Most bettors remember their wins more vividly than their losses. It is human nature. The practical consequence is that without written records, almost everyone overestimates how well they are doing. Keeping records is not optional if you are serious about a data-led approach: it is the foundation of knowing whether you are actually making progress.
---
What to record for every bet
At minimum:
| Field | What to capture |
|---|---|
| Date | When the bet was placed |
| Fixture | Which match |
| Market | WDL, BTTS, Over 2.5, Player shots, etc. |
| Selection | What you backed |
| Decimal odds | The price you took |
| Stake | How much you bet |
| Signal rating | ★, ★★, ★★★ or ✕ if using BetSignals |
| Model probability | The BetSignals probability at time of placing |
| Result | Win or loss |
| Profit/loss | Actual £ return minus stake |
The more consistently you capture these, the more useful your records become.
---
Why the model probability column matters
Recording the BetSignals model probability alongside your outcome allows you to track model calibration from your own betting history.
Over time, you can check: do bets I backed at a 55% model probability actually win around 55% of the time? If they are winning 65% of the time, the model may be underestimating those outcomes. If they are winning 45% of the time, it may be overestimating.
This is how you build genuine evidence about whether your approach is working, not from memory, and not from a handful of results.
---
Calculating your key metrics
From your records, you can derive:
Return on Investment (ROI): total profit ÷ total staked × 100. The most important single number. Positive ROI over a large sample (100+ bets) is meaningful evidence of edge.
Strike rate: percentage of bets that won. Interesting but not definitive: a low strike rate at long odds is not worse than a high strike rate at short odds if the ROI is the same.
Average odds: total of all decimal odds ÷ number of bets. Combined with strike rate and ROI, this tells you whether you are generating returns from value at competitive odds or from volume at poor odds.
Performance by signal rating: are your ★★★ selections outperforming ★ selections? If not, it may mean you are not applying the ratings correctly, or the relationship between signal strength and returns in your specific betting mix is weaker than expected.
---
Tools for keeping records
A spreadsheet is sufficient and preferable to memory or a notes app. Set up columns as above and add a row after every bet is settled.
Key things to include as formula columns:
- Profit/loss: `=IF(result="W", stake*(odds-1), -stake)`
- Running total: cumulative sum of profit/loss column
- ROI: running total ÷ total staked
Review the spreadsheet monthly. Look for patterns: markets where you consistently underperform, fixture types where the model has historically been strong, and any drift in your ROI over time.
---
What records reveal about behaviour
Beyond raw performance, records reveal things about your decision-making that you would not otherwise notice:
Stake creep. If you gradually increase stakes after wins and decrease them after losses, that shows up in the data. Erratic staking undermines even a good selection record.
Selection creep. If you are recording more bets per week than you planned, that suggests you are finding reasons to bet rather than waiting for genuine signals.
Recency bias. If your records show clusters of losses followed by over-staking, you can identify and address the pattern.
The discipline of recording forces a degree of accountability that cannot be faked.
---
Starting from scratch
If you have not kept records before, start now. Do not try to reconstruct historical bets from memory, they will not be accurate, and the exercise will be frustrating. Start clean from your next bet.
Pick a simple spreadsheet format, stick to it, and review monthly. After one season of honest records, you will have more genuine information about your betting performance than most recreational bettors accumulate in years.
---
Next reads
- Bankroll Management for Football Bettors: how records feed into stake management
- Staking Strategies Explained: choosing a staking approach that records can validate
- Responsible Gambling: using records to spot when betting stops being enjoyable
---
18+ | Gambling should be enjoyable. If it stops being fun, take a break. For support and advice visit BeGambleAware.org. See our full responsible gambling guide.