The Millennium PAO - an 8-digit and 2-card system


For years this was merely a dream of mine: Now the first step is done for my millennium PAO system. Basically it is a normal PAO but with 3-digit images instead of 2-digit images for the people and the objects. Using 1.000 different actions seems impractical, so I will only use a 2-digit list with 100 actions. This creates a cluster-image with 8 digits encoded (3 + 2 + 3). This will be perfect for a 40-digit row in memory competitions because you can memorize such a row with exactly five PAO images.

The big question in developing such a system for me was the codification. If I would rely on the common Major System or the Dominic System I would reduce the possibilities for good images to only a few choices. Especially when it comes to the 1.000 people it seems like a bad idea to simply take a name that fits. It is difficult enough to clearly imagine random people from a 100-list so they have to be very well chosen. Even if you create a 1.000 object list with the major code and then add a person to each object with natural associations, it would feel like a compromise I don’t want to make. The only reason why we use a code in the first place is to help our memory. But I rather have more effort learning this system but have the freedom of choice instead.

That is why I skipped the codification completely to create a perfect list of 1.000 people. I wanted characters I know and can easily imagine and all should be distinguishable from each other. Since I am already an experienced memorizer I thought it should be possible to learn such a list without any code. But it still should have at least some logic in it. So I sat down and created basically ten major categories of people like cartoons, movies, historic figures and more. Then I created 100 categories within these major categories with ten people each, like the Simpsons, the Muppets or Star Wars. For every person I googled a nice picture and put them all into order. You will find my video playlist with all of them below.

To learn this huge list I simply used my old 2-digit PAO system. I didn’t use the old 100 people but the list of 100 actions and 100 objects. I took the new person for the number XYZ and created a PAO scene the following way: The new person is doing something (XY) with some object (YZ). I did use the middle digit of the three digits twice - as the second digit of the first image and the first digit of the second image. That creates several advantages: Firstly you can memorize the entire list of 1.000 people with exactly all combinations of 100 actions and 100 objects from 00-99. Secondly each action encodes an entire category of ten people which makes it much easier to reproduce. And thirdly since you use the middle digit twice, the images are entangled and each action will be ultimately linked to only ten objects - that will help you to remember the right person a lot.

Examples of this evolution:
Step 1: 2-digit Major System: 74 = car; 40 = rose;
Step 2: 2-digit PAO: Michael Schumacher - racing - car; Sleeping Beauty - sleeping - rose
Step 3: 3-digit Person: 740 = Charlie Sheen

To memorize Charlie Sheen as the new #740 I split this number into 74 and 40. Charlie is the person § which is racing (action #74) into a pile of roses (object #40). That is a normal PAO image we are used to deal with (supposing you already using a 2-digit PAO). Later I can reconstruct Charlie Sheen by asking myself who is racing into roses.

That is actually pretty easy to do so. After some struggle of motivation I forced myself to learn a new set of 100 every day by just watching the 5-minute video I created for all major categories and memorize these stories on the fly. After the first run I was able to recall most of them immediately but with a relatively weak memory image. Then I would have been able to use a flash card system like Anki to train them right away. Though I waited to do so until I memorized all ten sets of 100 because I wanted to train all of them at once in a random order (because it is much easier if you train just all images from 500-599 for example). Now it is very comfortable to train all my people with Anki on my iPhone or iPad. Some of them are still difficult but I just have to improve the mental PAO image and I am fine the next time. Many are popping already right into my mind when I see the associated number. I finished memorizing all of them a week ago.

Well that’s it so far. I am creating the list with 1.000 objects already and it takes much less time than the people. I will keep you posted.

More about my system on my website:

Here is a link to the list with all names and categories:

Here is a link to my YouTube playlist with all 1.000 people.


Millennium PAO

I am finally done with my 1000 objects. And not just that: I put them all into 1000 collages together with my 1000 people as well as my action/object overlap images (see above). Everyone who wants to have a closer look at my system, can do so on my website, where I created a really neat presentation for ALL of them.

Millennium PAO

Just like the people I devided my objects into categories of 100. Each category is a related subject. Most categories are additionally encoding valuabale/interesting information.

The categories are:

000-099: The Periodic Table of Elements
100-199: Food
200-299: Countries of America, Europe and Middle East
300-399: Countries of Asia, Ocieania and Africa
400-499: Commodity Items
500-599: Toys, Sport Items, Instruments and Electro Gadgets
600-699: Artefacts, Tresures and Legendary Items
700-799: Plants
800-899: Four Animal Alphabet lists with mammals, birds, reptiles/amphibia/fish and dinosaurs
900-999: Esoteric Wisdom

I admit that the objects are not the easiest but they are beautiful and useful. It is fun to learn them and even more to use them. I am happy to have them all in my head and train with these amazing contents.

1 Like

It’s really really nice, Flauwy. Thanks for your great job.

