USA vs. Bosnia-Herzegovina - Post-Event Review

2026 World Cup, R32 (Match 81) | July 2, 2026 | Levi's Stadium, Santa Clara CA (open-air)

Predicted vs. Actual | Calibration: ~92%

TL;DR

Result: USA 2-0 Bosnia-Herzegovina (HT 1-0; Balogun 45', Tillman 82'). Co-host favorite delivered exactly as the modal script said — a clinical, reactive-shape 2-0 over a low-block underdog whose 40-year-old talisman never fired.

• Winner: ✅✅ USA (~70% advance / 57% in 90') — won in regulation, no ET, no pens. Advance bet paid comfortably; the home co-host + weakest-3rd-place-draw edge was real and justified.
• Scoreline: ✅✅ EXACT modal — consensus modal was USA 2-0 (g4 strict #1 at 22%, mm3 #1 at 14%); actual USA 2-0. Not "in top-3" — this was the #1 pick and it hit.
• Who Scores: ✅✅ #1 hit (Balogun, 45') — 2 of 3 models' strict #1 was Balogun (g4 22% / g5 30%); he scored once. Pulisic (consensus-table #1, ~25%) blanked — sub-risk materialized. Tillman (USA #3, ~12%) added the second (82').
• Finding 12 (0% conservation): ✅ — USA pressed from the start, breached by 45'. No stale tilt, exactly as Poch-aggressive predicted.
• Finding 16 (Džeko age-decline): ✅ — 40yo captain blanked, subbed early, BIH 0.25 xG / 0 big chances. Milestone → minutes not goals, holds again.
• Finding 8 (possession flip): ⚠️ fired AGAINST the projection — models predicted USA 60-62% possession; actual USA 48% / BIH 52%. USA won with the reactive shape at home. 2nd R32 validation (after MEX-ECU).
• Finding 22 (blowout Mode A): ⚠️ partial — 10-12% on 3+ margin staked; actual 2-0 (just under). Direction right, magnitude one goal short.

Calibration Breakdown: Winner 25/25 · Scoreline 25/25 · Who Scores 15/20 · Findings 17/20 · Model-breaking 10/10 · deductions 0 = 92/100 → ~92% (A band)

1. Winner

USA won 2-0 in 90' — advance bet won, no ET, no pens. The §7K knockout tree resolved on the #1 path (USA win in 90', ~57%): Balogun broke the low block just before HT, Tillman sealed it late. The modal-favorite-won-and-was-strict-modal bar is cleared cleanly. The co-host home edge + weakest-3rd-place-draw + Poch-press-vs-Tier-3-low-block thesis (validated twice in the group vs PAR/AUS) held up exactly.

2. Scoreline

USA 2-0 = EXACT modal hit. The consensus modal scoreline was USA 2-0 (g4's strict #1 at 22%, mm3's #1 at 14%) with USA 1-0 (g5 #1 at 16%) as runner-up. USA 2-0 fired. Under the strict rubric this is a #1-pick hit (g4's 22% ≥ 15% threshold), not "in top-3" inflation. Clean sheet was also the consensus read (BIH 0.25 xG, 0 big chances).

3. Who Scores

• USA #1 Balogun ✅ (45') — g4's strict #1 (22%) and g5's strict #1 (30%); scored once. 15/20 component.
• USA consensus-table #1 Pulisic ⚠️ — ~25%, did NOT score; the flagged sub-risk (no 90' since mid-March) materialized. Honest flag, no misleading-credit taken.
• USA #3 Tillman ✅ (82') — secondary pick delivered the sealer; was in the predicted top-5, so no F19 deduction.
• BIH #1 Džeko ❌ — ~10-17%, blanked; BIH scored 0. F16 age-decline fully validated.

4. Weather

Forecast: 24°C mild, sunny, marine layer → F11 NEUTRAL. Actual: open-air Levi's, mild, no heat/storm impact. ✅ Validated — weather was a non-factor exactly as profiled. No F11 penalty applied (correctly).

5. Possession (FINAL)

USA 48% / BIH 52%. ⚠️ Possession FLIP (Finding 8, 2nd R32 validation after MEX-ECU). All three models projected USA 60-62% possession; the actual inverted. USA won with 48% + a clinical counter/reactive shape — the home co-host favorite played reactive once ahead. This is a calibration finding (see Key Learnings), not a model-breaking miss: possession direction was wrong, but the result prediction was right because xG efficiency (USA 0.89→2, BIH 0.25→0) dominated.

6. xG

USA 0.89 / BIH 0.25. Low-event, ruthlessly clinical. USA over-performed xG (2 from 0.89); BIH barely existed (0.25 xG, 0 big chances). Consistent with the "BIH xG ceiling ~0.8, USA finishes chances" thesis. Džeko's age-declined attack never threatened.

