# Memory Champions Techniques Anyone Can Learn: The Science of Trained Memory At the 2019 World Memory Championships in Shenzhen, China, the competitor Ryu Song-i memorized 620 decimal digits in five minutes. She correctly recalled the order of a shuffled deck of 52 playing cards in 13.86 seconds. In a separate event, she memorized the order of ten shuffled decks, 520 cards total, after fifteen minutes of study. These performances sound superhuman. They are not. Research on trained mnemonists, pioneered by Anders Ericsson and extended by Boris Nikolai Konrad and others, has demonstrated repeatedly that world-class memory performers are ordinary people who have trained systematically with techniques available to anyone. A 2017 study by Martin Dresselhaus and colleagues showed that six weeks of structured training with the method of loci produced durable memory improvements in novice learners that persisted at four-month follow-up. This article explains the core techniques, the cognitive science behind them, and how to apply them to real-world learning. It is a practical guide to tools that have been known for over two thousand years but are, even today, used by almost no one. --- ## The Method of Loci The method of loci is the oldest and most powerful of the mnemonic techniques. Its origins trace to the ancient Greek poet Simonides of Ceos, who according to legend could identify crushed banquet guests by recalling where each had been sitting before the building collapsed. Roman orators Cicero and Quintilian systematized the technique, and it remained standard equipment for educated minds through the Renaissance. The technique works by exploiting two features of human cognition: our extraordinary memory for spatial information and our strong encoding of vivid imagery. To use the method of loci: 1. Select a familiar spatial route. Your home is a common choice, but any well-known path works. A walking route to work, the layout of your childhood elementary school, or the streets of a familiar neighborhood all serve. 2. Identify a sequence of specific loci along the route. These might be the front door, the coat hook, the hall mirror, the living room couch, and so on. Aim for 20 to 50 distinct locations that you can mentally visit in a consistent order. 3. Associate items you want to remember with those locations. The associations must be vivid, unusual, and multisensory. A mundane association fails; a bizarre or emotional image succeeds. 4. To retrieve the information, mentally walk the route and retrieve the item placed at each location. The technique is more reliable than pure rehearsal because spatial memory is one of the most robust systems in human cognition. Episodic memory for where things are located draws on the hippocampus and associated medial temporal structures that are deeply evolved and highly efficient. > "The method of loci is not a trick. It is a way of using the memory architecture that evolution gave us for tracking spatial and episodic information. It recruits systems that are already extraordinarily powerful, and directs them toward the kinds of content that verbal memory handles poorly." -- Joshua Foer, *Moonwalking with Einstein* (2011) --- ## The Major System for Numbers Numbers are difficult to memorize because they lack distinguishing features. The digit 7 looks like other digits, and a string of digits contains no natural boundaries for chunking. The major system, developed in the seventeenth century, solves this problem by converting digits to consonant sounds, from which you can construct memorable words. The standard mapping assigns consonants to digits as follows: | Digit | Consonant Sound | Memory Aid | |---|---|---| | 0 | s, z, soft c | "z" is the first letter of zero | | 1 | t, d, th | "t" has one downstroke | | 2 | n | "n" has two downstrokes | | 3 | m | "m" has three downstrokes | | 4 | r | "r" is the last letter of "four" | | 5 | l | roman numeral L | | 6 | j, sh, ch, soft g | a script "j" resembles 6 | | 7 | k, hard c, hard g | a "k" contains two 7s | | 8 | f, v | script "f" resembles 8 | | 9 | p, b | "p" is a mirrored 9 | Vowels and the letters w, h, y are free. So 31 becomes "mat" or "mud" or "maid." 426 becomes "rinch" or "ranch." 1492 becomes "turban" or "tray pin." With practice, numbers can be converted to words at speeds of roughly one digit per second, and these words can then be placed in a memory palace for later retrieval. Memory competitors often pre-memorize images for every two-digit number from 00 to 99, creating a personal vocabulary of 100 images that can encode any digit sequence. --- ## The PAO System For memorizing long digit sequences or shuffled card decks, the Person-Action-Object (PAO) system multiplies the power of memory palaces. Each of 100 two-digit numbers (or 52 playing cards, or any other discrete set) is pre-encoded with three elements: a Person, an Action, and an Object. For example, 34 might be Princess Diana (person), waving (action), a hairbrush (object). 