consistency which I personally like in a video game. A problem arise when the game is so consistent that you can easily predict the outcome of an action before it is even made or when there is only one choice to be made for almost every situation. I think a game has the most replayability when it has around 5+ meaningful degrees of freedom and perhaps 9 variables total. The meaningful part is the lone problem area since everyone has a different definition of what gives something meaning. But, for me if there are 5 or more major areas that have a cumulative and distinct affect on the direction a game can take I should be happy.