How 38-0 Simulates a 38-Match Season

When an XI is complete, 38-0 simulates one decisive 38-match campaign. It does not average 1,000 seasons and it does not claim to forecast how a historical team would perform in real life. The output is a playable football what-if: one run through a model where team construction matters and randomness ensures that a strong side is never guaranteed a perfect record.

This design keeps the final screen understandable. The wins, draws, losses, goals, scorers, clean sheets, and points all belong to the same virtual season. Replaying with the same XI can produce a different result because matches are sampled again, just as a real match is not determined by squad strength alone.

The engine first maps each selected player to the corresponding formation slot. Natural and explicit alternate positions receive full fit weight, compatible roles receive a reduced weight, and incompatible assignments receive a larger penalty. The combined fit becomes one input to chemistry alongside era diversity, legend status, and groups of players from the same club.

Players are then grouped by role to build attack, midfield, defence, and goalkeeper strength. Attackers lean on shooting, pace, and dribbling; midfielders on passing, dribbling, and physical ability; defenders on defending, physical ability, and pace. Chemistry scales the outfield units, so a collection of high individual ratings can still lose effectiveness when the formation is poorly filled.

A 38-game league season requires 19 opponents. The simulator selects an opponent pool for the chosen league and creates one home and one away match against each club. In all-leagues mode, the pool can draw clubs from across the five supported competitions. The resulting 38 fixtures are shuffled into a season order.

Opponent ratings are converted into attack, midfield, defence, and goalkeeper values with some match-run variation. Home advantage is applied to the home side. This means the schedule contains both structural difficulty from the opponent pool and uncertainty inside an individual run.

For every fixture, the user's attack is compared with the opponent's defence, and the opponent's attack is compared with the user's defence. Those gaps, plus home advantage, produce expected-goal values. Expected goals are not displayed as a promise; they are the rates used to sample a possible scoreline.

Goals are drawn from a Poisson distribution, a common starting point for football score models. A high expected-goal value makes multiple goals more likely but does not force them. That is how an elite XI can draw a match it was expected to win, or a weaker side can occasionally produce an unusually strong result.

Each win earns three points and each draw earns one. The game totals the record, goals for, goals against, and clean sheets, then estimates a final league position from the points total. It does not simulate a complete table of 20 independently managed clubs, so the position should be read as a familiar summary of the season rather than an official table reconstruction.

Goals and assists are distributed across the XI using player quality, position, and attacking attributes. Forwards are more likely to finish chances, creative midfielders are more likely to supply assists, and the selected goalkeeper receives clean-sheet credit. These events make the season readable while remaining downstream of the team result.

No compact browser simulation can reproduce tactics, injuries, substitutions, fixture congestion, managerial decisions, changing form, or the competitive context of different eras. The ratings and tuning choices are editorial, and historical players never faced the exact mixed opponent pools created by the game.