57 might be Michael Jackson, moonwalking, a sequined glove. When encoding numbers, you combine triplets. The six-digit sequence 345734 would be encoded by taking the person from the first pair (34, Princess Diana), the action from the second (57, moonwalking), and the object from the third (which happens to also be 34, so hairbrush) into a single composite image: Princess Diana moonwalking with a hairbrush. This single image encodes six digits and can be placed in one locus of a memory palace. A well-trained memory competitor can encode an entire shuffled deck of 52 cards in about 15 seconds using a PAO system. The technique requires substantial upfront investment in building the PAO set but produces enormous compression once internalized. --- ## Spaced Retrieval and Memory Consolidation Mnemonic techniques handle the encoding problem. They do not fully handle the consolidation problem. Even beautifully encoded information fades from memory if not retrieved over time. Research on spaced retrieval, extending from Hermann Ebbinghaus's pioneering 1885 work on forgetting curves to contemporary studies of the testing effect, has established that retrieval practice distributed over time produces substantially more durable memory than massed practice. The principle is simple: study less frequently, but retrieve more. Effective spacing depends on the retention interval. For short-term retention (a test tomorrow), retrieval practice a few hours apart is adequate. For long-term retention (expertise development), retrieval should be spaced across days, weeks, and months with increasing intervals as memory strengthens. Software implementing spaced repetition, such as Anki and SuperMemo, has made this technique widely accessible. Combined with mnemonic encoding, spaced retrieval produces the durable learning that expertise requires. This principle underlies the study approaches that work for certification preparation programs like those at [Pass4Sure](https://pass4-sure.us), where candidates build durable knowledge of complex domain material through deliberate retrieval practice rather than passive re-reading. Structured note-taking systems documented at [When Notes Fly](https://whennotesfly.com) complement mnemonic techniques by providing the external scaffolding that supports retrieval practice across sessions. --- ## Applications to Real-World Learning The memory competition context may seem remote from practical learning needs. In fact, the techniques apply directly to a wide range of real-world challenges. ### Languages Vocabulary acquisition is a classic mnemonic challenge. The keyword method, in which a foreign word is encoded through a native-language keyword that sounds similar and is linked by a vivid image to the meaning, has strong research support. To memorize the Spanish word "caballo" (horse), a learner might notice that it sounds like "cab-eye-oh" and imagine a horse with a taxi cab on its eye. Research by Richard Atkinson in the 1970s demonstrated that keyword method produces 50-100% improvement in vocabulary retention over rote memorization. ### Anatomy, Pharmacology, and Medicine Medical students have adopted mnemonic systems for decades. The cranial nerves, drug classifications, disease symptom clusters, and anatomical structures are all well-suited to memory palace techniques. The mental discipline of encoding complex material into vivid imagery also supports the kind of pattern recognition that clinical practice requires. ### History Historical dates and sequences respond well to the major system. Memorizing that the Norman Conquest occurred in 1066 becomes easier when 1066 is encoded as "touche-shoo-judge" or any other memorable phrase generated by the consonant mapping. Historical sequences can be placed along a timeline-based memory palace, allowing rapid retrieval in chronological order. ### Presentations and Speeches Orators from Cicero onward have used memory palaces to organize speeches. Each point of the speech is placed in a locus, and the speaker mentally walks the route during delivery. The technique produces the appearance of extemporaneous speaking with none of the memorization-induced wooden quality of scripted presentations. ### Names and Faces The most universally requested memory improvement is for names. Effective techniques combine visual attention (really look at the face, identify distinctive features), verbal encoding (repeat the name, say it back in conversation), and imagery (connect the name to a visual feature of the person or to their context). Regular practice reduces name-forgetting substantially. The following table summarizes the main techniques and their primary applications: | Technique | Primary Application | Learning Curve | Payoff Timeframe | |---|---|---|---| | Method of Loci | Ordered lists, presentations, history | Moderate | Days to weeks | | Major System | Numbers, dates, digit sequences | Steep | Weeks to months | | PAO System | Very long number sequences, card decks | Steep | Months | | Keyword Method | Foreign vocabulary | Easy | Immediate | | Peg System | Numbered lists up to 10-20 items | Easy | Immediate | | Linking Method | Short sequential lists | Easy | Immediate | | Spaced Retrieval | Long-term