From Struggle to Speed: My Insights to Faster Memorization

Memory Training Journals
1

Hi all, I’m Rado.

I’d call myself an intermediate memorizer, still figuring out what my mind can really do. I started this journal to track what I learn about different ways of visualizing, compressing, and storing information — and maybe some of it will click for you too.

Recently, I made two small changes in my mnemonic system (00-99 PAO) that made a huge difference:

  • Visualizing people from behind. I used to struggle to visualize the people from my PAO list. Faces were hard to hold in my mind. Then I tried picturing them from behind instead — their shoulders, the way they walk, how they dress, what shoes they wear, etc.. Suddenly, they became clear and easy to see. Suddenly I didn’t needed to force myself to visualize their faces no more!

  • Switching from PAO to PA. I dropped the object part and went with just Person-Action, as Dominic O’Brien suggests in How to Develop a Brilliant Memory.

These two tweaks simplified everything. Now I see the person and the action instantly — no clutter, no forced scenes. Sure, I use more loci (26 per deck, 2 cards per locus), but that’s fine. Memory palaces are built for that and they are meant to be used! (a favourite quote of mine from TheHumanTim)

The result was that my speed nearly doubled in just a few days. Memorization feels smoother, quicker, and less of a grind. More importantly, I feel more confident. I can trust my memory to show up when I need it instead of constantly second guessing myself.

7 Likes
2

If it may help you, as I had problems with the Object part, what I did instead was adding another person (PAP), that way you ‘save’ loci and keep the efficiency of having a Person. Seeing my mom (33) dancing (02) with the Hulk (59) is a pretty memorable image, much more than dancing with a… hammer?

I still need 17 loci for a deck, for in one locus I can put my mom (Jack of Hearts) dancing (Queen of Hearts) with the Hulk (9 of Hearts)!

6 Likes
3

This is a great idea! Will certainly try it.

What is your personal best? And did you eventually switched to PAO or you still are doing PAP?

4

How would you distinguish between 33–02–59 and 59–02–33? Both of these are the same two people dancing together.

(Whereas for some actions like “throwing”, it would be clearer which person is the object of the verb.)

5

Well I have 5 sequences, for example man, woman, animal, action and object.

The important thing about the faces is that they must have prominence, the images are shadows, you will not see a 4k image in your mind, if the face you imagine it flat and without emotion, ie a shelf and this does it with everyone, well, the brain gives zero importance to this, since a face has many more details than an apple and consumes more resources, but there are ways to force the brain to see it and that is facial expressions, angry, sad, embarrassed, etc., that every action committed has an expression, an anxious person touches a ray of light inside a cell and this with a sad face and with tears in his eyes… etc., it takes more time to describe this than to imagine it.

Because of the facial expressions:

The brain processes facial expressions quickly to detect emotions, with specific regions such as the amygdala and posterior superior temporal sulcus involved in emotion recognition.

3 Likes
6

Update (13/09/2025): I’m finally sub 5 minutes!

still very slow from competition standpoint but I visualized myself a few months ago where I was doing 15 mins per deck and sub 5 mins is a big accomplishment for me.

I will keep on grinding and keep this post updated in time!

Getting used to PAO as well hopefully soon I can replace my PA.

Thank you everybody for your support and advice!

4 Likes
7

My personal best is 85 secs, but that’s a problem of my lack of sufficient training, not the PAP :')

And yeah. Stick to PAP.

Edit: today, 14/09/2025, I memorized a deck in 84 secs.

8

Not at all. The first P and the second are not the same.

If 11 is Donald Trump, the Object (a second person for me) is Hillary Clinton.

So 1102XX is Donald Trump dancing with XX, but XX0211 is XX dancing with Hillary Clinton.

PAP works in the same way as PAO, just substituting the Object with a Person, but the second person of each XX is different, depending on the position.

3 Likes
9

The difficulty I’d have with this is having two different people for the same number and needing to remember that one is the “primary” person for that number and the other is the “secondary.” It seems to involve extra focus needed to parse that out when in the moment. And if one is related somehow directly to the number and the other is related to the person, there’s another mixed translation method to deal with.

I think duplicating a category would work fairly well with an OAO, where the first object in the scene is the “character object” that for all intents and purposes represents the “person” and is the thing that performs the action. This is an animated version of the object that is directly associated with the number somehow. The object that occupies the traditional object position in the scene would be the same object but its usual inanimate form that gets affected by the action.

So lets say 77’s object is a cake, and 77’s action is kicking, and 09’s object is a bar of Soap.

77-77-09 is a living cake character kicking an inanimate bar of soap.

09-77-77 is a living bar of soap character kicking an inanimate cake.

77-77-77 is a living cake character kicking an inanimate cake. No confusion even though it’s the same object.

09-77-09 is a big bar of soap character kicking a bar of soap.

With this in mind, there is no need for a “secondary” object that you also need to learn to associate with the number.

1 Like
10

It’s never been a problem to me.

2 Likes
11

Thanks Tim!

I owe you a lot man! Thank you for your kind assistance in my topics as well as answering my DMs <3

1 Like
12

i could try a pa (o?) like that as well, because the object part gets reduntant, good idea :smiley:

13

Update: (22/09/2025)

Sub 3 mins for the first time of my life!

I optimized my PAO. Eliminated most of my objects and connected them to my actions. This worked like a charm for me.

Example:

Old PAO image:

83 (Dominic System: HC)

Person: Hillary Clinton

Action: SHOVING ENVELOPES

OBJECT: BLENDER

New Optimized PAO Image:

Person: Hillary Clinton

Action: SHOVING ENVELOPES

Object: ENVELOPES (I visualize a stack of envelopes)

Another example:

Old image: 55 (EE per the Dominic System):

Person: Emilio Estevez (from The Mighty Ducks)

Action: PLAYING HOCKEY

Object: CHEWING GUM

Optimized image:

Person: Emilio Estevez

Action: PLAYING HOCKEY

Object: HOCKEY STICK

Basically connected all of my actions with their most obvious objects and I got faster very quickly from sub 6 mins to sub 3 mins (roughly 50% of constant improvement in time)

Got this suggestion from @climbformemory and his PAO system. Most of Nelson’s PAO images are following that exact structure.

Thanks Nelson!

4 Likes
14

Thanks! I appreciate this post.

PAP sounds so much easier.

With PAO I’d have multiple entertainers but eventually every singer has a guitar or a microphone.

If I have multiple scientists there’s only so many test tubes they can hold. :rofl: objects become repetitive very fast.

2 Likes
15

PAP is a really great method - which should hopefully be easier to avoid swapping the Ps (as separated by an object).

I do PO at the moment and wondering if I would like to change that to POO, or POP, if I want to change it up down the track.

Currently can do around 44 - 46 cards in 1 minute, I want to get to 1 minute as my next goal and see if i can go faster than that :slight_smile:

PAP, I think, is what Don Michael Vickers (@Bigdonnyv ) & POO is what @vishvaavishvaa1 does. Vishvaa won the last Slam on the weekend - the Asian-Oceanian Open 2026

16

@waverider Yes, I does POO, but I still have few actions in it though, like PAO/POO for some images, but I would still not recommend this method it’s slow, best option is trying to do a full PAO.

2 Likes