Perfect Game in Numbers
Perfect Game asks the same question in three rounds every day: which Nintendo game sold better, which players like more, which critics rated higher. This page shows how well players actually answer those questions. Not which games are popular, that is what the ranking is for, but how accurate the guesses are and where intuition keeps getting it wrong.
📊 Period covered: 2026-08-02 to 2026-09-28How many of the three rounds land
How completed games break down by number of rounds guessed correctly. Three means a Perfect Game, all three right.
Which ranking is hardest to guess
Share of guesses that landed on the real winner in each round. A high value means players are confident on that kind of question, a low one means they are often guessing.
Players struggle most with User rating: that is where the smallest share hits the real winner. They are most sure-footed on Critic rating. That is the whole appeal of Perfect Game. Sales success, player love, and critical acclaim often pull apart for Nintendo, and where they diverge the most, the most mistakes happen.
Whether the majority is smarter than the individual
Across all rounds, a single guess hits the correct winner 66.4 percent of the time.
Take instead only the most-picked answer in each round, the one the majority agrees on, and that is right 77.2 percent of the time.
So the majority hits more often than an average individual guess. That is the wisdom of the crowd in miniature: individual players’ errors cancel out, and the most-chosen title is the real winner more often than one random player’s pick would be.
How the numbers are produced
Everything here is based on completed runs of the daily Perfect Game in the normal three-card mode. Each finished game records which title was picked in each round. The correct answer of every round has been stored alongside the puzzle since August 2026, so each guess can be classified as a hit or a miss without recomputing anything.
Hardcore mode is deliberately excluded. It has six cards per round instead of three and therefore a completely different chance of a hit, and mixing the two would make any rate meaningless. Only games from the switch to three cards onward are counted, earlier rounds are not comparable.
No game names and no absolute counts per title are published, only shares and rates. Otherwise the numbers would reveal which answers reliably work in which rounds, and that would spoil the daily puzzle for everyone else.