How to use Scoutelligence for first-half goals
Scoutelligence provides transparent statistical estimates for the chance of 2+ goals in the first half, based on historical first-half scoring rates, venue splits, recent form, and competition averages. Estimates are experimental and not betting advice.
Understanding the probability
A 55% model estimate means the Poisson model assigns a 55% chance of 2 or more first-half goals, given the available sample. Always check the sample sizes shown (recent matches, venue matches, league sample). Small samples are noisier.
- Strong candidate — typically higher probability with adequate samples.
- Candidate — lower or more uncertain estimate.
- Missing data is shown explicitly; we never treat missing stats as zero.
Worked example: reading an estimate
Imagine the model shows the following for a Premier League fixture:
Home vs Away · Premier League
Model probability of 2+ first-half goals: 52%
Sample: Home recent 8 / venue 6 · Away recent 7 / venue 5 · League n=120
Rank: Strong candidate
A typical bookmaker price for “Over 1.5 first-half goals” might be 1.90. The implied probability is:
1 / 1.90 ≈ 52.6%
In this case the model (52%) and the market (52.6%) are almost identical. There is no meaningful edge after accounting for bookmaker margin. You would treat this as informational only.
If the model had shown 62% against the same 1.90 price, the difference would be larger. Even then you would still check:
- Sample sizes are large enough that the estimate is not driven by a handful of matches.
- The calibration bands in the performance tracker show the model is roughly reliable in the 55–65% range.
- You have a staking plan and are not treating any single estimate as a guaranteed outcome.
The point of the tool is to surface the model number and the evidence behind it so you can compare it to the market yourself — not to tell you what to bet.
Comparing to the market
Where price data is available, compare the model probability to the implied probability from bookmaker odds (1 / decimal odds). An edge exists only if the model probability is meaningfully higher than the market-implied probability after accounting for margin. Always use a staking plan and never bet more than you can afford to lose.
Using the performance tracker
The prospective tracker records every eligible prediction at the time it is made and later settles it against actual half-time scores. Look at hit rate, Brier score, and calibration bands. A well-calibrated model has observed rates close to predicted rates in each band. The model remains experimental until more data accumulates.
Scout Plus
Scout Plus (£4.99/month) unlocks full-time BTTS analysis, player-level rankings for fouls, shots and shots on target, and deeper match context. Player Cards and some match stats remain free until 1 December 2026.
Statistics and model estimates are for information only. They are not guaranteed outcomes or betting advice. Gamble responsibly.