Speed memory experiences and problems

Recently I found Ramon Campayo’s memory simulator v7 by chance, and I really love the functions it got.
Which I don’t quite comfortable with is the font.

I memorized a 22 digits number in 1.1 to 1.3 seconds using this app, and most comfortable with 1.3 seconds. Don’t know whether the grade is affected by the drug, I think it is a little disappointing.

video demonstration: https://www.youtube.com/watch?v=ZdGrYjqywGs

And here are some questions:

Will my grade improve through training?

How can I improve my speed memory grade?

Is there any special technique to achieve it?


That’s very impressive!

What’s method did you use to memorise it this fast?

I can memorize 10 digits in 1 sec .

Basically , If I have to remember 10 digits in 1 sec , I am not using any number system and memory Palace.
I just focusing on numbers and when I blink my eyes , I can see the 10 numbers.

I practiced this in speed reading memory app.

1 Like

Nice to meet you Erol!

I read through them.
I still didn’t find any good method could help to memorize faster.

1 Like

With 4 digits chunks I can memorise 12 digits in 2.5 seconds.

That s usually fast enough to memorise a quick phone number.

Competing is a different level . I mostly use mnemonics for day to day life.

I have a book I bought on kindle of ramon and it’s very interesting. I have to reread again to refresh.

1 Like

Thank you Erol! I will check it out later

1 Like

You could diverge the processes you are doing for a small boost.

For example you can read 20 digits then store another 10 in your visual working memory and encode some extra digits at the same time.

You could try learning to read bits in parallel, which has been successful for me lately.

You could even learn larger chunks for memory or speed up how quickly you are reading for speed.

Ramon is stated to do the diverging I referenced, so I think this would be easiest for the largest gain.

1 Like

Long time no see, Nagime!
I will take your advice. Thanks!

Most of the time I memorize just like that.
My visual memory didn’t vary much, but my reading speed of numbers dropped a lot since I left hospital. Now I’m more depend on the visual memory.

1 Like

Hi Nagime, sorry to disturb you again.

I used to read book in parallel for a period of time, but later I found it was much slower than just read straight forward, line by line, and I finally gave it up.
Now the same problem occur when I try to read numbers in parallel. Could you please explain how you tackle this problem? Thanks.

Generally when you read numbers in parallel to begin with, you usually can’t do it without training.
After training, you can do it but it’s too slow to use.
It’s usually when you can do this with 3-4 sets of numbers slowly, that 2 sets of numbers become quicker than 1 set of numbers on its own (using numbers as an example I initially did this with lines too).

I at least find that the best way to train speed quickly is to increase the parallel count as opposed to increasing speed directly. I find that if I use a lower count than my maximum, I perform much faster. There should be a way to make it still quicker without doing this, but this has been my experience.

You should also use a stopwatch, I found that parallel, after being able to do it with 3 sources was 120% quicker with 2 sources than 1 source for example. This at first made me think that it was still slower because I was working at 60% of a 1 digit rate just doing 2x the amount, but it wasn’t at all the case.

Hope this helps, it’s generally also important that you are not task switching but truly doing this in parallel. Sometimes task switching can become quite fast so it’s important not to use it on accident.

Thanks for you reply, Nagime. I will try it.

By the way, what is “task switching”? (Sorry if this question sounds stupid, my English is bad)

Task switching is in a literal sense switching the task particularly the attention you are giving to the task.
Task being what you are doing, for example reading.
Switching is changing from one thing to another.

Task switching usually happens when you try to do something in parallel without first learning to do it slowly.

For example when you are reading 2 lines and “task switching” you may read
“This is the first line
The second line is this one”
At the same time, in the form of:
“This is the the second first line is this one” as opposed to truly parallel.
In a simple sense it’s when you do something then do something else usually quite quickly, serially, rather than in parallel.

As you can get faster at this, you may not even realize if you are doing it. This is quite crucial because you can read 2 lines in a seemingly parallel way but only be task switching in reality.

For example, when you are much quicker at this, you may do something like, reading 1 word let’s call this X and another word Y. You may read this as : start X start Y, X ends ,while Y is ending; rather than start X&Y at the same time.

Thanks Nagime! You explained this explicitly. Love it.

It seems that one ‘line’ is missing from this sentence

Are you using parallel method for verbal memory or visual memory or both?
Parallel method works well and fast for my visual memory, but for verbal memory I need much more time to process. l’m double-handed, therefore I could practice processing information in dual every time I write. But even I practiced it, the speed is still not not fast enough to handle memory task.
In my case visual memorization in parallel is 100%+ slower(don’t exactly know, but only slower) than memorization straight forward.
Screenshot 2020-11-23 091343

It seems that Ramon uses parallel method for both visual and verbal memorization according to the screenshot. If I’m not mistaken he used verbal memory for ‘138567’ and visual memory for’6316441’.
He split numbers into small parts, Is parallel method works better with smaller chunks?

Thanks for your patience.

sorry it’s 'verbal memorization here, I made a mistake

There is indeed. This is a good example of what can happen with task switching when you process it this way since both lines have “line”.

I first used the parallel method for verbal memory. I then afterwards used the parallel method for visual memory. However, to be clear, you do not have to use it for memory only, it’s possible to calculate in parallel although this is significantly exhausting. I tested this with some simple additions (visual parallel) and have had positive results.

The visual parallel is indeed easier than the verbal one(not easier for calculating). In order for me to do the verbal one without interference I use different voices for each line which helps, distinguish it more. I don’t think this is required, but it made it easier for me. I found it easiest to train by first using a second language since I am bilingual, this prevented me from task switching.

