RecordEdge™

RecordEdge™ answers a fundamental question: when teams with similar records face off, who usually wins?

While GameStateEdge™ focuses on in-game situations, RecordEdge™ provides context about team strength based on their season performance entering the matchup.

What It Measures

RecordEdge™ calculates the historical win probability for the home team based on:

1. Home Team Record - All wins, losses, and ties of the home team entering the game
2. Away Team Record - All wins, losses, and ties of the away team entering the game

How It Works

When two teams with specific records face off, we look up all historical matchups where:

• The home team had the same (or similar) record
• The away team had the same (or similar) record

For example, if a 10-3 home team faces a 6-7 away team, we find all historical games where:

• Home team was 10-3 (±1 win/loss/tie)
• Away team was 6-7 (±1 win/loss/tie)

The Calculation

Base Query

For exact record matches:

$P_h = \frac{\text{home team wins in matching games}}{\text{total matching games}}$

Perturbation for Sample Size

To increase sample size, we also include ±1 record perturbations that maintain the same total games played:

Original Record	Included Perturbations
8-4-0	9-3-0, 7-5-0
6-6-0	7-5-0, 5-7-0
10-2-1	11-1-1, 9-3-1, 10-1-2, 10-3-0, etc.

This ensures we have enough historical data while keeping matchups comparable.

Why Records Matter

Team records are a proxy for team strength. A 12-1 team is typically better than a 4-9 team, and historical data reflects this:

Home Record	Away Record	Historical Home Win %
10-3	3-10	~81%
7-6	7-6	~50% (home advantage perhaps not statistically significant by some matchups, leading to a pure 50/50 toss up)
3-10	10-3	~21%

Data Sources

RecordEdge™ uses the same nflverse play-by-play data as GameStateEdge™, but joins it with weekly team record data. Records are calculated as of the week the game was played, not end-of-season records.

You can filter by time period:

• All Seasons (1999-present): Maximum historical context
• Last Season (2024): Recent competitive landscape
• This Season (2025): Current season trends

Interpreting the Values

RecordEdge™	Market Price	Interpretation
72%	60%	Market underpricing strong favorite
40%	55%	Market overpricing home team
55%	55%	Market fairly pricing the matchup

Edge Strength Guide

Edge Size	Signal Strength
±0-5%	Normal variance
±5-10%	Worth noting
±10%+	Significant divergence

Home Team Perspective

RecordEdge™ is always calculated from the home team's perspective. This is because:

1. Kalshi markets: We track home team contracts to avoid duplication
2. Home advantage: Historical data accounts for venue
3. Consistency: All edges use the same perspective

When viewing the away team's contract, we simply calculate 100% - RecordEdge™ to get their probability.

Limitations

• Record ≠ Strength: A 7-6 team that lost close games may be better than their record suggests
• Schedule strength: Doesn't account for opponent difficulty
• Injuries/changes: Mid-season roster changes aren't reflected in the record
• Small buckets: Unusual record pairings (e.g., 0-13 vs 13-0) have limited historical data

Combining with GameStateEdge™

The dashboard shows both edges for a reason. Each captures different information:

Metric	Best For
GameStateEdge™	Live game situations, momentum shifts
RecordEdge™	Pre-game analysis, team strength context

The Combined Edge on the dashboard averages both when available, giving you a balanced view of market mispricing.

---

Back to all edges | View GameStateEdge™