INTERACTION MODEL

This section considers the design of the player interactions and their perceived agency within the game world.

After receiving some simple instructions, including an invitation to decide how each victim had died and who was responsible, players registered with the game, by texting a keyword into their handset.  At this point, they were added to the database of players. 
 
Players resolving their clues


They subsequently chose whose death to investigate – Benjamin Blight in 1777, George Bodle in 1833 or Rose Mayflower in 1916 – by texting the initial name of the spirit.  The selected spirit then sent them a text message with a clue. 

There were three types of clues – those that required a specific action and directed the player to a new location to perform it (see Example 1); those that were basically names embedded within the text, indicating that another spirit could be contacted (see Example 2); those that gave detailed information about the story (see Example 3).  The first two types of clue required a text response from the player.  If performed correctly, these increased the players’ score. 

Examples

1.       “Greetings from George Bodle. I used to have a farm near here, in Plumstead.  Worth a pretty penny.  They all had their eyes on it.  Look around Artillery Square to uncover the perfect weapon for murdering an old man.  What is it? ”  Answer: Arsenic

2.       “Henry Adolphus, for the Crown.  We were relying on the evidence of James Marsh, the Chemist and Frederick Pound, the apothecary.  It was a disappointing verdict.”

3.       “Twas the year of Our Lord 1843.  Found guilty of fraud and blackmail, I was sentenced to 7 years hard labour and transportation.  I knew they could not try me twice, so I confessed to an older deed ... I, John Bodle the Younger, was the wretch that murdered my own grandfather with poison.”

Exploring the riverside

Often a message contained more than one type of clue, so players could choose their direction of progression.  Clues also alluded to information that would be useful, and players who remembered this would be in a better to position to take advantage of physical clues when they were found.
The structure of each game was different, but they all consisted of branching storylines and multiple opportunities for choice.  There were on average six directions to site-specific information (correct response worth 5 points) and nine embedded keyword clues (worth 1 point).  Several trails led to “dead ends”, so there was no obvious denouement. 
Each game progressed in this way until all the texts had been sent or players reached a conclusion with respect to murder or misadventure and who was responsible.
Interaction between competing sets of players was redundant as a game design feature, although it would obviously occur between players on the same team.  
Map showing ghost positions