Australia and Brazil drew 1–1. Bawler's model favoured Brazil at 53% before kickoff and got it wrong — it gave the draw 26%. It forecast 2.5 goals and the match produced 2. Australia had 7 shots (2 on target) to Brazil's 16 (5), with 35% of the ball against 65%. Expected goals finished 0.5 to 1.2.
Published before kickoff and settled afterwards — misses included. Every call is time-stamped to the Bitcoin blockchain before the match so it cannot be edited later.
What we expected before kickoff.
Each row is drawn on one scale, so a bar means the same length on both sides. Second-half goals are our full-match forecast minus our first half. Shots and corners are not shown for this match — one of these national teams has too few recorded games for us to forecast them.
What we published for this match, strongest first — our three strongest above the line.
Chance of at least one goal before half-time: 68%
Shots and corners are not shown for international matches yet — national teams play too few games for us to forecast them well.
Australia and Brazil drew 1–1. Bawler's model favoured Brazil at 53% before kickoff and got it wrong — it gave the draw 26%. It forecast 2.5 goals and the match produced 2. Australia had 7 shots (2 on target) to Brazil's 16 (5), with 35% of the ball against 65%. Expected goals finished 0.5 to 1.2.
Australia 1–1 Brazil, played on 25 Sept 2026 in the World: International Friendly. Bawler's model had gone into it with Australia 21%, the draw 26% and Brazil 53%.
Full-time: Possession 35%–65%, Expected goals (xG) 0.5–1.2, Shots 7–16, Shots on target 2–5, Shots inside box 3–8, Big chances 2–2, Corners 9–8, Saves 4–1, Fouls 7–8, Offsides 0–0, Yellow cards 1–0.
Yes — both Australia and Brazil scored, finishing 1–1. Bawler's model had put both teams scoring at 48% before kickoff.
Every call is published before kickoff and graded afterwards, misses included. On this match: Our call — we said Brazil or draw 79%, it finished Draw (26% called); Total goals — we said 2.5, it finished 2; Both teams to score — we said 48%, it finished Yes.