A Quest for Optimal Training: a detailed memory training journal

Context

Hi! My name is Maike Anthony, and I’ve been practicing memory techniques for almost 5 years now. However, I’ve never practiced speed memorization. After reading Joshua Foer’s Moonwalking with Einstein last year (I should’ve read it earlier), I got a strong desire to get faster.

My motivation to write this journal is threefold: to get advice from experienced athletes, share a detailed account of my practice to help others, and motivate myself to keep going. In fact, most of the time I shared an idea on the forum, I received great insights, impossible to get otherwise. Moreover, an additional goal of this thread is to apply the most efficient methods for training that I know of to get a better sense of how they work. In other words, this is a journal about discovering how to efficiently get better at something, whatever it is.


Disciplines

I’m going to hone the following skills:

  • 5-min Numbers
  • 5-min Binary
  • Speed Cards
  • 5-min Words
  • 5-min Names (WMSC style: international names, surnames, and faces)
  • Recognition speed (Anki) / Memory Palaces navigation speed

I won’t practice marathons yet because I want to build speed first. Also, I may exchange my speed practice for long-term goals. For instance, if I decide to memorize a book, I will use my memory training time.


Anki Deck and Systems

If you want to see my Anki deck and systems, you may access them here.

For cards, I’ll be using the Double-2-Block System together with Variable Spatial Encoding to differentiate the colors of the suits (2-card system).

Special thanks to @TheHumanTim , @Honje , @Bigdonnyv , and @Zoomy for their advice, criticism, ideas, and resources. They helped me a lot with formulating my methods and the Anki deck.


Memory Palaces

I have shared my list of Memory Palaces here. I hope to build others in the future, and my final goal is to have 15 Memory Palaces in total (3.510 loci) for memory training.

All of them are universal. That is, each of them will store information from any discipline. When I finish a trial, I’m simply going to start the next one at the next empty locus.


Websites

In order to practice, I’m going to use the following websites:

IAM Training Website is good too, so I may use it sometimes, but not very often because it doesn’t have auto advance.

Special thanks to @Aa123456 for sharing the Arabian Memory website here. I would never have found it otherwise.


Equipment

I’m going to use the following objects:

  • Blinders (swimming goggles painted black with holes in the middle to obscure peripheral vision)
  • Industrial earmuffs for sound isolation
  • Timers for Rubik’s cube (I don’t have this yet, but I intend to buy it for when I practice with physical decks)

The first two are the most important ones since they help a lot with concentration.


Memory Training Log

After each trial, I’m going to fill in the following Memory Log spreadsheet:

This is where I’m going to record all my progress and will share its updates with you guys. I hope you find this table useful for your own practice too.

The red columns are supposed to be filled right after the trial, while the blue ones are only filled at the end of the whole session.

The distraction column refers exclusively to external ones, like sounds or people in the background. Internal distractions should be placed in either the “Mind-wandering” or “Drowsiness” columns. (PS: this paragraph refers to columns that used to exist in the spreadsheet: the focus columns. I deleted them as I’ll explain in my second post.)

The deliberate practice columns are by far the most important ones, sometimes even more important than the practice itself. When I write down any insight or lesson, I’ll highlight the cell in red to make it easier to find.

In the well-being columns, I write “X” if I experienced any of those problems or write nothing at all if I didn’t. (PS: this paragraph refers to columns that used to exist in the spreadsheet: the well-being columns. I deleted them as I’ll explain in my second post.)

The short-term goals are the ones I’ll have in mind during every trial. When using auto advance, the short-term goals will be used to set the speed. The long-term goals, though, are the ones I hope to achieve in a longer time frame, like weeks or months.

I’m going to use this rule to determine whether I should increase the short-term goal:

After three trials all with at least 80% accuracy and one with at least 90%, I’m going to increase the short-term goal by 20% (i.e. multiply the amount by 1.2 or divide the time by 1.2).

(PS: I actually changed this rule a few weeks after the start of my training. I’ll tell you the new rule in my next update.)

For the long-term goals, they are currently the following:

  • 5-min Numbers: 265 digits
  • 5-min Binary: 500 digits
  • Speed Cards: 40 seconds for one deck
  • 5-min Words: 50 words
  • 5-min Names: 40 names

I hope it’ll take at most two years to reach all of them.


