Read: That gamespy rebuttal, That gamespy article by some guy trying to sell his book and Catnews

Here's how it is.

The players, because it's a game - want to have free will.

They are willing to accept some limitations (they will live within the basic rules of the gameworld) but they will always want to push those limits.

And now, just to be contrary, they come in and demand a story too. They can get fre ewill any time - but they want to focus upon a story and to be given everything.

Logistically, this is a nightmare. The player can go anywhere, do anything, and yet still be provided with a tailored story? Hard task. how do we approach it?

1. They really don't know what they want

The players don't want a tailored story that makes sense, they wanna have fun. This is fine. A totally different kind of game, but fine. Put comedy characters everywhere. Throw stuff at them. Give them a gun which shrinks stuff. Let them graffitti and throw bouncing balls around.

You have to be careful, pretty soon the players will get bored. There has to be a reward there too. But that's doable too. Have an audience that goes 'oooh' whenever they do something new.

While this is a perfectly valid genre (look at those stakeboarding games) I am going to ignore it. I'm wearing my storyteller hat today, dammit.

2. Sorry, you can't do that

This is a complete cop out.

It's one that can work. Constrain your players in some obvious way. Keep them in a side scrolling shoot 'em up. Stop them walking off the side of the screen.

Now, if the player is in the mood for a non-interactive game - yeah, this'll work.

Heck, you can intersperse fully interactive bits, provided they don't actually affect the outcome of the story.

As I say, this works. But it's just poor. Hugely limits replayability, and, as games like DOOM prove - if you have good interactive bits, you don't need story.

Ditch one side of the equation entirely, and you will please more people than if you try to keep both.

3. Put the story in, let the players meet it as they choose

Well - this works. Sorta.

See the final fantasy 7 onwards titles. You pretend to have an interactive story, but actually, you don't. The players can wander about the world, but the world advances its story only when they go to the right places. This limits replayability immensly, but does let you create a wonderful story the first time through.

The problem is when the players notice, they tend to lash back.

4. Let the players create it

I am optimistic. This could work. The problem with MMORG is twofold. Firstly, deliberate greif players. Lllamas, twinks, whatever.

And secondly, people aren't content with their role.

Here's an example (from the next big thing, games aimed at householding women) Set in elizabethan or victorian times, upper class society.

It's pure fanatsy fulfillment. Everyone gets to play the 'lady of the house' all with a tragically dickens backstory, and all come up smelling of roses in the end.

So, who plays the scullery maid? Who plays the butler? For every one person who fights off their rivals to win the hand of lord mucknstuff, three people have to fight for his hand and LOSE.

Now yes, some people are bright enough to build in such little failures - but co-ordinating them is going to be practically impossible.

It's hard enough with seven people, seven thousand?

Story and Free will. The cantabrigian method

Roleplaying. There, I said it. If you're still reading, tell the people who ran away screaming to come back.

Gaming is gaming, story is story. It doesn't matter the medium. I think I can safely say that the 'cantabrigian method' of play emphasises story above mechanic, to the extent where exposing randomness or method to the players is a dirty concept.

Have individual story elements. LOTS of individual story elements.

Make sure that they all point to each other.

Your players hop from one to the other. They have plenty of free will, as long as you have an event to cover their choice.

You don't even have to be that clever. Maybe the players decide to go to a location that happened much earlier. I don't know why, perhaps they are lost, perhaps they are looking for leads, or maybe they think it's more important to the overall plot than you intended.

Fine, just run a random event there. One that they wouldn't have seen otherwise, or maybe one they missed earlier. It's still linked into a part of the plot, it still furthers the plot - it just wasn't very planned.

Now - you have to be careful not to foster apathy. If the players figure that they can go anwyhere, do anything, and you'll still show them the whole plot - then they'll end up sitting at home and watching the pretty colours.

Well - in a couple of years, this may be fine. After all, if they want to watch a movie, let them. In a few years, maybe, we'll be able to make a game that is also a movie, and they can switch modes seamlessly. But not now. Right now we have to make them play the game.

Timing is the most important thing for this. You have to have tempo. If the players are just relaxing, then the plot moves slowly. Few random events, fewer still plot events.

If the players are actively following, furthering, and creating their own plot - then go faster. If the players are fighting against your plot, then you've got them - you can go into overdrive, and watch them scuttle about, trying to head off its every advance.

Tension is another subject - it's very hard to create, you will ahve to be careful with any automatic system that it can spot the difference between a tense player and a relaxed player - both go slowly, but there is a huge difference.

Pick up on things that the players pick up on. Don't make EVERYTHING that the player focusses on important to the plot - only about 70% of them.

The thing with this is, that players like to think they are clever, and have picked up upon a small clue. Now yes, you can put lots of small clues in - but very often conincidence will do this for you. Pick up on the fact that your players think something is a clue. Throw further references in to whatever it is. Have events mention it - or not mention it. Throw away some later 'key' item, and substitute the one your players have already found for it. Watch strategy guide writers tear their hair out figuring this one out.

Success and player skill.

This is critical to enjoyment. The players want to feel that they have made an impact. Not their avatars in the game world, THEM. Puzzles, reaction speeds, they all make an impact. But don't just throw them in as an extra scene - they have to count throughout.

Success is important. A player should suceed in about 70% of the things they attempt.

That's a hard and fast rule, feel free to bend it. The important things about it are:

1. The player should succeed in most of the things they attempt.
2. The player should not succeed in everything.

The player should always get to an end of the game. But they should always beleive that their choices had an impact on the ending. And I don't mean just a trite 'will you join us yes/no' either.

And, yet again, I'm gonna advise you to fake it here.

Have a number of endings. Twelve, maybe. Make them adaptive, certain key items, people, whatever, are placed in them decided upon the foci of earlier. Don't have different 'succeed' and 'fail' endings - make them all ambiguous. This leads to sequels.

Allow the player to act during the ending scene. Maybe the player doens't WANT an end now. Maybe he plans a huge betrayal. If the player wants to blow everyone up at the end, log it.

I advise fading to black anyway, but you need to know where they want to go for the sequel.

If you have a truly great random encounter engine, then you might just be able to cope with this, but you've just used up all of your set plot. Call for a break, and get your tame monkeys to write new stuff.

And if your game has enough posisble paths through it, don't be afraid to just start again at the beginning, or midway through, but with randomly renamed objects.

Credit to James Palmer for the MMORG idea, I hope to see it made some day