How SALMO Records Every Football Prediction Before Kickoff

The sports prediction industry suffers from a pervasive trust deficit. Social media channels and commercial tipping websites routinely delete losing picks, alter recommended odds after matches conclude, or create synthetic track records using hindsight data.

A prediction service that cannot mathematically prove when its predictions were generated is unworthy of serious capital.

SALMO was designed on a foundational principle: Zero Fabrication. Every recommendation must be publicly recorded and cryptographically frozen before a football match kicks off. Once recorded, records can never be edited, hidden, or deleted.

In SALMO, a prediction is not a casual text post; it is a structured, auditable data object that travels through an immutable seven-stage lifecycle:

Stage 1: Pre-Kickoff Ingestion & Feature Construction: Ingest official team, fixture, and historical data from canonical football APIs within a dynamic 7-day UTC rolling window.

Stage 2: Independent Model Probability Derivation: The quantitative engine computes P_model and fair odds for Asian Handicap, Over/Under, and BTTS.

Stage 3: Sharp Reference Odds Snapshot: Real-time benchmark quotes from Pinnacle and leading Asian bookmakers are ingested and timestamped.

Stage 4: Point-in-Time Database Ledger Entry: The prediction object—containing fixture UUID, selection, model probability, reference odds, net edge, EV, and UTC timestamp—is committed to our production ledger.

Stage 5: The Automated Kickoff Gate: At match kickoff, the record is locked permanently. The ingestion engine automatically rejects any prediction where timestamp_prediction ≥ timestamp_kickoff.

Stage 6: Automated Match Settlement: Following the final whistle, official match scores are fetched and processed through QuarterLineSettler.

Stage 7: Performance Metric Reconciliation: Profit/loss, ROI, and Yield are updated across the ledger with zero human intervention.

Hindsight bias is the enemy of quantitative credibility. If an algorithm adjusts its parameters or logs predictions after a match has started, historical backtests become meaningless.

The SALMO Kickoff Gate is an automated software guardrail. If an API provider experiences latency and delivers odds or match parameters after scheduled kickoff, the gate activates immediately. The fixture is flagged as KICKOFF_PASSED, and no prediction can be created or updated.

Every prediction on SALMO is guaranteed to represent a forward-looking forecast formulated strictly under pre-match uncertainty.

Unlike consumer apps that display simulated winning streaks or synthetic "90% win rate" banners, SALMO adheres to a strict Zero Fabrication Policy:

• We do not invent performance statistics.

• We do not claim an ROI, Yield, or win rate unless it exists in verified production database records.

• If a capability is pending or in progress (such as our automated closing-odds runner), we state "Currently unavailable" rather than fabricating placeholder numbers.

Transparent, unvarnished data is the only foundation upon which serious sports bettors can allocate capital.

When reviewing betting performance, terminology matters. Many websites conflate ROI and Yield to inflate their apparent returns.

• Yield (Turnover Efficiency): Defined as (Net Profit / Total Stakes Placed) * 100. In sports betting, Yield measures model quality. A sustained Yield of +4% to +7% over 500+ wagers in liquid Asian markets is considered institutional-grade.

• Return on Investment (ROI): Defined as (Net Profit / Starting Bankroll) * 100. ROI reflects bankroll growth and depends heavily on staking size and turnover velocity.

SALMO tracks both metrics rigorously, ensuring that all published figures reflect true mathematical definitions.