Yes, I’m posting a card system, again… After my first 3-card system, I came up with a second one, and then a third one (the best so far), and now I’m going to share the first 4-card system ever created. I recommend you to read the post of my third 3-card method before reading this one.
There are 52x52x52x52 = 7,311,616 quadruples of images, so most people would think it’s impossible to have a system for that because of the exponential increase of combinations. However, after reading @gavino ‘s Massive Memory Palace and @Nerto ‘s Inverted Memory Palace, I felt inspired to build a 4-card system.
In order to apply this system, we need a memory palace with 13 rooms/regions, each containing 13 loci. For each of them, we are going to divide into 4 subloci: front left, front right, back left, and back right.
Now here is the trick: for each sublocus, you need to create a memory palace with just 13 loci. For each of them, we are going to divide into 4 subloci again: front left, front right, back left, and back right. In order to connect the sublocus with its respective memory palace, you’ll place an image that reminds you of that mini memory palace in its respective sublocus in the main memory palace. To understand this better, I recommend you to read Gavino’s Massive Memory Palace post.
The last thing you need is a 2-card system. To exemplify, I’m going to use the Double-2-Block System.
In order to encode, you’ll go to the room that corresponds to your image (if it’s the first image, then go to the first room; if it’s the second image, then go to the second one; and so on). Now, go to the locus that corresponds to the first card’s rank (if it’s A, go to the first locus; if it’s 2, go to the second one; …; if it’s Q, go to the second-to-last locus; and if it’s K, go to the last one). In that position, you’ll encounter 4 subloci, each corresponding to a different memory palace. The right one will be decided by the suit of the first card: if it’s 
, then go to front left sublocus; if it’s 
, go to front right one; if it’s 
, go to the back left one; if it’s 
, go to the back right one. At that sublocus, you’re going to find an image to remind you of the memory palace you’re looking for, enter in it.
Inside this mini memory palace, you’re going to place your image in the locus according to the second card’s rank: if it’s A, place it in the first locus; if it’s 2, place it in the second one; and so on. To choose what sublocus exactly you’re going to place the image, take it according to the second card’s suit just like we did before.
The image itself will encode the last two cards. To build it, we are going to use the Double-2-Block System, but it requires a way of distinguishing red-first and black-first suit-pairs. To do so, we can apply the Agent-Observer strategy just like in my third 3-card system (read that post to understand it better).
Memorizing the cards would be relatively easy; however, recalling could be troublesome: you’d need to peek inside all 13x4=52 mini memory palaces of each room. To make matters easier, I would recommend you, instead of creating 52 independent mini memory palaces per room, to build them based on scenes from the website Animation Screencaps. This way, each region/room of your main MP will be related to a different movie, making the navigation much easier.
What do you think of this system? In my opinion, it’s certainly more difficult than my 3-card one, but it’s still not as impossible as most people believed.