Wow, very cool system, Flauwy. At present, this is only for numbers, correct? You mentioned at some point that you’d be working up to a full 2704 image system for cards by adding extra images. Have you already done that as well?

The cards are coming next. I was considering back and forth whether I am going to do it or not. But now I will!


That’s my way of the ninja…


Webmaster’s note: This was a question about: how to find actions when memorizing numbers with the PAO system.

Check out the page on PAO systems.

You just use the major system like this:

05: S L
Person: SauL
Action: SLides
Object: SLime

Or, you could match the person with the AO, like this;
Einstein writing on a chalkboard, the pope praying to a painting of jesus, etc


rock-lee cock-island
this is perfect!
i will send me an e-mail with that picture!

I am half way through the next step: the Double Cards system. It fits perfectly into my 1000 people and objects system. Since that is category based with 10 images each, I only need to add 3 fitting characters to make it a full 13 suit. I create 200 of these groups, using my entire 2000 images. Finally I add 8 new categories with 13 each to make it to the 2704 images. Each group represents one out of the 52 cards plus the full suit of the second one. Assigning the fitting numbers 1=Spades (one stem), 2=Hearts (two halfs), 3=Clubs (three bubbles) and 4=Diamonds (four corners) orders the four suits. For the index I keep 2-9 as they are, transform Ace into 1 and 10 into 0, which I find most intuitive. J, Q and K become 10, 11, 12.

Now I multiply the index value with a four (because each first card can be combined with one out of four suits with the second card, this is the same for all groups, except the eight complete new ones). Then I add the value of the suit to my result. This gives me a 2-digit number that I translate into the according action. For example:

10 of Spades (1st card) plus Hearts (2nd card) = 0 x 4 + 2 = 2.
5 of Spades (1st card) plus Diamonds (2nd card) = 5 x 4 + 4 = 24

In my Millennium PAO all actions define a cateory of ten people and ten opbjects, never to be used again for any other category. Therefor reminds me the action immediately of the entire group. Cards Ace-10 I translate right away with ease. J, Q, K are now to be learned, which feels easy enough.

When I reach King - 1st card and Heart - 2nd card, I have a round 50 with 500 images of my old system and 150 new images. The last two groups for King of Spades and Diamonds are new. Because then I can freshly start calculating anew with the next suit for the first card. Up until here I assigned 676 images for the entire first suit for the first card. I used 52 groups of 13 images each, and reused exactly 50 of my categories. But only the people of each category, not yet the objects. So alltogether I used 25% of my system for 25% of the Double Cards, by filling in the gaps.

Now the first card switches to Heart = 2 = red. From here on I start over with my calculation result with the cards but instead of people I am now using the objects. This way I will always know already with the first card, whether I am dealing with an object or a person: black = person, red = object. So now I fill another 50 categories plus two new ones and have 50% of the Double Card System. I used all people and objects from 000-499.

The remaining 50% will be executed in the same way, but to the calculation I add a 50, ending up with a number between 51-00. Again I will do the people with the black Clubs and the objects with the red Diamonds. Finally I used up all my 2000 images perfectly in exactly the same logic of assignment to digits as I have created them for the number system. And all new 704 images fit in the categories as if they always were meant to be there in the first place.

The calculating part is kind of slow in the beginning, but this is only a temporal problem, until I recognize all 208 categories on the spot. I think that will not be very difficult and I will focus on that first, to get fast with this system.

Here are a few exampls of my new collages:





Bateman is right. I created my actions by associating a fitting action to my Major System list, which contained out of mixed people and objects. That way you can learn it in a day.

You can read my article about the PAO system.


Great idea!!
You are becoming a memory teacher!
You are a sensei of taijutsu memory!

You discovered a secret prohibited ninja technique!
If I become a teacher I will tell my students about your memory ninja techniques.

Not sure if you are still active… but I’d like to see if I understood your system correctly. I am a visual thinker, so reading a text without examples is quite hard for me to understand.

You have 1000 persons, each one with a number between 000 and 999
You have 100 actions, each one with a number between 00 and 99
You have 1000 objects, each one with a number between 000 and 999

So a number like 7407400132940087 you would break into two pieces of 8 digits each, so 74074001 and 32940087
740 - 74 - 001 would then be something like ‘Charlie Sheen - racing - helium’
and 329 - 40 - 087 would then be something like ‘Shenlong - sleeping - radium’
(if I got your actions and objects right)
Those two chuncks you would then place on your Loci, like Charlie Sheen driving a race car with helium balloons as tires down the driveway and Shenlong sleeping on a watch with radium-covered dials in the front yard, or something like that

Would that be about right? or am I missing your point?


I’m not sure that there are still people alive in this topic. But I still want to leave my question here. What about binary numbers? I use standart PAO and my PAO images contain 18 digits. This is an advanced method of Dominic O’Brien`s system, who used 12-digit blocks.
Given that Millenium is a PAO system that uses a thousand images, one Millenium image must contain just a huge number of binary digits.