Schedule

I’m going to practice three sessions before lunch and one session after it. Each of them takes 50 minutes, followed by a 15-min break.

Inside each session, there’ll be multiple trials without breaks between them. To apply varied practice to my training, I am going to choose the disciplines of each trial at random using Google’s Random Number Generator. Therefore, I may end up memorizing a deck of cards, then words, then numbers, then names, and so on, but I won’t do the same discipline again before doing each one at least once that day. Applying interleaving and varied practice means that my improvement will be very slow at first, but I hope the long-term benefits will make this strategy worth it.

The only defined practice will be the recognition speed/memory palaces navigation speed, which I will always practice in the session after lunch, and only during it.

I’m going to practice 5 days a week (usually one weekday and one weekend for breaks), so it’ll be 16 hours and 40 minutes of training per week.

For planning, I’m going to use habit tracking (to motivate myself) and time-blocking (to organize my schedule). These two methods are crazily good for habit formation and discipline. In fact, they are what practically cured my maladaptive daydreaming.

If some unforeseen event, like sickness or loss, happens, forcing me to stop practicing, get worse performance, and feel like giving up, I’m only going to allow myself to give up the project after I get as good as before the event.

All this schedule is valid only until the middle of August. Since I’m going to start undergraduate education in the USA as an international student, I don’t know how my training will change when the fall semester starts.


Breaking Through Plateaus

I may apply one or more of the following methods when I face plateaus:

  • Increase difficulty: I may increase the amount or the speed to force myself to get better. When one tries harder, they fail at first, but they should inspect the mistakes because they are the weak points, which should be worked on.

  • Practice differently: by changing the training, one can spot things that are hard to see, which will allow for improvement. For instance, I may train on paper instead of digitally, practice without auto-advance, memorize in public, decrease the amount but increase the speed, etc.

  • Stop practicing: when I’ve tried everything but still can’t spot anything to work on, it’s probably better to take a break from that discipline. After coming back, it’s common to experience a sudden improvement, which may be caused by the incubation effect. I’m going to take either 4 or 5-day breaks.


Health and Well-Being

I’m doing all these things:

  • Sleep: 10 PM to 6 AM (8 hours) every night;
  • Sleep hygiene: no electronic devices after 9 PM (except for my Kindle);
  • Siesta: 30-min nap after lunch;
  • Water: 4 liters per day;
  • Alcohol/cigarettes/drugs: none;
  • Coffee/energy drinks: I may drink them very rarely, about once or twice a month;
  • Omega-3 supplements: 1000 mg of DHA and 300 mg of EPA per day;
  • Gym: 6 days a week (1 hour of weightlifting and 20 min of aerobic exercise);
  • Diet: following the gym’s nutritionist (I’m not strict. When I hang out, I don’t mind eating some sweets, for instance.);
  • Sun exposure: 10-min sunbath, 4 days a week;
  • The Wim Hof Method: breathing technique + cold showers, in the evening and 6 days a week (I’ll be doing this mostly for health reasons, but I still decided to mention it because it may have psychological advantages as well).

Posting

I don’t promise any frequency, but it will probably be every week or two. It depends on whether I’m going to have any insight to share. In any case, I’ll be updating the Memory Log Spreadsheet in this post with my performance and lessons, so you can check that instead. Finally, if I make any big decision, like changing the disciplines or giving up, I’m going to let you guys know.

My training started on May 2, 2024. My next post will include what I learned during this time.

6 Likes

I love everything… except this. I highly recommend against this.

I use similar in competition, however training by myself Id encourage you to open yourself up to distractions. You cannot always control your environment and you need to be ready for this.

On the note with Kindle are E-Books fine for you? I recently bought a sleep mask with bluetooth for $25 and it’s one of my favourite purchases ever.

2 Likes

Thanks for your feedback, Don! I do intend to practice with distractions as well, but more as a way to make it more diverse and engaging. I’ve also noticed that wearing blinders when I’m already tired makes me even sleepier.

