Roosters v Knights

Grand Final — one game for the 2026 premiership. The Roosters come the long road after a Qualifying Final loss to the Panthers, having since put 46 on the Sharks and beaten the Dolphins 36-20 at Suncorp in the Preliminary Final, with Sam Walker back and the squad at full strength. The Knights have broken every barrier these finals — the Allianz bogey against the Rabbitohs, the Warriors' 5-0 H2H at Eden Park, and the three-time premiers' 4-1 H2H at Accor — but arrive without Dylan Brown at halfback and with Phoenix Crossland suspended.

Headline averages from every completed match this season (Roosters n=26, Knights n=26). Bars diverge from centre — longer bar wins on the metric (green where lower is the goal, like points conceded or penalties).

1H = first half, 2H = second half. 1H/2H averages across the full 2026 season. "1H win → match win" is how often leading at half-time converts into the win.

PCS (Player Contribution Score) is a correlation-informed rating built from play-by-play stats — run metres, line breaks, tackle breaks, missed tackles and so on. Each stat is weighted by how strongly it correlates with winning margin across 1,000+ matches. Aggregated to position group over a 15-game rolling window.

Why position groups matter: the spine (FB, halves, hooker) controls structure, the forwards set the ruck and tempo, the backs finish chances. A team can be middling overall but elite at one group — and that often shows up as a directional edge in the result.

Strictly the very first scoring play of each match — try, penalty goal, or (rarely) field goal. Teams that open the scoring tend to lead at half-time about 65% of the time and win about 60% — so the opening play matters. Penalty-goal openings are shown as raw counts because most games never open with a PG; the position-group breakdown only fires when the first play was a try.

Consistency Index (CI) measures how predictable a team's results are. It's built from the standard deviation of their margin and their points-conceded across the season. Higher CI = tighter range of results, more dependable.

Disruption Score (DS) measures how far the named squad is from each team's best 17 — weighted by position importance (a halfback out hurts more than a bench prop) and by how fresh the absence is. Higher DS = more disrupted, less the side coaches game-plan around.

Anytime Try Scoring — Matchup Likelihood: each side's position groups ranked by how often they score against how often this opponent concedes.

Roosters Backs are the standout anytime bet — the top combined matchup score in this game at 1.96/game (Roosters backs score 2.26/game; Knights concede to the backs at 1.67/game).

The Roosters' 5-0 H2H dominance is the strongest signal in the match, but three of those five meetings were decided by 2 points. The Knights have defied every model prediction this finals. Grand Final compression and the Knights' superior consistency (CI 0.553 vs 0.493) halve the model's +12.7 margin. Crossland's suspension deepens the Knights' spine crisis — Brown and Crossland both absent for the biggest game of the year.