Rules

October 2026 experiment rules

• The starting bankroll is £100 and the initial unit is £10.
• Football and horse racing are the focus.
• Every decision is recorded and timestamped before its event. Every result, including losses, remains visible.
• Profits stay in the bankroll. There are no October top-ups.
• The experiment ends on 31 October 2026 or when the bankroll reaches £0, whichever comes first.
• Confirmed outcomes update the published bankroll, profit or loss, and return on settled stakes automatically. Corrections are appended to the audit trail.

What the figures mean

Profit or loss is the total of settled outcomes. ROI is profit or loss divided by total settled stakes. Pending decisions are excluded until settled. A void decision contributes neither profit nor settled stake.

This is a documented experiment, not a betting service or financial advice. Past results do not guarantee future outcomes.

Analysis and prediction methodology

The membership research programme begins with football and horse racing. Basketball and tennis are planned additions, not currently promised coverage.

1. Record the event, competition, start time and data sources. Display times in Europe/London, including daylight-saving changes.
2. Check data freshness, injuries, line-ups, withdrawals and event conditions.
3. Explain supporting and opposing evidence and the limits of the analysis.
4. Publish a selection only when the evidence supports it. No selection is a valid outcome.
5. Record any recommended market, available price, source and publication time before the event. Show subsequent changes as corrections.
6. Review the settled result and the reasoning, including losses.

Probability and confidence

Precise probabilities will be published only when supported by a tested model. Qualitative confidence is not a measured probability. Historical testing and simulations will be labelled and separated from live results.

Separate performance records

Personal experiments, assistant experiments and platform selections must be identified separately. Singles, Bet Builders and horse-racing markets should be reported separately where data permits. Deposits are contributions, never winnings.

Two return measures

Stake ROI: settled net profit divided by eligible settled stakes, multiplied by 100. Pending and void stakes are excluded.

Return on contributed capital: settled net profit divided by total contributed capital, multiplied by 100. This is a separate measure and must not be labelled stake ROI.

Win rate must state how voids and pushes are handled. A high win rate alone does not demonstrate profitability. Available cash and money committed to pending selections should be distinguished.

Publication standards

Keep wins and losses visible. Retain the October archive after the experiment ends. Do not describe past results as a guarantee of future returns. A timestamp on this website is not independent third-party proofing.