One of the great challenges of our genre is the ability to tell a story in real time, exactly as it unfolds. This is a powerful device, but it's also what is responsible for a lot of the logistical problems encountered. That's because in the conventions of alternate reality gaming, events *have* to spill over into the physical world, and that introduces a huge element that the storyteller can't control.
An example? Sure. Let's say you were making a game that cleverly hinged on getting players to answer payphones at GPS spots you had coordinated for them. (I'm looking at you, Beekeepers.) There are lots of real-world things that could go wrong with that kind of scheme. Like, say, broken payphones, or bad addresses, or hurricanes.
Now, it's true that these kinds of uncontrolled real-world events have led to some of the most fantastic player experiences in gaming. So how does that happen?
First let's take a look at what goes into writing this kind of event. In general, you don't want to put all of your eggs in one basket, just in case an uncontrollable event transpires. A sensible puppetmaster will plan for this sort of contingency. That's why there was more than one axon/payphone in Bees, for example.
There will be times, though, when you simply can't do a lot of contingency planning, or where all of your planning will come to naught. Let's say, for example, that your story hinges on players talking an actor into performing some action for them over the telephone. What happens if the actor contracts pneumonia that day, or the telephone number is disconnected, or the server with the clue to the phone number is struck by lightning? These are all things that could happen, sure, and I bet some of them have. This is the sort of thing that can keep a puppetmaster awake in the dead of the night.
Given the real-time nature of the story, it's generally not possible to throw up your hands and try again a week later -- especially if you're well into that story arc before your uncontrollable event hits you.
Is the answer to not take these risks, then? Absolutely not. Your audience can detect wishy-washy writing. If you're afraid to ever roll the dice and try to do something great, then you'll be lucky to end up with a story that's even OK. The key is to maintain flexibility, and to treat your story like a Bonsai tree.
You can't perfectly control how your story will develop, or how your tree will grow. You may have a general shape in mind, and all you can do is nudge and trim and watch what happens, organically, all on its very own. This kind of responsiveness is a large part of the spine-tingling appeal of the cross-media game -- on both sides of the curtain.