Show the model. Show the misses.

TAC forecasts are editorially informed estimates, not the output of a fully automated, independently validated statistical model. We consider recent team performances, underlying chance quality where available (including xG), home advantage, opponent strength, injury and suspension news, fixture congestion, likely rotation and tactical matchups.

We weigh those inputs using football judgment; we do not currently publish a fixed mathematical weighting formula or claim that each percentage comes from a reproducible simulation. When we provide win/draw/loss percentages, they represent our best pre-match assessment and should sum to 100%. Projected scorelines and xG are estimates, not guarantees.

We distinguish confirmed team news from reported information and our own assumptions. A forecast can be updated before kickoff when important news changes, but every update should retain its timestamp and previous published values.

Our projected XI considers player availability, recent starting selections, minutes played, recovery time, competition priorities, manager rotation patterns, positional roles and opponent-specific tactical needs. Expected substitutions can inform the analysis, but only the 11 predicted starters count toward the starting-XI accuracy tracker.

We identify uncertain selections as judgment calls rather than reporting them as confirmed. The official matchday starting XI is the benchmark for scoring, regardless of whether a predicted player subsequently appears as a substitute.

Match outcome: one correct prediction when the published win, draw or loss selection matches the final result in the competition's applicable match period. Exact score: one correct prediction only when both teams' predicted goal totals match the actual score. Cup extra time and penalty-shootout rules must be stated for each applicable pick.

Starting XI: count each correctly named starter against the official starting lineup. The denominator is 11 per match, so five tracked lineups represent 55 selections. A correctly predicted substitute does not earn starting-XI credit.

We report the sample size alongside accuracy. Small records such as 1/2 match outcomes or 38/44 lineup selections should not be interpreted as proof of long-term predictive skill. As the archive grows, we aim to evaluate probability calibration and error, not merely correct picks.

Forecast lock: predictions are timestamped when published. Pre-kickoff revisions must be clearly labeled and archived; after kickoff, the forecast is frozen for scoring and cannot be retrospectively edited to improve the record. Results and official starting XIs are checked against competition or club match reports.

Source transparency: we attribute external statistics and team news where they are used, identify the date of each data snapshot and distinguish reported facts from editorial inference. No claim is made that TAC operates a proprietary live-data feed.

Limitations: football outcomes are noisy. Red cards, refereeing decisions, individual errors and late lineup changes can overturn reasonable pre-match expectations. These are analysis and editorial forecasts, not betting advice or bookmaker odds.