History-Dates Technique for Three-Digit Numbers

I’ve recently been reading a German memory study from 1987 that compared the efficacy of two different approaches to memorizing numbers, both using a memory palace based on 30 major sites in Berlin. (For those who’d like to look it up, it’s called Mnemonic Training for the Acquisition of Skilled Digit Memory by Kliegl, Smith, and Hechhausen.)

They looked at two test subjects. One learned how to encode numbers based on the Major system. (They call it the “Figure” system.) And the other learned a system I’ve never heard of before: “History-Dates.” I’m curious if anyone here has heard of it or uses it.

It would be a fairly challenging system to acquire. The idea is that you would memorize an historical event associated with every year from 1000 to 1999. So, yes: 1,000 historical dates. For every grouping of three digits that you need to remember, you would imagine a “1” in front of it, to make a four-digit date. You would then store the corresponding historical event in your memory palace.

Here’s an example. Say you’re wanting to memorize pi in 3-digit chunks, ignoring the initial “3.” So you’d start with [3.]141592. You’d take “141,” add a leading “1” to it to get a four-digit year, and end up with 1141. You’d then take whatever event you’ve previously memorized that happened in 1141 and put that in the first location in your memory palace.

This technique seems highly creative. But I can’t imagine anyone embracing it, since it does require memorizing 1,000 historical events and can’t be easily adapted for image creation on the fly, as Major can. Since I’d never heard of it before, it’s also baffling to me that the researchers in this study considered it one of two main ways mnemonists remember long strings of numbers.



Interesting! Thanks a lot !

The idea is to have a 1000 things (in this case events) that you associated with the given number. I am failing to see what is different about this that makes it better. Because of are relation to events particularly? I have not noticed any difference on my end and I use a mixture of dates, shapes, people, major, and more tricks to associate to my numbers. But not enough dates to confirm there is a difference in efficiency for me.

I will definitely try more dates for my remaining, and who knows? It would be great to learn some history + number at the same time!

I have more than a 1000 numbers linked myself and the way I’ve been successful with is because I have first placed all of them in a memory palace, Or I should say many interlinked memory palaces. With time, as I went through the ones I know as well as I could now, I started to simply have to use whatever I want from the space where it is and place that in another palace.

Example : My 1978 is in my first memory as a child palace, in a plane, and it’s shaking, my seat is 1978, my birthday, and if I ever need to remember 978, I place a piece of that seat cloth in the palace I am building at the moment to remember the number sequence( in this case the touch sense is what works, but in most cases it is more visuallly exuberant)

P.S 2978 is the first birthday cake I remember

1 Like

I personally don’t think it’s better in any way. It’s just a different approach. And, if I’m reading the research correctly, it was more problematic for the one subject to learn and use effectively than the Major system was for the second subject.


Here’s a link:

Clicking “open access” leads to this PDF:

1 Like

90 digits at a 1 second rate is pretty impressive but 1000 images is a pretty overwhelming place to start.

It seems like there is something missing from this study in terms of the effort made to engrain the underlying method. You would expect with equal preparation time learning the triplet method would take more than ten times the effort of learning the two digit method. Similarly with equal preparation time you would expect the major system to provide quicker imagining for chaining/loci.

I can accept that a fully trained individual with a 3 digit system is going to outperform one with 2 digits and the 1 second encoding rate is impressive. It generally takes me somewhere around 3 seconds to do 2x2 multiplication when I have to think.

The drastically lower result of the major system over the 3 digit system surprises me however in a specific way. People who work simple systems aggressively, soroban, criss-cross method of multiplication seem to become very effective quickly at the method. I expect this is mid-term to long-term limiting but I would have thought that in a short study with 2 novice memory students that the major system with loci would outperform 1000 pegs with loci.

I am going to have to read this study a few more times and maybe send the author some questions.

Does Reinhold lurk on this forum?

Another link to his research


If I remember rightly, Tony Buzan’s Major system wasn’t just 3-digit. It was a Master Memory Matrix (or something like that) in which 3 digits became 4 or 5 by adding colours and smells. I think Kevin Horsley uses this system and has 0 - 50,000(?) mapped out.

Also, didn’t Harry Lorrayne just make up images/sentences/stories on the fly for telephone numbers? So when I hear about people using 3-digit Major systems I always wonder what happened to the 5-digit system. Why don’t the pros use it?

1 Like

Just to clarify: This study is not referring to a three-digit Major system. It’s comparing the results of memorizing using a three-digit history-dates system versus a two-digit Major system.


1 Like

In terms of following up with the authors, do keep in mind that this study was published over 30 years ago…

What surprised me most about it—apart from the researchers using a memory technique I’d never heard of before—was that this was just a two-person study. While they took pains to reference prior research and thoroughly document their approach and conclusions, I’m not sure this study really proves anything, since the conclusions are based on the experiences of just two individuals. Still, I found it fascinating to read and thought it might be interesting to folks here.

BTW, I found this study thanks to the bibliography in Moonwalking with Einstein. There is a wealth of information there.



oops. 30 years old is a little long in the tooth to email the author you are right.

1 Like