Exactly. My goal with sleep hygiene is just to get rid of blue light and distractions coming from my phone and computer. Reading E-books or books helps me get calmer before going to sleep.

I’ve never heard of that before, but it seems interesting. The only things I do to help me fall asleep are recollecting what happened throughout the day and counting my pegs. I do wear a sleep mask too, though. If it’s during my nap, I gotta wear industrial earmuffs as well because there’s more noise in the afternoon, but they aren’t very comfortable, so your idea sounds quite nice.

Also, I thought you were gonna scold me for using universal memory palaces, hahaha.

Hi mike !
Good luck with your training !
and the only tip that I can give it to you is :

Trust in your memory !
This goals you can reach it in 2 months
Not 2 years !

I remember when I started memory training I was train for 2 months on words only
and can you guess ?
I reached the World record for WMSC 130 words !

I think the main reason is : that I was Focus only on words !
and I didn’t say that will be hard to reach the world record in words !
I was even didn’t know about the WR
But when I saw that I reached 100 words I research about the world record then training on break it !

So you can also focus on make a system for numbers
100 images and just focus on it !

Then you can reach 500 binary easily in few days if you are faster with numbers system !

also cards you can use your numbers system for cards !
And it will be easier !

Words i think you. Can reach it in 1 month of training !
50 words in 1 month is so easy !

and last thing that I want to say is : good luck again with your training !

1 Like

Maybe 260 digit in 5 min numbers will take more time !
But
It is better to training on memory league to be faster with your system !

1 Like

Thanks for the feedback, Naser!

That is really awesome!!! Congrats!!!

I already got a 3-digit number system and a 2-card system memorized. Are you recommending me to ditch it for a 2-digit system and a 1-card system?

1 Like

This is a follow-up of my training.

From May 2 to 11, I only reviewed my images on Anki and committed two of my memory palaces to memory. During the second week of the month, I also didn’t practice for four days because I was in a trip to get my student visa. In total, I got 12 days of real training (not counting the breaks).


Performance

• 5-min Numbers: from 98 digits to 148 digits (51% increase in 12 trials)

• 5-min Binary: from 102 digits to 235 digits (130% increase in 13 trials)

• Speed Cards: from level 1 to level 5 in Memory League (23 trials)

• 5-min Words: from 7 words to 39 words (457% increase in 14 trials)

• 5-min Names: from 3 people to 6 people (100% increase in 14 trials)


By far, the hardest ones for me are cards and names. The former is difficult because I should review more on Anki, while the latter is mostly because of spelling.

It’s been tough to memorize those names. I usually get through 10 people, but there are so many misspellings that I end up with only 6 ones totally correct at the end. I have no problem with the faces, but the names and surnames are difficult. My current method is to imagine the person doing an action with the name’s peg on the surname’s peg, and placing only one person per locus. Sometimes, the spellings are so strange that I have to use two images for a single name/surname, which makes the scene crowded and forgettable. I wonder if there’s a better strategy, so I would love to get any feedback on this one.


Changes

I’m exchanging my previous short-term goal rule for the next one: if my accuracy is at least 90% in a trial, then I should increase the difficulty of my short-term goal by 20%. The previous rule made my progress MUCH slower.

I also deleted the columns about focus (mind-wandering, drowsiness and distractions) and well-being (sickness, sleep deprivation, stress, lack of exercise, hunger, thirst, and extreme temperature) because they were too constant, which made them useless.

About my diet, I’ll add that I have been taking whey protein (25g per day) and creatine (5g per day) since May 17. I want to inform you about this because the latter may have a positive correlation with cognitive performance.


Lessons

• In the numbers discipline, it’s much better to make a strong connection between the person and the object rather than between the object and the locus. This “intra-connectivity” seems much more important for accuracy because in the latter situation it’s easier to forget the person.

• I use PO for numbers, but I tested POP too. Well, it sucks so badly for me. PO is quite good, though.

• I think @TheHumanTim may like to know that I’ve tested VSE for longer now. Well, my conclusion is that Variable Spatial Encoding sucks. It’s just too easy to make mistakes with it. On May 20th, I decided to move to Variable Image Stacking instead.

