Eight correct. Two missed. Value still unproven.
This retrospective covers an archived private Top 10 research screen from 25 October, originally recorded at 15:59 Athens time. These ten markets were external probability scenarios, not an independently validated AIONORA PRO Final Call. No certified bets or accepted bookmaker tickets exist in this archive.
Every selection and its outcome
The original market, estimated probability and reference quote were preserved. Outcomes were compared with independent finished-match sources. Scores are historical; no after-the-fact changes were made to the forecasts.
| # | Event · audit explanation | Market | Prior P* | Odds** | Final | Audit |
|---|---|---|---|---|---|---|
| 1 | AZ Alkmaar – Twente Two late Twente goals (84′ and stoppage time) carried the Over 1.5. The outcome does not validate the pre-match 88% external probability. | Over 1.5 goals | 88% External | 1.16 Reference | 0–2 | WIN |
| 2 | Sporting CP – Académico Viseu Three total goals; the underdog won outright. The totals outcome should not be mistaken for support of the Sporting win market. | Over 1.5 goals | 87% External | 1.16 Reference | 1–2 | WIN |
| 3 | Rennes – Strasbourg Rennes scored all three. One-team scoring sufficed, so this says nothing about BTTS reliability. | Over 1.5 goals | 87% External | 1.15 Reference | 3–0 | WIN |
| 4 | Marseille – Le Havre Seven goals; the wide score range illustrates totals tail-risk. It is not proof of accurate calibration. | Over 1.5 goals | 86% External | 1.19 Reference | 4–3 | WIN |
| 5 | Union SG – Cercle Brugge Seven total goals, including two own goals and a red card. Unusual events must not be used as information allegedly known prematch. | Over 1.5 goals | 86% External | 1.16 Reference | 5–2 | WIN |
| 6 | Real Sociedad – Levante Three goals cleared the threshold, but independent model stress and available odds were not sufficient to certify the forecast. | Over 1.5 goals | 83% External | 1.21 Reference | 2–1 | WIN |
| 7 | Volos – Panathinaikos Only one goal: the 83% external estimate overpredicted this match. Analyze one-team-zero risk, lineup and low-goal tails. | Over 1.5 goals | 83% External | 1.23 Reference | 0–1 | LOSS |
| 8 | Toulouse – Troyes Goalless match: the 82% external estimate missed the draw and zero-zero mass. Test this systematically, not by hindsight retuning. | Over 1.5 goals | 82% External | 1.22 Reference | 0–0 | LOSS |
| 9 | Auxerre – Le Mans Two goals just crossed Over 1.5; the hypothetical win margin does not reveal the model’s probability quality. | Over 1.5 goals | 82% External | 1.25 Reference | 1–1 | WIN |
| 10 | Göztepe – Çorum Both goals came from Göztepe. Over 1.5 passed; this would not have been a successful BTTS pick. | Over 1.5 goals | 82% External | 1.23 Reference | 2–0 | WIN |
*External research probability, not independent AIONORA p_model. **Site reference odds, not a verified bookmaker price executable in Greece. There is no defensible individual 0–100 Validation Score for these archived picks, because the six mandatory component inputs were not recorded. All ten: NOT CERTIFIED.
Learning, not hindsight optimisation
Toulouse–Troyes · 0–0
Loss: the 82% Over 1.5 external estimate failed. The model screen missed the zero-zero outcome. Study lower-tail and low-event-rate calibration in a pre-registered holdout.
Volos–Panathinaikos · 0–1
Loss: the 83% Over 1.5 scenario failed as only one side scored once. Check opponent-adjusted chance creation and player availability, using data known before kick-off.
AZ–Twente · 0–2
Win, but late: goals at 84′ and stoppage time. A correct total is not evidence that the 88% pre-match forecast was calibrated. Capture time-of-first-goal sensitivity.
The sample consists of ten bets from a single market family: goal totals Over 1.5. Ten correlated modelling assumptions are not ten independent proofs of forecasting quality. No model weights have been changed from these events.
Ex-post success is not pre-match certification
The archived external research optimizer identified no 2- or 3-leg set meeting the 2.00 reference-odds requirement.
Two archived reference combinations would have passed on actual outcomes. Their original independent-minimum EV scenarios were −36.0% and −39.8%, respectively. Neither was certified or placed.
The 4-leg reference combination was 2.00158; the 5-leg reference combination 2.136086. These are old site reference products, not executable prices or certified probabilities. No retroactive winning slip, return or official track-record entry was created.
A hit rate alone cannot establish value
Worksheet from ten externally reported probabilities. NOT AIONORA-certified calibration.
Small-sample descriptive measure; no OOS model passport.
Not an actual monetary result, CLV or verified executable-price EV.
Certification stays on hold until the independent market-specific model, component-level Validation Score, confirmed player status, Greece-executable pricing and positive conservative EV are all documented. The authenticated PRO Final Call freeze and full NO-BET ledger also remain to be reconciled.
Auditable inputs, explicit limits
Original research file: AIONORA_PRIVATE_TOP10_80PLUS_RESEARCH_2026-10-25.json, recorded 25 October at 15:59 Greece time. The source was owner-only, and is not served for download on this page.
Settlement evidence: the result source link in each row opens an independently published fixture score. Archive source SHA-256 is stored in the internal audit receipt; published scores were checked independently on 26 October.
Limits: no authenticated AIONORA PRO Final Call publication receipt, no prospective approved p_model, no current Greece-executable bet price and no real stakes. Original Morning CORE/PRO/NO-BET forecasts were not modified. Future daily PRO Audit updates require authenticated frozen source snapshots plus result readback.
Independent informational analysis only · 21+ · Responsible Use · Privacy.