The mathematical distinction
A permutation is an ordered arrangement. If A-B-C and C-B-A count as different results, order matters and the problem involves permutations. A combination is an unordered selection. If choosing A, B, and C represents the same group regardless of listing order, the problem involves combinations. Traditional formulas such as nPr and nCr apply when arranging or selecting items from one set.
Why prediction cards are a product of choices
A multi-match prediction card normally chooses one outcome from each match. Those matches already occupy distinct positions, so the central calculation is the product of the available choices at each position. Two outcomes for match one and three for match two create 2 × 3 = 6 complete paths. This structure is more precisely a Cartesian product than a textbook nCr calculation.
When order does not create a new card
Suppose a card contains a home outcome for Match A and a draw for Match B. Displaying Match B first does not change the underlying predictions, so it should not create another card. Counting both display orders would duplicate the same path and overstate the total coverage.
When permutations really do matter
Order matters in questions about sequences: the possible order of scorers, the finishing order of teams, or a step-by-step event path. In those cases A then B differs from B then A. It can also matter for display or processing rules, but reorganizing an ordinary set of match predictions does not usually change the represented outcome path.
Why tools still say permutation generator
In everyday product language, a sports prediction permutation generator often means a tool that expands every allowed outcome pattern. That label is understandable because the output lists many arrangements across match positions. For accurate counting, focus on the rule used: one selected outcome per match and multiplication across matches.
A quick terminology check
Ask two questions before choosing a formula. Are you selecting items from one shared set, or choosing one option from several separate match sets? Does changing only the display order create a meaningfully different prediction? For standard multi-match cards, separate choice sets and the multiplication rule usually provide the correct model.