Deep Cut gives you one question a day and five slots to fill, and it inverts the usual instinct: the obvious answer is worth almost nothing. Every player on the day’s list is scored on how obscure he is next to the others on that same list, from 1 for the most famous up to 99 for the man nobody remembers - that number is our own estimate of how well known he is, not a fan survey. Name somebody who genuinely fits the question but sits outside the day’s list and he is credited too, as a deep cut, without filling a slot. Search a player and select him; if he fits, his name lands in a slot and a beat later his rarity arrives. Fill all five slots and your total is your score. Wrong guesses cost nothing at all and there is no clock, so the only real decision is how brave you are - reaching past the three names everyone thinks of first is the entire game. Beat the target score and the day is a win, which means five obvious answers can fill every slot and still lose.
Deep Cut is airball.gg's inverted trivia game: the rarer the answer, the better. It shares one scoring model with Crowd Favourite, used upside down, so the two games can never disagree about which players are obvious. That model is an OBVIOUSNESS score, not a survey - we have never asked a room of fans anything, and nothing here claims we did. It is derived from the recognisability score already used to build our eligible player pool, then ranked within each individual question's answer list. Ranking within the list matters: raw recognisability is heavily skewed, with most players clustered together and a handful of superstars far above, so printing raw numbers would give a board reading 5, 5, 6, 29. Ranking spreads them across a readable range and says the truer thing - on a list of MVPs, being the least famous MVP is not the same as being obscure. Questions are computed from the database and do not repeat within 90 days.
What exactly is the rarity number?
How well known a player is relative to the others who fit the same question, on a scale of 1 to 99. It is derived from our own recognisability scoring, not from asking anybody. It is never 0 or 100, because neither 'nobody knows him' nor 'everybody does' is a claim the model can support.
Did you survey real fans?
No, and the game never says we did. This is our own estimate of how obvious an answer is, computed from the player database.
Can I fill all five slots and still lose?
Yes, and that is the point. The five most obvious answers on a typical day total around thirty between them, and the bar is a hundred and fifty. The game rewards reaching past the names that come to mind first.
What happens if two players share a surname?
The surname is rejected rather than guessed, and you can type the full name instead. Two players on the same list can score very differently, so crediting the wrong one would not be a rounding error - it would be the round.
Is it the same question as Crowd Favourite?
Never on the same day. Both games draw from the same pool of computed questions, and the generator guarantees the two differ.
Player stats, rosters, and career data reflect the most recently completed NBA season until the new season is underway.
2026 airball.gg • Privacy • Terms • Contact
airball.gg is an unofficial fan project, not affiliated with, endorsed by, or licensed by the NBA, the NBPA, or any of their teams. Player names and statistics are historical public record, used for informational and entertainment purposes only. Simulated games, results, and stat lines are fictional.