retention of any content | Easy to implement | Weeks to permanent | --- ## The Neuroscience of Trained Memory Neuroimaging studies have revealed what happens in the brains of trained memory experts. A 2003 study by Eleanor Maguire and colleagues scanned memory competitors during encoding tasks and found that they engaged brain regions associated with spatial navigation, particularly the right posterior hippocampus, medial parietal cortex, and retrosplenial cortex. Control subjects performing the same tasks showed activation mostly in regions associated with verbal rehearsal. The finding confirmed that mnemonic techniques work by recruiting spatial memory systems for tasks that would otherwise rely on weaker verbal memory. It also suggested that the techniques are available to anyone with intact spatial memory, which is to say, almost everyone. A 2017 study by Dresselhaus and colleagues took the work further. They randomized novice learners to six weeks of method of loci training or control conditions. Trained learners showed substantial memory improvements that persisted at four-month follow-up, and neuroimaging revealed increased functional connectivity between the hippocampus and medial prefrontal cortex similar to patterns seen in memory competitors. The study settled a longstanding question: mnemonic training produces durable functional brain changes in ordinary learners. The techniques do not merely provide clever tricks; they engage neural systems that develop with practice. > "What we found is that ordinary people, given six weeks of training, can develop brain activity patterns that resemble those of world memory champions. This is not about innate ability. It is about access to techniques that most people have simply never been taught." -- Boris Nikolai Konrad, *Neuron* (2017) --- ## Common Pitfalls and How to Avoid Them Several patterns undermine success with mnemonic techniques, especially early in training. ### Weak Imagery Vivid, multisensory imagery is essential. Mental images that are bland, static, or poorly visualized produce weak memory traces. Successful imagery is unusual, emotional, interactive, and engages multiple senses. ### Inadequate Palaces A memory palace with indistinct or poorly remembered loci cannot support reliable retrieval. Build palaces that you know in detail. Walk the route mentally until every locus is clear. Many competitors maintain multiple palaces, each dedicated to different content domains. ### Skipping the Upfront Investment Systems like PAO require substantial initial investment. Resisting this investment and trying to "wing it" produces weak performance. Building a 100-image set for two-digit numbers might take 20-40 hours of focused work; once built, it pays dividends for years. ### Neglecting Spacing Encoding alone does not produce durable memory. Even beautifully constructed memory palaces fade without retrieval practice. Plan the retrieval schedule from the beginning, using spaced repetition software or simple calendar reminders. ### Applying Techniques to Wrong Content Mnemonic techniques excel at discrete, factual information. They are less useful for conceptual understanding, which requires elaborative processing, integration with existing knowledge, and flexible retrieval in varied contexts. Using memory palaces to learn calculus would waste effort; using them to memorize chemical formulas or historical dates is appropriate. --- ## Tools and Environmental Support The ancient techniques combine well with modern tools. Several pairings are particularly useful. **Spaced repetition software** implements the retrieval schedules that turn short-term encoding into long-term memory. Anki is the most popular free option. Commercial alternatives provide more polished interfaces but similar core functionality. **Digital text analysis** helps when processing dense material before encoding. Tools like the word-count and text-analysis utilities at [File Converter Free](https://file-converter-free.com) can help identify key terms and structure for mnemonic encoding. **Focused workspaces** support the concentrated attention that mnemonic encoding requires. Working environments that balance focus and stimulation, similar to the cafe settings described at [Down Under Cafe](https://downundercafe.com), are well-suited to the deep work of building memory palaces. **QR codes and link systems** can serve as external retrieval cues, extending internal memory with external prompts. Tools at [qr-bar-code.com](https://qr-bar-code.com) can embed prompts in physical spaces or printed materials for supplementary retrieval practice. **Writing and template systems** support the articulation of memorized content. Clear writing structures, such as those curated at [Evolang](https://evolang.info), allow memorized material to be communicated effectively. The act of writing about memorized content also serves as a retrieval practice session. --- ## Memory Techniques Across Species Comparative cognition research has documented extraordinary memory performance in non-human species through mechanisms that parallel human