• About my system for binary numbers, I’m LOVING it. I thought it would suck, but I like it very much.

• Also, I initially thought that affixes memorization would be useless: I was wrong. Having images for common suffixes is very useful, but the prefixes images are indeed useless.

• About the fact that I’m choosing my disciplines randomly to apply interleaving and varied practice, I’m LOVING it too! It’s so fun not to know what will happen next. My progress is much slower, for sure, but the long-term performance will be better (and transfer too!).

• My memory palaces are universal (they are used for all disciplines), and that is also another thing that I’m loving so much. This is another way to promote interleaving and deeper familiarity with the loci.

• I hate names and faces because I suck so much at it, but this only means that I SHOULD practice it.

• About cards, it took me a LOT of time to notice that my struggle was mostly caused by lack of practice on Anki. I still got to do more of it, but I think I’m on the right track.


I updated my Excel spreadsheet with my performance and lessons, so you can read it if you want more information. The insights are highlighted in red.

Not too surprised about this. I haven’t worked with it much, but I can definitely see that at speed it would be really tough.

1 Like

أعتقد أن 3 أرقام للمبتدئين أكثر من اللازم!
إذا كنت لا تريد الحصول على بطل العالم وتريد فقط التدريب على الذاكرة
النظام المكون من رقمين هو نظام مثالي!

لكن المشكلة مع رقمين هي أن المقايضة هي الأكثر أهمية
يمكنك تبديل الصور برقمين!
أيضا مع 3 أرقام ولكن في 3 أرقام هو أقل من نظام رقمين!

لكن يمكنك استخدام النظام المثالي لك كما أرى

نظام PO!
البطل الهندي فيشفا راجاكومار يستخدمه!
أيضا DMV
ويصلون إلى 15 ثانية معها!

ولا يمكنك التبديل معها!
لأن الرقمين الأولين دائما سيكونان شخصا والثاني يعترض!
أيضا مع البطاقات
البطاقة الأولى هي الشخص والثانية هي الشيء!
لذا لا يوجد مبادلة هناك!

لذا يمكنك تجربتها وأخبرني إذا كنت تريد تجربتها!
لأنني أعتقد أن نظام البطاقات المكون من 3 أرقام و2 بطاقتين سيستغرق عامين كما قلت بالفعل!
ولكن مع السلطة الفلسطينية يمكنك الوصول إلى جميع أهدافك في غضون بضعة أشهر!
لكن لا تتدرب على ثنائي!
فقط ركز على نظام الأرقام والتدريب لتكون أسرع معه في دوري الذاكرة!
ثم إذا أصبحت أسرع، فيمكنك التدريب على البطاقات والثنائية أيضا في غضون أسابيع قليلة ويمكنك الوصول إلى أهدافك!

So my tip here is continue with your PO system
وشكرا لك
ناصر علي

1 Like

Hey,
Thanks for all the sharing. I’m impress by the amount of effort you are putting in this. It give me motivation to organize my training set up some drill and start a journal too.

I’m fairly new in this memory art world. I tried to read the other post you’ve shared. For number when you say you use a PO system you mean person object right?

Also you talked about Recognition speed (Anki) / Memory Palaces navigation speed. What do you mean by that? What kind of drill are you doing?

1 Like

Thanks for the feedback! I’m gonna keep with my 3-digit PO and 2-card system then.

Exactly. For instance, if I need to memorize “325 679 516 749”, I visualize:
325 679 : Manuel (325) shock absorber (679) → Manuel trying to swallow a shock absorber
516 749: Wild Jango (516) corbel (749) : Wild Jango playing with a corbel like a cat

Recognition speed means trying to get faster with your images using Anki. In my post about my systems, I shared my Anki deck. I use it to build filtered decks that are random selections of 50 flashcards from my system. I take these randomly selected cards and flag them. Later, I activate Anki’s auto advance feature and go through all the cards from the filtered deck. When I finish, I decrease the speed and rebuild the deck (now using the flag to make Anki select the same ones). I do this about three times in total, and then I change to another random selection of 50 flashcards. I’m currently practicing only cards, people, and objects on Anki.

It just means reviewing my memory palaces. I may use this time to build or memorize new ones as well.

