I recently crossed a 3000 puzzle rating on Chess.com. That puts me in roughly the top 0.15 percent of players on the platform. I am not writing this to talk about chess. I am writing it because the muscles that get you to 3000 are the exact muscles product strategy runs on, and almost nobody connects the two.
Pattern recognition is most of the game
Strong puzzle solvers do not calculate every position from scratch. They recognize shapes — a back-rank weakness, an overloaded defender, a pinned piece — and the candidate moves come for free. The calculation is spent confirming, not discovering.
Product is the same. After enough cycles you stop analyzing every feature request from zero. You see the shape: this is a buyer-versus-user mismatch, this is a vanity metric dressed as traction, this is a habit-loop problem wearing a feature costume. Pattern recognition is what lets a senior PM reach the right answer in minutes that takes a junior PM a quarter.
Calculation under constraint
In a puzzle you have a fixed position and a finite clock. You cannot try everything. You pick the two or three candidate moves most likely to matter and you calculate those deeply, ignoring the rest. Calculate too broadly and you flag; calculate too narrowly and you miss the refutation.
Roadmapping is candidate-move selection under a budget. You cannot build everything, so you identify the two or three bets most likely to move the outcome and you go deep on those. The skill is not generating options — it is ruthlessly pruning to the ones that matter and committing real analysis there.
Seeing the opponent's best reply
The mistake that ends most puzzle attempts is assuming the opponent plays the move you want. Strong players calculate the best defense, not the convenient one. Every plan is stress-tested against the strongest reply.
In product, the "opponent" is reality: the user who behaves differently than your deck assumes, the competitor who responds, the edge case that breaks the happy path. A strategy that only works if everyone cooperates is not a strategy. The discipline of asking "what is the strongest thing that could go wrong here, and does my plan survive it?" is pure chess.
Sitting in the uncomfortable position
Sometimes the puzzle has no flashy combination. The right move is quiet — improve your worst piece, fix the weakness, accept a slower win. Beginners reach for the sacrifice; strong players take the boring, correct move.
Product has the same trap. The flashy launch, the AI feature for its own sake, the rewrite — these are the sacrifices that feel like progress. Often the correct move is quiet: fix the onboarding step that loses 30 percent of users, upgrade the thing nobody demos. On Listen2RE the highest-leverage move one quarter was not a feature at all — it was upgrading voice quality. Boring, correct, and it beat everything flashy.
Why this matters for hiring
Strategy is not a vibe. It is calculation, pattern recognition, and disciplined pruning under uncertainty — trainable, measurable skills. A 3000 puzzle rating is one objective proxy that those muscles are strong. I bring the same ones to a roadmap.