mnemonic techniques. Research at sites like [Strange Animals](https://strangeanimals.info) has catalogued these findings. Clark's nutcrackers, a species of corvid, cache up to 30,000 seeds across thousands of locations and retrieve them months later. Their performance depends on spatial memory that dwarfs typical human ability, achieved through specialized hippocampal architecture that expands seasonally during caching. Chimpanzees have demonstrated extraordinary short-term memory for visual sequences. A young chimpanzee named Ayumu, trained by Tetsuro Matsuzawa at Kyoto University, outperforms human adults on spatial memory tasks involving briefly flashed numerals. The performance depends on encoding spatial arrangements as unified perceptual patterns rather than as sequential verbal information. These cross-species findings suggest that the cognitive architecture supporting extraordinary memory is evolutionarily ancient and broadly available. The human contribution is not better hardware but better software: symbolic systems like the major system and conceptual tools like memory palaces that direct existing memory machinery toward content it would otherwise handle poorly. --- ## Professional and Business Applications For professionals, the techniques have direct applications beyond memory competition. Sales and customer-facing roles benefit from name and face memorization techniques. Clients notice when their names are remembered, and the cognitive overhead of tracking relationships is reduced when names attach reliably to faces. Legal and medical professionals handle large volumes of discrete factual material that mnemonic techniques process efficiently. Drug interactions, case citations, diagnostic criteria, and statutory language all respond to structured mnemonic encoding. Entrepreneurs and business founders manage diverse information across legal, financial, and operational domains. Reference resources like the business formation guides at [Corpy](https://corpy.xyz) present substantial discrete content (entity types, filing requirements, tax treatments by jurisdiction) that mnemonic techniques can encode durably. The initial investment in encoding saves repeated lookup time over months and years. Academic and certification test preparation is another clear application. Students preparing for standardized exams often face thousands of discrete items that must be retrieved accurately under time pressure. Mnemonic encoding, combined with spaced retrieval, outperforms pure repetition by substantial margins in controlled studies. --- ## A Realistic Expectation These techniques are powerful. They are not magic. Building durable mnemonic skill requires sustained practice over weeks and months. The first few memory palaces will be built slowly. The first attempts at the major system will be clumsy. The first PAO set will take substantial time to construct. Expecting immediate fluency leads to abandonment; expecting gradual fluency leads to mastery. The techniques also do not substitute for conceptual understanding. A memory palace full of cellular biology terms does not make someone a biologist. Deep understanding requires elaborative processing, integration with existing knowledge, and application to novel problems that simple memorization cannot provide. What the techniques do provide is reliable access to discrete information, compressed into retrievable form, without the effortful rehearsal that traditional memorization requires. That capability, once developed, supports serious learning across virtually any domain. The world memory champions demonstrate its upper bounds. The research shows that its lower bounds, available to any trained learner, are far higher than most people suspect. --- ## References 1. Maguire, E. A., Valentine, E. R., Wilding, J. M., & Kapur, N. (2003). Routes to remembering: The brains behind superior memory. *Nature Neuroscience*, 6(1), 90-95. https://doi.org/10.1038/nn988 2. Dresler, M., Shirer, W. R., Konrad, B. N., et al. (2017). Mnemonic training reshapes brain networks to support superior memory. *Neuron*, 93(5), 1227-1235. https://doi.org/10.1016/j.neuron.2017.02.003 3. Roediger, H. L., & Butler, A. C. (2011). The critical role of retrieval practice in long-term retention. *Trends in Cognitive Sciences*, 15(1), 20-27. https://doi.org/10.1016/j.tics.2010.09.003 4. Atkinson, R. C. (1975). Mnemotechnics in second-language learning. *American Psychologist*, 30(8), 821-828. https://doi.org/10.1037/h0077029 5. Ericsson, K. A., Chase, W. G., & Faloon, S. (1980). Acquisition of a memory skill. *Science*, 208(4448), 1181-1182. https://doi.org/10.1126/science.7375930 6. Bower, G. H. (1970). Analysis of a mnemonic device. *American Scientist*, 58(5), 496-510. 7. Karpicke, J. D., & Blunt, J. R. (2011). Retrieval practice produces more learning than elaborative studying with concept mapping. *Science*, 331(6018), 772-775. https://doi.org/10.1126/science.1199327 8. Inoue, S., & Matsuzawa, T. (2007). Working memory of numerals in chimpanzees. *Current Biology*, 17(23), R1004-R1005. https://doi.org/10.1016/j.cub.2007.10.027