main exact score. Index 0.119948; count probability 0.088917.
Project Turtle / postgame audit
Belgium-USA finished outside the locked C4.
Final: USA 1-4 Belgium. The locked pregame C4 was 2-1, 1-2, 3-1, 1-1. Turtle read the hot Belgium/BTTS game state, but failed the exact-score test by truncating the Belgium cushion ladder at 3-1.
Locked C4
Belgium-USA C4: 2-1, 1-2, 3-1, 1-1
main exact score. Index 0.087791; count probability 0.074896.
main exact score. Index 0.083238; count probability 0.053757.
containment score. Index 0.082245; count probability 0.098048.
Node success means the final regulation score lands inside these four score nodes. No market price entered this C4.
Postgame analysis
Belgium 4-1 USA: C4 miss, support reads hit.
This is graded plainly. A late or error-assisted fourth goal has no special exception. The final score is outside C4, so the practical protocol test failed.
| item | settlement | audit note |
|---|---|---|
| Locked C4 | MISS | 2-1, 1-2, 3-1, 1-1 did not contain 4-1 Belgium. |
| Belgium direction | HIT | The favorite win/cushion branch was real. |
| BTTS thread | HIT | Both teams scored; the BTTS-heavy read was directionally right. |
| Over 2.5 thread | HIT | The game cleared three goals. |
| Exact-score engine | FAIL | 3-1 implied an adjacent ladder check; 4-1 was not formally evaluated. |
Failure classfavorite-cushion ladder failure
Once Belgium by 2+ and BTTS survived recursive attack, the proof needed to test 4-1, 3-2, and 4-2 before locking. It did not.
Repairadd R-ladder
If a favorite-cushion BTTS score is selected and P(over_3_5) >= 0.40, at least one adjacent ladder score must enter C4 or be rejected by a stronger numeric inequality.
Verdictpause exact-score locks
This match fails the practical standard. Turtle should not lock the next exact-score card until the ladder rule is implemented and tested on prior files.
Prediction engine
Evidence files were created before page design.
| metric | value | formula_or_rule |
|---|---|---|
| fixture_status | pregame_valid | lock_time=2026-07-06T18:57:54.364698-04:00 < kickoff=2026-07-06T20:00:00-04:00 |
| lambda_BEL | 1.8137 | ((A_BEL + D_USA)/2)*knockout_caution*belgium_h2h_pressure*volatility_release |
| lambda_USA | 1.5277 | ((A_USA + D_BEL)/2)*knockout_caution*usa_home_balogun*volatility_release |
| A_BEL | 2.2025 | 0.18*GF50 + 0.20*GF10 + 0.25*GFWC + 0.37*realm_GF |
| D_BEL | 0.9687 | 0.18*GA50 + 0.20*GA10 + 0.25*GAWC + 0.37*realm_GA |
| A_USA | 2.0840 | 0.18*GF50 + 0.20*GF10 + 0.25*GFWC + 0.37*realm_GF |
| D_USA | 1.4742 | 0.18*GA50 + 0.20*GA10 + 0.25*GAWC + 0.37*realm_GA |
| P_BEL_win | 0.4463 | sum P(a,b), a>b |
| P_draw | 0.2265 | sum P(a,b), a=b |
| P_USA_win | 0.3272 | sum P(a,b), b>a |
| P_BTTS | 0.6552 | sum P(a,b), a>0 and b>0 |
| P_over_3_5 | 0.4284 | sum P(a,b), a+b>=4 |
| human_signal_before_lock | none_named_for_this_match | Human Signal Preservation gate inactive but recorded |
| clean_sheet_override | inactive | no extreme verified non-concession streak or opponent failed-to-score gate found |
Opponent-threshold realm
Belgium is tested against USA-realm rows; USA is tested against Belgium-realm rows.
| team | target_opponent | epsilon_used | realm_rows | readiness | GF50 | GA50 | GFWC | GAWC | realm_GF_shrunk | realm_GA_shrunk | realm_scored_rate | realm_clean_sheet_rate | realm_BTTS_rate | realm_over_3_5 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Belgium | United States | 0.75 | 9 | medium | 2.040 | 0.860 | 2.250 | 1.000 | 1.818 | 1.092 | 0.673 | 0.323 | 0.455 | 0.375 |
| United States | Belgium | 0.75 | 8 | medium | 1.980 | 1.260 | 2.500 | 1.000 | 1.737 | 1.777 | 0.792 | 0.184 | 0.609 | 0.409 |
Recursive contradiction audit
Every branch is attacked before it can survive.
