This is the complete pre-match research workflow used by serious football bettors on KiqIQ — from checking the predictions hub through to sizing your stake with Kelly. Every step uses a specific tool. We will walk through a real example throughout.
Running Example Throughout This Guide
We are analysing Brighton vs Arsenal — a Premier League fixture where Brighton are priced as home favourites at 2.30, with Arsenal at 2.80 and the draw at 3.40. We will work through every step to determine whether any outcome represents value.
Start at the Football Predictions Hub. This gives you a structured pre-match overview for any upcoming fixture — including the xG model output, form context, and key factors.
The predictions hub is not a tipster service — it is an analytical starting point. It shows you what the data suggests, not what to bet. Your job is to decide whether you agree with the model, and whether the odds reflect it.
In our example:
The predictions hub shows Brighton with a slight home advantage based on their xG record this season (1.6 xG per home game), while Arsenal have been creating well away (1.4 xG per game away) but conceding more than their season average would suggest (1.3 xGA per away game).
Before running any model, ask the AI assistant about the xG context for both teams. This is where you surface nuances the raw numbers do not show — form trends, injury impact on defensive shape, whether the xG is being driven by low-quality shots or genuine chances.
Example prompt to the AI:
"Brighton have averaged 1.6 xG at home this season and Arsenal 1.4 xG per game away. Brighton's last 5 home xG games have been 1.2, 1.8, 1.1, 2.1, 1.6. Is Brighton's home xG consistent or driven by one outlier fixture? And how does Arsenal's away xGA of 1.3 compare to their season-long average?"
The AI will break down whether the averages are stable or skewed, flag any recent trend shifts, and identify whether the match-up creates any structural xG advantages (e.g. Brighton's high press vs Arsenal's build-up pace).
What to look for:
With your xG inputs established, open the Poisson Distribution Calculator and enter the expected goal rates for both teams. The calculator outputs the probability of every scoreline — and from those, every market outcome.
Inputs for our example:
Brighton xG (home avg)
1.6
Arsenal xG (away avg)
1.4
Model output (approximate):
Brighton Win
38%
Draw
27%
Arsenal Win
35%
The calculator also outputs Over/Under and BTTS probabilities from the same model.
Note the total goals expectation: 1.6 + 1.4 = 3.0. This makes over 2.5 goals approximately 59% likely — relevant if you are considering goals markets.
Now convert the bookmaker's odds to fair implied probabilities using the Implied Probability Calculator. This strips the margin and shows you what probability the bookmaker is actually pricing in for each outcome.
| Outcome | Book Odds | Raw Implied % | Fair (no-vig) % | Model % | Edge |
|---|---|---|---|---|---|
| Brighton Win | 2.30 | 43.5% | 41.6% | 38% | -3.6% |
| Draw | 3.40 | 29.4% | 28.1% | 27% | -1.1% |
| Arsenal Win | 2.80 | 35.7% | 34.2% | 35% | +0.8% |
Reading the table: The bookmaker is pricing Brighton slightly above what our model suggests (41.6% fair vs 38% model — Brighton looks overpriced). Arsenal shows a marginal edge (+0.8%) but it is within noise range. This is a low-confidence market — no obvious +EV bet on the 1X2. We look at goals markets next.
With your Poisson output in hand, the AI assistant can go deeper on specific markets, contextual factors, and historical patterns that the model does not capture automatically.
Goals market value check
""My Poisson model gives over 2.5 goals a 59% probability for Brighton vs Arsenal. The bookmaker has over 2.5 at 1.75 (57.1% implied). Is this strong enough edge to bet, or should I look at the over 3.5 line instead?""
BTTS / timing pattern
""Brighton have scored in the first 30 minutes in 6 of their last 8 home games. Does this support a BTTS bet given Arsenal's away form?""
Qualitative context beyond the model
""Are there any tactical factors in Brighton vs Arsenal that could suppress or inflate the xG — pressing intensity, defensive line height, set-piece threats?""
The AI's response will synthesise the quantitative inputs with qualitative context — and often surface market angles you had not considered. In this case it might confirm the over 3.5 goals market at a higher price offers better value relative to model confidence.
Once you have identified a +EV bet, use the Kelly Criterion Calculator to determine the optimal stake size. Kelly takes your edge and the odds as inputs and outputs the mathematically optimal fraction of bankroll to risk.
Example: Over 2.5 goals at 1.75 odds, model probability 59%
Odds
1.75
Model probability
59%
EV
+3.3%
Kelly output: Full Kelly recommends ~8.7% of bankroll. Most practitioners use quarter-Kelly (2.2%) or half-Kelly (4.4%) to reduce variance while preserving most of the edge. At £500 bankroll, a quarter-Kelly stake is approximately £11.
After placing the bet, record it in the KiqIQ bet tracker with the odds you got. After kickoff, note the closing line (the final odds available at kickoff) and calculate your closing line value (CLV).
If you got 1.75 and the match closed at 1.68, your CLV is +0.07 — you beat the market's final assessment. Consistently positive CLV over 200+ bets is stronger evidence of skill than ROI in the same sample. It tells you your process is working before your results sample is large enough to confirm it.
Steps 1–2
Context
Predictions Hub + AI
Steps 3–4
Probability
Poisson + Implied P
Steps 5
Depth
AI Assistant
Steps 6–7
Execution
Kelly + Bet Tracker
The key inputs are: each team's average xG scored and conceded (home and away split), recent form over the last 5–8 games, and the current bookmaker odds. From these you can calculate a probability estimate, compare it to the bookmaker's implied probability, and identify whether a bet has positive expected value.
With a systematic workflow and the right tools, a thorough pre-match analysis takes 10–15 minutes per match. The KiqIQ platform consolidates the calculators, AI analysis, and probability tools you need in one place — cutting research time significantly.
Yes, the process works for any league where xG data is available. It is most effective in leagues with good public data coverage: Premier League, La Liga, Bundesliga, Serie A, Ligue 1, and the top Championship and Eredivisie leagues.
Every calculator in this workflow is free to use — no account required. Start with the Poisson calculator and see how your own probability estimate compares to the bookmaker's implied price.
