In 1946, Dutch psychologist Adriaan de Groot conducted a famous experiment. He himself was a master-level chess player, and he took a scientific approach to the question of what differentiates great players from average ones.
De Groot gathered players of various levels, from GMs to beginners, and conducted the following experiment:
This seemingly simple experiment yielded remarkable discoveries.
When testing with positions from actual games, significant differences emerged:
| Player Level | Pieces Correctly Placed |
|---|---|
| GM | ~93-95% |
| Master | ~70-80% |
| Intermediate | ~50% |
| Beginner | ~30% |
GMs could recreate over 20 pieces almost perfectly after just 5 seconds of observation.
Here is the core of the experiment. De Groot also conducted the same test with pieces placed randomly in "meaningless" positions. Meaningless positions refer to unrealistic arrangements that could never occur in an actual chess game.
The results were striking:
| Player Level | Accuracy with Random Positions |
|---|---|
| GM | ~30% |
| Master | ~30% |
| Intermediate | ~30% |
| Beginner | ~30% |
Everyone performed at nearly the same level.
These results overturned common misconceptions about GM abilities:
The idea that "skilled players like GMs have larger short-term memory capacity than average players—meaning they can memorize more items at once" proved incorrect. If that were true, they should excel at meaningless positions too.
What skilled players excel at is recognizing "meaningful patterns." Real game positions contain patterns learned from experience: opening formations, pawn structures, typical piece placements. Skilled players recognize these instantly and remember them as "meaningful chunks" rather than individual pieces.
This research provides important insights for blindfold chess training:
Memorizing randomly placed pieces is difficult even for GMs. Practice using positions that appear in actual games is more effective.
Learning chess patterns—openings, endgames, tactical motifs—is not merely acquiring knowledge. It enhances your ability to "see" the board itself.
Rather than memorizing 32 piece positions individually, recognizing them as meaningful formations—"typical King's Indian middlegame" or "Sicilian Dragon Variation"—significantly reduces memory load.
De Groot's experiment scientifically proved that chess improvement is not simply a matter of memory capacity. The ability to play blindfold chess is likewise not an innate special ability, but a skill that can be trained through pattern recognition.
By studying many positions and accumulating patterns, you too can develop the ability to play chess without seeing the board.