| audit | branch | candidate_scores | supporting_fact | strongest_opposing_fact | numeric_test | decision |
|---|---|---|---|---|---|---|
| favorite_clean_sheet | Belgium clean sheet | 1-0, 2-0 | Belgium clean-sheet rate realm=0.323; count mass bel_clean=0.182 | USA scored rate realm=0.792, current WC GF/match=2.500, Balogun eligibility restored | USA realm scored 0.792 >= 0.70 and WC GF 2.500 >= 1.50 | excluded_from_C4_unless_extreme_clean_override; no override active |
| underdog_clean_sheet | USA clean sheet | 0-1, 0-2 | USA clean-sheet rate realm=0.184; count mass usa_clean=0.128 | Belgium scored rate realm=0.673, current WC GF/match=2.250, March H2H Belgium scored five | Belgium realm scored 0.673 >= 0.65 and WC GF 2.250 >= 1.50 | excluded_from_C4_unless_extreme_clean_override; no override active |
| BTTS | both teams score | 2-1, 1-2, 3-1, 1-1, 2-2 | BTTS thread=0.655; Belgium realm scored=0.673; USA realm scored=0.792 | combined clean-sheet mass=0.345 | BTTS 0.655 > non-BTTS 0.345 | passes; C4 allowed to be BTTS-heavy only after clean-sheet branches are explicitly attacked |
| no_BTTS | at least one side blanked | 1-0, 2-0, 0-1, 0-2, 0-0 | non-BTTS mass=0.345 | both teams show high realm and current-tournament scoring; no extreme non-concession override is active | non-BTTS 0.345 < BTTS 0.655 | not promoted to C4 |
| low_draw | low draw | 1-1 | draw thread=0.226; draw_low spread=0.134 | Belgium H2H 5-2 and both current attacks can break 1-1 upward | 1-1 remains in S10 and is highest draw-low representative | included as C4 containment branch |
| high_draw | high draw | 2-2 | over3.5 thread=0.428; USA recent volatility; Belgium conceded in current WC knockouts | under3.5 mass=0.572 and knockout caution suppresses chaos | 2-2 is below 3-1 by TurtleIndex and duplicates the over-3.5 slot without winning the draw slot from 1-1 | excluded from C4 because 3-1 has higher TurtleIndex and represents the stronger over-3.5 branch |
| favorite_by_one | Belgium by one | 2-1 | BEL win thread=0.446; bel_by_1 spread=0.205; March H2H 5-2 | USA home setting and Balogun availability make a clean Belgium favorite card unsafe | 2-1 is the highest Belgium-by-one BTTS representative in S10 | included in C4 |
| underdog_by_one | USA by one | 1-2 | USA win thread=0.327; usa_by_1 spread=0.173; current WC GF/match=2.500 | Belgium strength rating and H2H win prevent USA from becoming favorite branch | 1-2 is the strongest USA-by-one representative and remains above clean-sheet branches by TurtleIndex | included in C4 |
| favorite_cushion | Belgium by 2+ | 3-1, 2-0, 3-0 | bel_by_2plus spread=0.241; March H2H 5-2 | USA home scoring and knockout volatility make a two-goal Belgium cushion less stable than by-one | bel_by_2plus 0.241 > bel_by_1 0.205 and 3-1 has higher index than 2-2 | include 3-1 as the favorite-cushion over-3.5 branch |
| upset_branch | USA win | 1-2 | USA win thread=0.327; USA current scoring remains live | Belgium is not a soft opponent; H2H and rating keep upset to by-one only | USA by one 0.173 > USA by 2+ 0.154 | include 1-2, reject USA cushion |
| over_3_5_tail | four or more total goals | 2-2, 3-1, 1-3 | over3.5 thread=0.428; both teams have current-tournament 3+ goal games | under3.5 thread=0.572 | Only one over-3.5 score can survive; 3-1 outranks 2-2 and 3-2 by TurtleIndex | include 3-1; reject 2-2 and other over-3.5 tails |
Clean-sheet override
No forced clean-sheet branch, but the audit is explicit.
Neither team has a Spain-style non-concession streak. Clean-sheet scores are plausible, but they are excluded only after USA and Belgium scoring evidence attacks them numerically.
S10 candidate table
Every candidate gets one final status.
