1. Expected goals and goal-rate inputs
Expected goals (xG) generally estimates the likelihood that shots become goals. Cateo's current fictional demonstration uses hypothetical scoring-rate inputs. Those values must not be confused with measured team xG or real match conditions.
2. Scoreline and total-goal probabilities
A simple Poisson model can turn assumed home and away goal rates into probabilities for individual scores. Summing these outcomes gives estimates for home/draw/away, over/under 2.5 goals, and both teams scoring (BTTS). These outputs depend strongly on the assumptions, which may be wrong.
3. Why two models can disagree
Two approaches may use different inputs, model families or calibration methods. Disagreement is a reason to inspect data quality and uncertainty, not evidence of a profitable selection. Cateo presents the two perspectives separately instead of averaging them into false precision.
4. The prospective evaluation plan
Before any real-data publication, models need licensed source inputs, time-stamped pre-kickoff forecasts, immutable data snapshots and verified outcomes. Planned measurements include Brier score, log loss, calibration and baseline comparison. Cateo has not published a verified prospective track record.
5. What the site does not claim
No fictional probability is a bookmaker quote, actual prediction, guaranteed result or demonstrated positive expected value. The site does not present historical returns or official soccer results unless they have passed independent approval and source-rights checks.
6. Data coverage
Cateo's research roadmap spans major football competitions, including Liga MX, MLS and European leagues and tournaments. Publishing specific matches or features depends on authorized data access, sufficient quality and evidence-backed model evaluation.