---
title: Add a slate-wide scoring factor to the simulation
type: task
status: cancelled
description: >-
  Add a small shared weekly factor across all NFL games in a slate, pursued only if the calibration
  backtest shows team scores co-moving more than simulated.
base_uri: user:task/league/add-slate-wide-scoring-factor-to-simulation.md
created_at: '2026-09-27T00:27:29.348Z'
entity_id: 2c098d7c-11be-4c7a-a1d4-7392a9a6aba7
finished_at: '2026-09-29T00:14:21.141Z'
observations:
  - ████████████████████████████████████████████████
owner_identity_uri: █████████████████████████
priority: Low
public_read: true
relations:
  - blocked_by [[█████████████████████████-██████-██████████-██████████████]]
tags:
  - user:tag/league-xo-football.md
updated_at: '2026-09-29T00:14:21.141Z'
visibility_analyzed_at: '2026-11-05T23:48:18.788Z'
---

# Add a slate-wide scoring factor to the simulation

## Problem

Each NFL game is simulated with its own matrix and seed (`libs-server/simulation/simulate-league-week.mjs`, the per-game loop), so games in the same week are fully independent. A league-wide scoring environment (weather week, officiating trend) cannot lift or depress every team together.

## Intent

- Add a small shared weekly factor across all games in the slate.
- Only pursue if the calibration backtest shows realized team scores co-moving more than the simulation predicts; the expected effect is small.

Gate: the week co-movement z-score variance in [[user:text/league/simulation-calibration-baseline.md]]. The 2026-09-27 baseline reads 1.12 over 79 weeks (2020-2022 1.33, 2023-2026 0.84), close to the 1 a captured co-movement gives, so there is no current evidence for this factor.