Soccer Desk — Model Anchor Calibration Prompt

Context: Our MLB Player Props council (2026-04-03) established the universal framework: per-possession/per-PA rates as foundation, Bayesian shrinkage toward career prior, Log-Odds opponent adjustment, Gaussian copula for correlated combo props, dual-anchor system (sportsbook + model) tracked via Brier scores, confidence tiers. This prompt asks for the SPORT-SPECIFIC parameters to plug into that framework.

Soccer Desk — Model Anchor Calibration

We are building model anchors for soccer betting markets on Kalshi: 1X2 (3-way ML), Asian handicap/spreads, over/under totals, BTTS.

We already use a Dixon-Coles bivariate Poisson model. This prompt is about calibrating the MODEL ANCHOR to run alongside our sportsbook-anchored edges.

Questions:

1. What team metrics best predict goal output? (xG, xGA, non-penalty xG, shot quality?)
2. How many matches until team xG metrics stabilize? k values for Bayesian shrinkage?
3. How should the Dixon-Coles rho parameter be estimated per league? How many matches needed?
4. How should home advantage vary by league? (EPL ~0.38, Bundesliga ~0.43 — are these stable?)
5. How should squad rotation and fixture congestion adjust projections? (UCL midweek = rotation for weekend)
6. How should promoted/relegated teams be handled at season start? (No prior data in new league)
7. How to model BTTS: from Dixon-Coles scoreline matrix or separate model?
8. How should managerial changes affect team priors? Reset? Gradual adjustment?
9. Transfer window structural breaks: how to handle mid-season squad changes?
10. How should Elo ratings interact with xG-based projections? Blend or separate?
11. EWMA alpha for team attacking/defensive metrics?
12. League-specific parameters: avg goals, draw rate, BTTS rate — how many matches to stabilize?
13. Give specific numbers, formulas, and implementation-ready recommendations.