The important decision can happen between the first and second reveal. Before the turn, you may have a guaranteed pair available. After an unfamiliar first card, that same pair cannot be collected with the one reveal left. The value of a choice depends on what you currently know and where you are in the turn.
Measure what remembering changes →
Build a small map with three kinds of entries
The Matching Pairs board has twelve cards: two each of CLOUD, RIVER, ACORN, COMET, FERN and WREN. Use row-and-column addresses on its four-column grid. Keep three categories in your notes or mental map: both locations known, one location known, and not yet located. Retire an entry when its pair is claimed.
Do not upgrade an uncertain recollection into a known location merely because it feels familiar. “RIVER was in the middle row, perhaps column 2” is different from “R2C2 is RIVER.” If a guessed location happens to match, the turn succeeded, but that does not show the location was remembered correctly before the reveal.
At the start of a turn, price the opportunity you are giving up
If both CLOUD locations are known, revealing them guarantees one pair in one turn, assuming the memory is accurate. Starting with an unseen card instead delays that guaranteed progress in exchange for a chance to learn or match elsewhere. This is an opportunity cost, not automatically a mistake: the choice depends on the practice goal.
For a low-error round, collect the known pair and reduce the map you need to maintain. For a deliberate information exercise, explore first and state that goal in advance. Do not declare one choice universally optimal from a single lucky match. Comparing total expected turns would require a specified knowledge state and assumptions about the unknown arrangement.
After the first reveal, ask a different question
Suppose your first card is RIVER and you accurately remember its mate at R2C2. Taking R2C2 finishes a guaranteed match now. Choosing an unseen second card abandons that certainty for additional information and may waste the turn. The fact that exploration can be useful does not make it useful at every stage.
Now suppose the first card is ACORN and its mate is unknown. You also know a complete CLOUD pair elsewhere. Revealing either known CLOUD as the second card guarantees a mismatch and supplies no new location information. The CLOUD pair cannot be collected during this turn because only one reveal remains. An unseen second card can instead reveal a new location and may be the ACORN mate.
That comparison is about the information state, not a promise of a match. If the new card is FERN, the turn still fails to clear a pair, but you can preserve two useful addresses for later. Update the map while both cards are visible rather than immediately planning the next click.
Measure decisions separately from lucky outcomes
The game counts one move when the two-card attempt is completed. Each claimed pair contributes 100 points, so clearing all six pairs produces 600 whether the route was short or long. The completion turn count is therefore more informative than the final points when comparing recall efficiency. There is no timer penalty in this Matching Pairs rule.
Keep two notes after a doubtful turn: what you believed before selecting the second card, and what the reveal actually showed. This distinguishes a correct recall, an exploratory discovery and a mistaken recollection. A lucky exploration should not erase the fact that a guaranteed match was available, just as a useful exploratory mismatch should not be dismissed as learning nothing.
Treat a rebuilt or restarted board as a fresh arrangement. The game can retain claimed-group progress, but it reconstructs the shuffled deck when mounted; it does not promise that every remembered address survives a reload. Re-establish locations before trusting a map from a different board session.
Two experiments to try
EXPERIMENT 1
A known pair at the wrong point in the turn
Use a practice position in which CLOUD’s two locations are known, but an ACORN revealed first has no known mate.
- Pause after that first ACORN reveal and count how many selections remain in the turn.
- Compare selecting a known CLOUD with selecting a previously unseen card. Say what each option can teach before seeing its result.
- Reveal the chosen card, then record its address even if it mismatches. Claim the known CLOUD pair on a later complete turn if you choose.
What to observe
A complete pair in memory is not collectible with only one reveal left unless it matches the current first card.
Why this comparison is useful
The unseen card provides at least a newly observed location. The known mismatching CLOUD provides neither a pair nor new location information, under the stated accurate-memory assumptions.
Your change and observation:
EXPERIMENT 2
Audit three kinds of success
Play a round with a simple row-and-column naming scheme.
- Before selecting a second card, label the decision “known mate,” “uncertain recall” or “exploration.”
- After the reveal, keep that original label rather than rewriting it as knowledge because it matched.
- At completion, compare the turn count and repeated address mistakes; use the fixed final score only as confirmation that all pairs were cleared.
What to observe
A shorter round can result from both better recall and fortunate exploration. The decision labels help separate them.
Why this comparison is useful
One round does not estimate an optimal strategy. This protocol makes a specific memory error visible and gives the next practice round a clear target.
Your change and observation:
Check your decision
Your first reveal is ACORN. Its mate is unknown. You know both CLOUD locations. What does selecting a known CLOUD as your second card achieve?
Answer and explanations
Use the current first card and remaining reveal allowance. Knowledge that was valuable before the turn may not be actionable now.
- It claims the known CLOUD pair.
Not supported. Only one reveal remains, and the open first card is ACORN.
- It guarantees a mismatch and adds no new location information.
Supported. That is the consequence if the CLOUD location is accurately known.
- It proves where ACORN’s mate is.
Not supported. Seeing an already-known nonmate does not locate the hidden ACORN.
Your next practice session
Try one round focused on eliminating repeated address mistakes. Track uncertain recollections honestly rather than trying to make every turn look like deliberate recall.
Scope and checks
This guide describes the current Playfield implementation. The experiments are proposed ways to practise, not measured player results. Rules were checked against the game source; the written scenarios and explanations received a separate review.
- Playfield Arcade game implementation — Primary source for this browser version’s board rules, scoring and state transitions.
Playfield Arcade · Updated 4 October 2026 · Original practice design, AI-assisted writing and code.