Double handed as in, being able to write with both hands at the same time or being able to write with either hand? I personally trained my left hand since I am originally right-handed so now I am ambidextrous, I also practiced writing with both hands recently at the same time and found decent although not as quick results (particularly since I can’t write with 3 hands to speed up 2 hands). I find that there is some transfer (benefits in general), for example if you can parallel verbal, you noticeably become better able to comprehend multiple audio sources at a time compared to prior. Though it’s not quite as broad that you get the full scale benefit.

Parallel with 1 verbal and 1 visual together rather than either exclusively, is much easier and doesn’t take too much training. From Ramon’s books I believe what he did was something like:
Read verbally ‘35477213’
Visualize images of ‘8567’
Glance with visual working memory at ‘6316441’

So he did not even much of any parallel, other than the visual one of visualizing and keeping working memory active. It’s generally much easier to do different tasks in parallel even if they use similar components. For example it takes very little training to do 1 verbal and 1 visual in parallel.

In some cases, to my experience, it’s easier although harder to train, to do parallel as a combination of visual and verbal. For example you can visualize text to process the text verbally in parallel.

The reason for the chunks, is because it’s easier to visually recall a chunk in working memory in this form, for the latter and you make less errors in chunks (particularly if you visualize images).

Parallel doesn’t actually make much of a difference in chunks. At best keeping the chunks different so you clearly know where to start in parallel.

It’s indeed something that you do have to train. If it is difficult that usually means you get the training benefits quicker, if it’s too easy you should increase the amount of digits rather than attempt to increase speed would be my advice. I tend to test a lot of things so I tend to make a lot of discoveries, so I personally think it’s good to try a range of things to see what works best for you.

Thanks for your reply, Nagime.

When I’m calculating easier arithmetic tasks , I mainly use visual parallel, for tasks tougher than 8*8 I verbalize a lot. But there is no clear distinction between them, sometimes my visual and verbal memory will rewrite each other-----It’s quite hard to explain it clear, but I think my system works a little bit different from yours.

Simultaneously. I mainly write in Chinese. When I was young I was double-handed, I use my hand randomly, eating , playing balls, writing etc. But back then I write separately, until recent years I found writing simultaneously quite fun and have the potentiality to write faster. Now I write with them without intermission, It’s all about muscle memory I think.

In practical writing with both hand at same time will face some crucial problems, like alignment and stroke calculation, which make it impossible to write 100% faster than writing separately, and inconvenient in many situations.

Totally agree with that.

I don’t quite understand this. I think it’s not your fault Nagime, I think my memory pattern works a little bit different from that, this method doesn’t really fit me.

This is indeed true. I will try to do my visual memorization in parallel.

Is there a lot of people faster than Ramon Campayo in speed memory? Why don’t they attempt to challenge him?

I think you once stated that you silent read, if you are silent reading then this can become quite notable. I have trained silent reading too separately from parallel, it kind of makes the gap between visual and verbal a lot less clear (although it is still there).
I don’t particularly use one way to do something, I’m always learning and trying new methods. So I believe I know what you mean.

It is quite good training.
The principle should also still apply to any general parallel tasks so if you are doing similar training you should be able to also do this with say verbal or so for similar results.

I’m not sure if this makes it easier to understand but I have seen that he also chunks verbal as ‘354772’ and visualizes images as ‘138567’ and glances with visual working memory at ‘6316441’

I’m not entirely sure what part is not understandable, but what ramon does is different than what you have to do of course. His memory method is entirely aimed at short term results, if he wants to actually remember what he is doing for a longer while he will only use images.

I think the basic idea of his method is that

  1. You can always still use working memory to hold digits aside from your encoding.
  2. You can do something verbally and do something entirely different visually.
  3. Working memory and encoding are different.

I think, while there are people faster than him, the drive to beat him is low. For starters the records he holds are not common records so most people actually do not know about him or his records.

Secondly I think that his records are perceived as very quick even by people faster than him because they assume he is doing this the same way people for example do cards. So it is as if someone is saying they are memorizing 20 cards in 1 second. The main issue with this is that he uses, working memory (verbal and visual) so his results are going to be higher with a starting boost. In each case, for example, it doesn’t take much to memorize 7 digits visually in a second, neither does it to verbally remember 7 digits. Visualizing 3 image encodings in 1 second may be a bit more taxing but, even from my own experience a burst of 3-4 is very different than maintaining this rate for 100samples and very feasible. So in each case he is doing what he can already do, but doing these things at the same time for different things. Therefore it’s very doable to beat his record, in fact you have even actually beaten his record for example and you are using an entirely different method.

I got what you mean, Nagime. But It seems that adopting his method will take me a lot of time, I’m not even sure I can do that fast.
I will try it anyway. It’s always interesting to try something new.

Thanks for all of your valuable advice and explanation, Nagime!


Hi, @Nodas
You have mentioned in a previous thread that Ramon could find some patterns in numbers, could you please explain it a little bit? Thanks.

It is not patterns in the numbers per se, (like figuring the properties of the number).

But rather, an extremely efficient chunking technique based on the science of human central and peripheral vision.

I have that Speed Memory simulator for many years (in order to train reading numbers fast for the calendar contest).

But certainly, I am not an expert on Speed memory, because I usually focus on the operations between numbers.

If you ask the top Speed-mnemonists , they may tell you which chunking does Ramon Campayo use. But beware. This chunking may be effiecient for him, but not for everybody.

We all have different preferences on how do we prefer to read the numbers, vertically or horizontally. And how many digits at time, or how many chucks and subdivisions.

1 Like