Home/Sports Betting/xG Interpretation

xG Interpretation

Model-Based BettingLevel 2 — Informed

Unlocks

Effect IdentificationHome Field Advantage MonitoringInjury & Lineup ImpactModel Building & CombinationModel Validation & OverrideStyle & Matchup AnalysisTeam Rating DriftVariance vs. Skill Decomposition

What It Is

Reading and applying expected goals (xG) data to evaluate team quality, identify over/underperformance, and make betting decisions based on process rather than results.

Correct Execution

You evaluate teams by xG differential (xGD) per game rather than actual goals or table position. You separate attack xG from defense xG. You recognize when results diverge from xG and understand what that means for future betting. You account for score effects, set pieces, and goalkeeper over/underperformance when interpreting xG.

Progression Levels

Diagnostic Tree

Coaching Cues

  • "First place with a +4 goal difference? Faaaaade." — when results far outstrip process, Ted Knutson
  • "A hilarious matchup between the great xG underperformers and the great xG overperformers." — regression collision, Ted Knutson
  • "My vibes bet did not pay off, but it was absolutely correct in the xG (a good indicator of whether a bet was likely correct or you got lucky)." — process over outcome, Ted Knutson

Common Errors

  1. Betting on table position instead of xGD: Table lies in small samples → "Wolves are still not as bad as their table position" → Always check xGD
  2. Ignoring score effects: Most shots came after going 2-0 up → inflated xG → Account for game state in xG interpretation
  3. Treating xG overperformance as skill: "12 goals from 5 xG" is not sustainable → Regression is real → Bet against inflated results
  4. Not separating home/away xG: Teams can have wildly different profiles by venue → Check both splits

Edges

🔑 Hidden Causal Lever

Score Effects Contaminate Your Data

model-bettingxg-interpretation →

Teams trailing early rack up flattering xG while chasing — shots accumulated after falling behind look impressive on paper but are low-quality, game-state-inflated numbers. If you don't strip out these score effects, you systematically overrate losing teams and underrate dominant ones who concede garbage-time chances.

What most people do
Look at raw xG totals and conclude a team "competed well" or is "better than results show."
What the best do
Discount xG accumulated after falling behind, recognizing that a team chasing a game gets flattering numbers that don't reflect true quality.
Why it's an edge: Most xG tables present aggregate numbers without game-state context. Bettors who adjust for score effects see a different and more accurate picture of team quality.
How to exploit: When evaluating a team, check the match timeline. If most of their xG came while trailing by 2+ goals, discount it heavily.
"Tons of score effects present in the Atletico Madrid game. Yes, Seattle put up 1.71 xG, but they were a goal down before the 10th, and ATM were never really out of control." — Ted Knutson, CWC Sunday and Monday 22 June 25
⚡ Conventional Wisdom Is Wrong

Defensive Overperformance Has a Historical Ceiling

model-bettingxg-interpretation →

Burnley's 18-goal defensive overperformance exceeds what Leicester achieved while winning the Premier League (~11 goals in StatsBomb data). When a non-elite team's overperformance exceeds what champions sustain, regression is near-certain. This turns "regression" from a vague concept into a quantified, actionable threshold.

What most people do
Either blindly trust overperformance ("they just defend well") or blindly bet against it ("regression is coming") with no framework for magnitude.
What the best do
Compare overperformance size to historical precedents. If it exceeds what elite champions achieved, the probability of regression is extremely high.
Why it's an edge: It gives you a concrete benchmark: "this level has basically never been sustained." Most bettors have no sense of scale for when overperformance is meaningful vs. unprecedented.
How to exploit: Track goals vs. xG allowed for every team. When defensive overperformance exceeds 10-12 goals, bet on regression aggressively.
"I honestly can't remember a defensive overperformance like this since Leicester won the Premier League and even that was only like 11 goals in StatsBomb's data. This is 18." — Ted Knutson, Good Friday Champ + EPL
🔑 Hidden Causal Lever

Goalkeeper PSxG+/- Separates Sustainable from Unsustainable Defense

model-bettingxg-interpretation →

A team that "looks terrible on xG but keeps winning" could be elite goalkeeping (partially sustainable over 1-2 seasons) or random finishing variance (fully regresses within weeks). PSxG+/- (Post-Shot Expected Goals plus/minus) isolates the goalkeeper's contribution. Without this decomposition, all defensive overperformance looks identical — but the sustainability profiles are completely different.

What most people do
Treat defensive overperformance as a single category. Either bet against it ("it will regress") or accept it ("they're just good defensively") without decomposing the mechanism.
What the best do
Decompose defensive overperformance into goalkeeper PSxG+/- (partially persistent — elite keepers sustain 0.05-0.10 xG saved per game over multiple seasons) and non-goalkeeper overperformance (random, reverts within 10-15 games). Only bet against the non-goalkeeper component.
Why it's an edge: The bettor who decomposes defensive overperformance correctly knows which teams will regress (non-GK driven) and which won't (GK driven), while the market treats them identically.
How to exploit: For any team overperforming on goals-conceded vs. xGA, check the goalkeeper's PSxG+/- on FBRef. If the GK is +3.0 or better over the season, a significant portion is sustainable. If the GK is near 0, the overperformance is random and will regress — bet against the team.
"Martinez being PSxG+/- of about positive 6. That's a significant chunk of the overperformance explained by an elite goalkeeper." — Ted Knutson, EPL analysis, 2025

Sources

  • Ted Knutson, multiple Variance Betting newsletters — xGD as primary evaluation metric
  • Ted Knutson, "English Champ Analysis 21st Sept Weekend" — score effects, contaminated metrics
  • Ted Knutson, "EPL and Champ Weekend 31Oct2025" — goalkeeper PSxG+/- impact