· Full time · Lumen Field, Seattle, WA
Right call — Belgium through (won 4–1).
A knockout can't end level. This is each side's chance to go through — win in 90 minutes, in extra time, or on penalties. It's a different question from the win probability above, which is who wins the match in regulation.
Heads up: this tie has a real chance (about 16%) of going to penalties, and a shootout is close to a 50/50 — they're near-random, so we don't pretend to call them. The deeper a tie goes, the more it's a coin flip.
Who wins in normal time — the 90-minute win/draw/loss. “Who advances” above folds in extra time and penalties.
USA 33%, draw 28%, Belgium 38%.
This is our latest read, with current confirmed team news folded into the model (see “What we considered” below). Our track record always grades the locked pre-match call we made before lineups were posted — so this number may differ slightly from the one on the results page. We never rewrite a locked prediction.
Every factor the model weighed for this match — including the ones worth nothing. Each is capped so it can nudge, never flip, a clear matchup. A “no change” line means we considered it and had no reliable data — that transparency is the point, not a gap.
USA is the host nation: +55 Elo home-field advantage (applied in the goal model).
Belgium comes in with a subtle but real edge in pedigree and expected output, yet this is no coronation. The home crowd gives USA genuine lift, and the model flags genuine upset risk here — fine margins, high stakes, one goal separates them.
Belgium's midfield dominance against USA's ability to turn home advantage into tempo and transition.
Home field narrows a gap that would otherwise favour Belgium; upset risk runs through the entire match.
AI-generated context based on a statistical model. Full-time/regulation result only. Not betting advice; no specific current team-news is used.
Display-only — none of this is used in the prediction above.
Context only — these confirmed lineups are not used in the prediction above. By kick-off, public team news is already reflected in any market line.
Lineups not available. We only show a confirmed starting XI when we can match this exact fixture to a Flashscore event — we never guess or invent one.
Venue not available. We only show a stadium when we can match this exact fixture to a Flashscore event or it's in our scheduled-venue data — we never guess one.
Venue is shown for information only and is not used in the prediction.
Real measured stats for this match — shown for transparency. They are not part of the prediction above, which is a pre-match estimate from team ratings.
Match stats not available. They appear once a match is under way and Flashscore is carrying live stats for it — we never invent them.
Screenshot to share — the honesty note travels with it.
Predictions are estimates, not advice. WisePicks is a fan dashboard, not a betting product. The numbers come from a simple statistical model. They carry real uncertainty and can be wrong. We don't show odds or encourage gambling.
Expected goals and control are derived estimates from the model's ratings — indicative leans, not measured match stats.
Over 2.5 goals: 48% · Under 2.5: 52%.
Who advances
No confirmed lineup, and no tracked key players for USA (0).
No confirmed lineup yet — not applied (0). We never guess at availability; we wait for the confirmed XI.
shorter-than-average squad (5 tallest outfielders ~189cm vs a ~190cm field) → -1 Elo set-piece nudge. Estimated from squad height (a rough proxy, weighted toward aerial defence); modest by design.
taller-than-average squad (5 tallest outfielders ~192cm vs a ~190cm field) → +4 Elo set-piece nudge. Estimated from squad height (a rough proxy, weighted toward aerial defence); modest by design.
Over 7 tournament matches: xG-for 1.10/g, xG-against 0.92/g (xGD +0.19/g, +0.2 SD vs the field). Shrunk for a 7-game sample → +2 Elo. The team created better chances than it converted; a bounded, modest chance-quality nudge.
Over 19 tournament matches: xG-for 1.39/g, xG-against 1.16/g (xGD +0.23/g, +0.3 SD vs the field). Shrunk for a 19-game sample → +5 Elo. The team created better chances than it converted; a bounded, modest chance-quality nudge.
Disabled by flag (0).
Disabled by flag (0).
Disabled by flag (0).
Considered — deferred (0). Extreme heat/humidity mainly lowers tempo and scoring, not who wins; we don't apply it as an outcome factor.
Considered — no weather data for this venue/date (0). Weather only nudges the goal total; it does not move who wins.
Host-crowd effect is already counted inside Home advantage — not double-counted here (0).
Player availability uses a curated key-player heuristic against the confirmed starting XI (via Flashscore), not a fitted impact model. Weather is real match-day data from Open-Meteo and only nudges the goal total — its measured effect on who wins is ~zero. We only ever apply a factor on real, confirmed data; otherwise it is exactly 0.
At least one factor moved the prediction on real confirmed data; the rest were considered and contributed nothing.