I have 2 versions of a linking system with person, action and location.
Basic preparation:
You can start by creating a person and action in a location for every 2 digit number. For example, 23 might be Michael Jordan dribbling a basketball at your high school basketball field. Because the 3 different number translations form a unity, you don’t need to learn them seperately, so learning this takes not much more time then learning a 100 person list.
Linking method 1 (person + action in location):
You start with a location, then you visualize a person in that location based on the 1st 2 digit number, then that person teleports to the location determined by the 2nd 2 digit number and performes the action connected to that location. At that location another person determined by the 3th 2 digit number appears and teleports to yet another location determined by the 4th 2 digit number and performs the action of that location, and so on. So in short it is person + action, but the action is always performed in the most appropriate location.
Linking method 2 (person + location + action):
Here you also start with a location and person based on the 1st 2 digit number and that person also teleports to the location determined by the 2nd 2 digit number. However, instead of performing the action of that location, the action is now determined by the 3th 2 digit number. After this another person appeas determined by the 4th 2 digit number and so on.
What I am going to do:
I have been experimenting with an in hindsight to complicated system called 2 digit 3 category system with easy 10 digits per location and I have taken from this system the person and action combinations and like what I usually do to learn a system, I have put all those combinations in locations in my memory palace. So I have all the basics for the above described linking methods and I just need to improve my translation speed and the quality of my visualisations. So in short, I am going to try out both methods with some serious training.