The easiest way to describe message stax content is that it should be “evergreen”. That is to say, it can be used anywhere during a scenario as a reaction to the emerging plot but not creating or in conflicting with it.
Writing stax is as straightforward as writing content in the MEL and you can even import from CSV or Excel. However, you should be aiming to create libraries of stax so that you can very quickly create exercises by reusing stax you have.
To allow the greatest reuse of stax, although they need to be written in the same storyworld, write them to be independent of the scenario (i.e. independent of the plot) and independent of each other. What do we mean by this?
Stories and Storyworlds
A story has a beginning, a middle and an end. A storyworld is all the characters, locations, props, language, tropes that supports the story. The storyworld of Harry Potter contains wizards, wands, English boarding schools, muggles... but the storyworld isn't a story. Harry Potter and the Philosopher's Stone is a story.
The storyworld of James Bond is casinos, gadgets, exotic locations, Martinis shaken not stirred. Dr No is a story and so is Moonraker.
The diagram below illustrates the idea that stories exist in a particular storyworld.
When you write your stax it's completely fine to write them as part of the story and to be tightly associated with the scenario in the MEL. But for the greatest reuse, write them as part of the storyworld so that you can re-use them over and over with different scenarios. Think of the scenario as the blueprint or foundations of the story but the story itself will emerge through (1) the actions take by players and (2) the stax content you publish as a consequence of those actions.
Making Stax Related but Not Dependent on the Scenario
A simple memory aid to have in mind when you write your stax is "consequences and not causes". That is, you want to write the stax content so that it is not always tied to a particular scenario event: you want the flexibility to publish the stax at any time and have the players assume the content relates directly to their actions.
If the stax content describes a reason then you'll have to specifically wait for that reason to arrive from player actions. This is totally fine so long as you recognize it limits the use of the stax.
If you need to report deaths, have your stax title say "Deaths" and then write reactions to the deaths. Don't say how or why the deaths occurred.
If it's important to know how the deaths happened, create a new stax with content on the event that caused the deaths but don't include any information about death.
When you run the exercise, blend two (or more) stax by publishing them at the same time.
Now you might get something like:
"2 people died right next to me" /death stax"Earthquake is biggest the country has seen""The death toll keeps rising. #heartbreaking" /death stax"This has to be the worst #earthquake in living memory, or?"
You can see here that taken together our minds create meaning from the blended feed and will assume that the deaths are from the earthquake. Of course this is our intention but now those same "death tweets" can be reused in other scenarios
"2 people died right next to me" /death stax"Bomb shook the building, glass everywhere""The death toll keeps rising. #heartbreaking" /death stax"Terrorism has no place in modern society"
There's no requirement to use Smartwords with stax (as it adds an additional layer of complexity) but the Pattern of Life feature was inherently designed to be used with Smartwords as these can allow additional portability and re-use of stax.