Brain Science
Memory Match trains short-term visual memory, but the technique that actually works isn't visual. It's spatial.
Memory Match in MyRin follows the classic format: all cards are revealed for a few seconds, then flipped face-down. Tap two cards at a time. Matching pairs stay face-up; non-matching pairs flip back after a short delay. Clear the board by finding all pairs with the fewest flips possible.
The game appears to test visual memory, whether you can remember what image was on each card. But the skill that actually differentiates high performers from low performers isn't image recognition. It's spatial memory: the ability to remember where something was, not just what it looked like.
During the reveal phase, the cards are visible for only a few seconds. In that window, your brain processes two kinds of information about each card: the image identity (what the card shows) and the location (where on the grid it sits). Both can be remembered, but they're stored differently.
Image identity is stored in the visual cortex's object recognition pathways (the ventral "what" stream). Location is stored in the parietal cortex's spatial processing pathways (the dorsal "where" stream). These are anatomically distinct systems. The critical finding from memory research: the spatial "where" system is more robust under rapid encoding conditions, it captures and retains location information more reliably after a brief exposure than the "what" system captures the identity details.
This is why the strategy of thinking "the star was at top-left, row 2 column 3" outperforms the strategy of thinking "there's a star somewhere." When you flip an unrecognized card mid-game, the useful question is "where have I seen this before?", a spatial query, not "what does this card show?", an identity query. The spatial answer is available; the identity answer may not be.
During the initial reveal, don't try to scan the entire grid. The grid is too large for complete encoding in a few seconds, and attempting to encode everything results in encoding nothing well, scattered attention produces scattered memory.
Instead, assign yourself one zone, typically one corner or one row, and memorize it completely before the reveal ends. "Top-left quadrant: star (row 1, col 1), moon (row 1, col 2), star (row 2, col 1), sun (row 2, col 2)." If the reveal allows more time, extend your zone to a second quadrant. But complete knowledge of one zone beats fragmentary knowledge of the whole grid.
The rationale: your first moves will clear the zone you fully memorized, which opens space and frees cognitive resources for the zones you didn't fully encode. Partial memory of the rest of the board is supplemented by information gathered during play, every non-matching flip reveals a card's identity and position, which gets added to your running knowledge of the board.
When you flip a card mid-game and it doesn't match what you're looking for, the natural impulse is to note its identity for later. "Oh, that's the anchor. I'll remember it's an anchor." But in a large grid with many card types, identity alone isn't enough, you need to know where the anchor is.
The more effective encoding: "anchor at row 3, column 4." Position is the primary key; identity is the secondary attribute attached to that position. When you later flip a card and see an anchor, the question to ask yourself is "do I know where the other anchor is?" If yes, go directly there. If no, this flip has given you half a pair's location, encode it as "anchor at [current position]" for when you encounter the second one.
Spatial anchoring is trainable. Players who practice explicitly naming the grid position of each flipped card, even silently, improve their pair-finding efficiency measurably within a few sessions. The habit of attaching position to identity, rather than identity alone, is the core skill the game is training.
After flipping the first card of a pair attempt, most players immediately commit to the second card. This is the most common source of errors, rushing to the second flip before fully searching memory for the matching card's location.
The better approach: after seeing the first card's identity, pause for one to two seconds and scan your memory for its partner's location before tapping anything. If you know the location, go there. If you don't know the location, flip a card you haven't seen before as the second tap, even knowing it won't match, you gain new information (a card's identity and position) that may help you later.
This "reconnaissance flip" strategy, deliberately using a non-matching second flip to gather information, is significantly more efficient than guessing at a card you vaguely feel might match. The information-gathering value of a deliberate reconnaissance flip is positive; the information-gathering value of a wrong guess made with low confidence is near zero.
Memory Match targets short-term visual memory with a spatial encoding component. The scientific term for the specific capacity it trains is spatial working memory, the ability to hold and use location information in active memory.
Spatial working memory is measured in cognitive assessments through Corsi block tapping tests (tapping blocks in sequence to reproduce a pattern) and virtual Morris water maze tasks (finding a platform's location in a pool using surrounding cues). Performance on these tasks correlates with everyday navigation ability, the ability to follow directions involving multiple spatial steps, and certain aspects of reading comprehension that require tracking where you are in a text structure.
Memory Match's escalating difficulty, more cards, shorter reveal time, more similar images, systematically increases the spatial working memory demand at each level. The game is more than a casual card game; it's a structured training protocol for one of the most practically useful components of working memory.