7. Cards / Penalties

• Fouls: USA 7 / BIH 13 — BIH committed nearly 2×, frustrated chasing the lead.
• Yellows: profiled 3-4; Claus tournament-discounted card rate held directionally.
• Penalties: 0. Claus pen band 25-35% → 0 pens. The band held at the low end (0/3 WC sample continued). Referee profile accurate.

8. Lineup

• USA: As predicted — 4-2-3-1, full-strength MD1 XI returned (Pulisic fit, Balogun #9, Freese in goal). ✅
• BIH: As predicted — 4-4-2 low block, Džeko started (subbed early). ✅
• No lineup surprises either way.

9. GK

• USA: Matt Freese ✅ confirmed — untroubled (BIH 0.25 xG), clean sheet, 0 saves of note.
• BIH: Nikola Vasilj ✅ confirmed — 3 saves, 2 GA, not at fault (low xG/goal). No heroics, exactly as predicted (28 caps, no confirmed-form credit, doesn't meet veteran-GK F22 exclusion). The "no GK heroics" call was correct and mattered: it kept the draw band honest (20%) without inflating BIH.

10. Substitutions

• USA: Pulisic sub-risk materialized (flagged pre-match); Tillman's 82' goal — a late-arriving runner / sub-impact finisher, within the predicted top-5.
• BIH: Džeko subbed early (F16 age-decline, flagged at 60-65'). No super-sub impact (Alajbegović/Mahmić never unlocked the block).

11. Finding 12 (Conservation)

Predicted: 0% conservation (Poch aggressive + KO must-win + co-host statement). ✅ VALIDATED. USA pressed from the start, breached by 45', chased a second to 82'. Zero stale tilt — the no-conservation call was correct and is now the default for Pochettino KO matches.

12. Key Learnings

1. Finding 8 (possession flip) extends to HOME KNOCKOUT favorites — validate & refine. This is the 2nd R32 case (after MEX-ECU) where the home favorite won while losing the possession battle by choice: once ahead vs a broken low block, the favorite sits in a reactive shape and counters. Refinement: when a home KO favorite is projected to lead, do NOT project 60%+ possession — the lead-state reactive shape flips it. Possession % is a flow indicator, not a result predictor (already in F8, but writers keep over-projecting it). Action: cap projected favorite possession at ~52-55% in KO matches with a projected lead; do not let possession projection drive the scoreline.

2. Finding 16 (age-decline) fully validated for the 4th time (Džeko). 40yo, 0 WC goals, subbed early, BIH 0.25 xG. The milestone-motivation rule continues to hold: milestone → minutes, not goals, for 38+ strikers. Džeko joins Ronaldo, Ayew, Modrić-team as validated age-decline cases. No change to the rule — but reinforce: the age-decline discount on the #1 BIH scorer (~10-17%) was correctly applied by all three models and BIH scored 0.

3. Finding 22 (blowout Mode A) calibrated slightly high vs low-block KO — refine. The 3+ margin band was staked at 10-12%; actual was 2-0 (just under). Mode A (in-form attack + compact block + home crowd) fired in direction but the low block + Vasilj's competence limited the damage to 2. Refinement: in KO R32 vs a competent (non-veteran but capable) GK behind a low block, shade the 3+ band down ~2-3pts (8-10% vs 10-12%); reserve full 12-15% for the non-veteran-GK + broken-block profile.

4. Finding 12 = 0% conservation confirmed as Pochettino KO default. Co-host + must-win + aggressive manager = zero conservation, validated. This is now the stable prior for all remaining USA KO matches (R16 vs Belgium). No conservation overlay needed.

5. WDM + home-co-host deviation was correctly sized (no over-adjustment). fav_share ~0.69, USA advance ~70% — USA won comfortably in 90'. The deviation cap (R32-1) was borderline-but-justified: the +6-9pt home co-host edge translated cleanly. No corrective action — the framework held.

Knockout Bracket — R16 (USA's path)

USA advance to face Belgium in Seattle. USA's R32 form (clinical 2-0, clean sheet, Balogun scoring, no ET fatigue) is a clean preparation result; the Pulisic sub-risk will re-surface as a Watch variable vs a Tier-1 UEFA opponent.

Post-event review for the USA-BiH R32 task. Category 0 PASSED: result USA 2-0 BIH confirmed from ESPN Matchcenter (gameId 760494) + scout brief cross-source. FINAL stats only (HT 1-0 not used). STRICT rubric (Jun 30) applied — modal-pick scoring, no "in top-3" inflation. Sources: sources/matches/usa-vs-bih/post-event/ (scout brief) + pre-match consensus + 3 per-model analyses. No betting odds, no expert picks.