| score | poisson_probability | turtle_index | result_branch | spread_branch | total_bucket | btts_clean_bucket | recursive_attack | status |
|---|---|---|---|---|---|---|---|---|
| 2-1 | 0.088917 | 0.119948 | BEL | bel_by_1 | exactly3 | btts | rescued_by_Belgium_H2H_and_favorite_by_one | MAIN |
| 1-2 | 0.074896 | 0.087791 | USA | usa_by_1 | exactly3 | btts | rescued_by_USA_home_balogun_and_current_scoring | MAIN |
| 3-1 | 0.053757 | 0.083238 | BEL | bel_by_2plus | o3_5 | btts | no_extra_penalty | MAIN |
| 1-1 | 0.098048 | 0.082245 | DRAW | draw_low | exactly2 | btts | no_extra_penalty | SCORE_4 |
| 2-2 | 0.067921 | 0.070809 | DRAW | draw_high | o3_5 | btts | rescued_by_high_draw_current_tournament_and_usa_volatility | EXCLUDED_BY_TOTAL_THREAD |
| 3-2 | 0.041063 | 0.060391 | BEL | bel_by_1 | o3_5 | btts | no_extra_penalty | EXCLUDED_BY_TOTAL_THREAD |
| 2-0 | 0.058201 | 0.040568 | BEL | bel_by_2plus | exactly2 | bel_clean | attacked_by_USA_realm_and_WC_scoring | EXCLUDED_BY_BTTS_CLEAN_SHEET_THREAD |
| 1-0 | 0.064179 | 0.039900 | BEL | bel_by_1 | u1_5 | bel_clean | attacked_by_USA_realm_and_WC_scoring | EXCLUDED_BY_BTTS_CLEAN_SHEET_THREAD |
| 0-1 | 0.054059 | 0.025754 | USA | usa_by_1 | u1_5 | usa_clean | attacked_by_Belgium_realm_and_WC_scoring | EXCLUDED_BY_BTTS_CLEAN_SHEET_THREAD |
| 0-2 | 0.041294 | 0.020200 | USA | usa_by_2plus | exactly2 | usa_clean | attacked_by_Belgium_realm_and_WC_scoring | EXCLUDED_BY_BTTS_CLEAN_SHEET_THREAD |
Proof ledger
The theorem introduces no new judgment.
D1definition
Let G={0,...,7}x{0,...,7}. S10 is the ten highest count-kernel score cells P(a,b)=Pois(a;lambda_BEL)Pois(b;lambda_USA).
D2definition
A score survives only after Recursive Attack: the strongest opposing fact for its branch is named and compared numerically.
D3definition
Clean-Sheet Override is active only under an extreme verified non-concession streak or strong opponent failed-to-score signal; it is inactive here.
D4definition
Human Signal Preservation is recorded. No Neohm branch was named before this lock.
Data 1data
Threads: BEL=0.446, DRAW=0.226, USA=0.327, BTTS=0.655, O3.5=0.428.
R1rule
Select one Belgium-by-one representative, one USA-by-one representative, one Belgium-cushion/over-3.5 representative, then one low-draw containment representative if it survives the BTTS and low-total attacks.
L-2-1inclusion
2-1 is included because branch=bel_by_1, P_count=0.0889, I=0.1199, and its recursive attack is survived: rescued_by_Belgium_H2H_and_favorite_by_one.
L-1-2inclusion
1-2 is included because branch=usa_by_1, P_count=0.0749, I=0.0878, and its recursive attack is survived: rescued_by_USA_home_balogun_and_current_scoring.
L-3-1inclusion
3-1 is included because branch=bel_by_2plus, P_count=0.0538, I=0.0832, and its recursive attack is survived: no_extra_penalty.
L-1-1inclusion
1-1 is included because branch=draw_low, P_count=0.0980, I=0.0822, and its recursive attack is survived: no_extra_penalty.
E1exclusion
Clean-sheet scores 1-0, 2-0, 0-1, 0-2 are excluded because both sides clear the realm-scoring attack and no Clean-Sheet Override is active.
E2exclusion
2-2 and 3-2 are excluded because the over-3.5 slot is won by 3-1 under TurtleIndex comparison, while the draw slot is won by 1-1.
Theoremtheorem
Under D1-D4 and R1, the locked C4 is {2-1, 1-2, 3-1, 1-1}. If the regulation score is outside this set, the current Turtle protocol fails this practical test.
LOCKC4 = {2-1, 1-2, 3-1, 1-1}If outside C4, Turtle pauses for method rebuild.
Support markets after lock
Analysis-only; no allocation guidance.
| market | model_probability | analysis_status | protects_scores | hurts_scores | note |
|---|---|---|---|---|---|
| BTTS Yes | 0.6552 | full_card_aligned | 2-1, 1-2, 3-1, 1-1 | none inside C4 | analysis-only; no market price or sizing used before lock |
| Over 1.5 goals | 0.8463 | full_card_aligned | 2-1, 1-2, 3-1, 1-1 | none inside C4 | analysis-only; no market price or sizing used before lock |
| Under 3.5 goals | 0.5716 | partial_card_aligned_only | 2-1, 1-2, 1-1 | 3-1 | not full-card support because it kills the favorite-cushion branch |
| Over 2.5 goals | 0.6486 | partial_card_aligned_only | 2-1, 1-2, 3-1 | 1-1 | not full-card support because it kills the low draw |
Failure standard
What breaks this card is written before kickoff.
The written failure list missed the actual failure: Belgium cushion extended beyond 3-1 to 4-1. That is not a lucky exception; it is a missing ladder rule. The current protocol is paused for rebuild.
Sources