I mean
PO
2 digit system !
Not 3 digit PO !
This will take a years to be good with it !
What I meant is
100 person
100 object
This is good for you !

1 Like

But I already got it memorized. I take less than one second to recall most of the people/objects already. I agree that it’ll take longer, but it’s gonna be a life-long investment. I don’t have any time constraints either, so I don’t mind if it takes me two years to reach the goals.

1 Like

It’s been 17 days of training already, and my progress has slowed down a LOT. Tbh, I wasn’t expecting it to get slower that early in my training. Some observations:

  1. Reviewing during encoding time is bad. Period. Firstly, that doesn’t help much since it isn’t active recall, just rereading. Secondly, it becomes a crutch that fosters worse visualization because you feel like, “okay, I just need to check it again later.” It’s important to get the image right the first time.
  2. I’ve thought of giving up names and faces many times. I can’t get those spellings right no matter what.
  3. Taking a break after just 2 days of training isn’t a good idea. Last week, I trained for 2 days, took a break, trained for other 2 days, and took another break. My performance and motivation fell drastically after returning. My current guess is that the ideal would be to take a day off after 3 or 4 days of training.

The End

I don’t want to practice this anymore. I started to ask myself what the point of this whole project actually is. I won’t compete, so what is the point?

  • These skills aren’t that important for my long-term goals;
  • I don’t believe they transfer;
  • Even if they did transfer, being very fast at them doesn’t seem that important, anyway;
  • If my speed drops that much just because I stopped for a few days, then it isn’t something worth practicing now since I won’t manage to maintain my training in the next months;
  • If it requires that much effort, then speed memorization isn’t really that useful for educational purposes, which is my true passion and mission;
  • I don’t find memory training pleasant. What I really like is the feeling of knowing that I learned something valuable, like when I memorized all the countries and the periodic table.

Therefore, my conclusion is that I should focus on long-term knowledge instead. For instance, spending my last months in Brazil learning a new language would be enormously more useful than practicing speed memorization.

I believe this very short project taught me a few very valuable things:

  • It seems that the best amount of deliberate practice is at most 1-5-25 (1-hour sessions, 5 hours per day, and 25 hours per week);
  • I started taking more care of my own body and health because of this project since I wanted to make it optimal, like doing siesta and sleep hygiene. I should keep doing these things;
  • It seems that 90% is indeed a good threshold accuracy before increasing the difficulty by 20%. I wonder if that is true for all skills as well;
  • Having a 3-digit PO system is awesome.

Thank you for sharing your reflections!

I would have guessed that for deliberate practice you would be in the high end of would be productive in terms of hours spent per gain. So it is interesting to me how you experience it.

I second your opinion that if long term knowledge is a goal, I do not believe speed memorization techniques is the most efficient way of working. It might be a useful tool sometimes, but it is probably more effective for long term learning to figure out how to actively engage with the content. That is, if the goal is learning, not memorization :smiley:

1 Like

I do feel that the practice was indeed optimal, except for maybe the number of disciplines that was too high. Varied practice slows down progress at the beginning, much more than I initially thought. For instance, on average, I would do each discipline once every day. Consequentially, if I had some insight, I would need to wait at least 24 hours to implement it. For this reason, maybe three disciplines would be better. Moreover, varied practice is optimal when the different subjects/skills have an underlying mechanism, so the best would have been to practice numbers, binaries, and cards since words and names are very distinct.

In conclusion, for any future goals of mine, I’ll follow the deliberate practice guidelines I shared on this thread.

Yes. I actually already knew that before I started my practice. However, some of my long-term goals (not knowledge) would profit from fast memorization. The following was crucial, though:

Also, this one:

Because of these two reasons, I think it’s better to practice my long-term goals directly, instead of getting fast at memorization first and only later working on the true objectives.

Drawabox is a free website teaching people to draw. They have a lot of focus on deliberate practice. They have a fifty fifty rule. Use fifty percent of your “learning time” on the deliberate practice and fifty percent on doing things. I think that is a sound way of thinking. Learning is maximized when you do not only practice the fundamentals, but also apply them to real scenarios.

1 Like

Hey, do you still practice?