Soccer predictions: actual data vs predicted data
Slate 20260913 · 6 match(es)
How to read this report. Each match shows the model's headline prediction, the stage-by-stage lambda (goal-rate) build, both teams' standing/form inputs, and every starter's projected line next to their real season baseline. A moderate / large deviation badge flags a projection that strays far from the player's season form — a quick way to spot when the inputs look right but a projection does not. When the real result is known an actual-vs-predicted banner appears.
All matches overview
| Match | Score | Winner | H / D / A | O2.5 | BTTS | Conf | λ H/A | Actual | Result |
|---|
| Real Sociedad vs Atl. Madrid LaLiga | 0-1 | Atl. Madrid | 24.2% / 25.2% / 50.6% | 48.5% | 51.0% | 0.67 | 1.02 / 1.59 | - | pending |
| Levante vs Barcelona LaLiga | 1-3 | Barcelona | 7.8% / 9.9% / 82.3% | 88.3% | 70.4% | 0.77 | 1.27 / 3.84 | - | pending |
| Coventry vs Brighton Premier League | 0-1 | Brighton | 23.9% / 24.6% / 51.5% | 50.8% | 52.7% | 0.66 | 1.05 / 1.66 | - | pending |
| Getafe vs Deportivo LaLiga | 0-1 | Deportivo | 32.3% / 29.7% / 38.0% | 36.5% | 43.5% | 0.63 | 1.02 / 1.13 | - | pending |
| Celta vs Málaga LaLiga | 0-1 | Málaga | 28.4% / 32.5% / 39.1% | 28.1% | 36.0% | 0.64 | 0.82 / 1.02 | - | pending |
| Man Utd vs Man City Premier League | 1-2 | Man City | 22.8% / 19.1% / 58.1% | 74.9% | 70.6% | 0.68 | 1.49 / 2.42 | - | pending |
Match detail
Real Sociedad vs Atl. Madrid
LaLiga · 20260913 · 4-4-2 vs 4-4-2
Predicted 0-1Winner: Atl. MadridConfidence 0.67λ home/away 1.02 / 1.59Most-likely P 11.6%
Actual vs predicted: result not available yet (upcoming fixture).
XGBoost Reasoning & FBref Dataset TracingAntagonist + XGBoost match analysis for Real Sociedad vs Atl. Madrid. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 4-4-2 vs 4-4-2 (duel lambdas: Home 0.943, Away 1.441). Final calibrated Poisson lambdas: Home 1.020, Away 1.593 (most likely score 0-1).
Home Starters 11/11 XGBoost (9 FBref) Away Starters 11/11 XGBoost (11 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 0.94 / 1.44 Final λ (H/A) 1.02 / 1.59
FBref Team Priors: Real Sociedad (Gls/90: 0.75, GA/90: 0.75 → E[λ]: 1.294) vs Atlético Madrid (Gls/90: 2.333333333333333, GA/90: 2.33 → E[λ]: 1.296) | Blend Weight: 0.228
Modeling Pipeline Stages (5 layers)
- 1. XGBoost player regressions for volume & event rates with market-value quality scaling
- 2. Multi-source prior blending (FBref 2026-2027 tables + Sofascore season history)
- 3. Antagonist formation duel matrix resolving 1v1 and zonal matchups
- 4. Team-level prior calibration (FBref possession and attack/defense efficiencies)
- 5. Bivariate Poisson matrix with Dixon-Coles low-score correction
Standing & form context
| Metric | Real Sociedad | Atl. Madrid |
|---|
| League position | 9.00 | 6.00 |
| Points per game | 1.40 | 1.75 |
| Attack index | 1.25 | 1.46 |
| Defense (concede) index | 1.25 | 1.67 |
| Form multiplier | 0.98 | 1.02 |
| Standing λ multiplier | 1.20 | 1.45 |
Lambda decomposition (how the goal rates were built)
| Stage | Real Sociedad λ | Atl. Madrid λ |
|---|
| FBref Squad Prior E[λ] (weight 0.228) | 1.294 | 1.296 |
| Pre-mismatch | 0.943 | 1.441 |
| Pre-competitive-floor | 0.915 | 1.593 |
| Pre-calibration (Micro-duels) | 0.943 | 1.441 |
| Final λ | 1.020 | 1.593 |
competition calibration scale 1.0competition observed total goals 0.0quality gap applied Truemismatch score pressure applied Truecompetitive underdog floor applied Truestage discount 1.0
Scoreline distribution & markets
| Top 5 scorelines | P |
|---|
| 0-1 | 11.6% |
| 1-1 | 12.0% |
| 1-2 | 9.5% |
| 0-2 | 9.3% |
| 0-0 | 7.4% |
| Home win | 24.2% |
| Draw | 25.2% |
| Away win | 50.6% |
| Over 2.5 | 48.5% |
| BTTS yes | 51.0% |
Pregame fair line — favorite Atl. Madrid (50.6%), fair total 2.61, fair spread -0.57
Projected players vs season baseline
Proj = projected for this match · Season = this season's per-game average from match data · deviation badge compares projected xG to the player's season xG/game.
Real Sociedad — projected line vs season baseline (4-4-2)
| Player | Pos | Min | Proj G | Proj xG | Season xG/g | | Proj Sh | Season Sh/g | Proj Rtg | Season Rtg | Rec | Model / FBref Prior |
|---|
| Álex Remiro | G | 90 | 0 | 0.00 | - | n/a | 0 | 0.0 | 7.1 | 52.2 | 6 | FBref Prior G_service |
| Jon Aramburu | D | 85 | 0 | 0.00 | 0.01 | 100% | 1 | 0.7 | 6.7 | 55.0 | 6 | FBref Prior D_service |
| Luken Beitia | D | 88 | 0 | 0.00 | - | n/a | 0 | 0.0 | 6.6 | 11.4 | 5 | FBref Prior D_service |
| Mamadou Sarr | D | 61 | 0 | 0.00 | - | n/a | 0 | 0.0 | 6.9 | 17.8 | 4 | FBref Prior D_service |
| Sergio Gómez | D | 88 | 0 | 0.00 | 0.02 | 100% | 1 | 0.5 | 6.7 | 56.8 | 6 | FBref Prior D_service |
| Mikel Oyarzabal | F | 66 | 0 | 0.20 | 0.08 | 150% | 2 | 1.2 | 7.1 | 55.7 | 6 | FBref Prior F_service |
| Orri Óskarsson | F | 71 | 0 | 0.31 | 0.36 | ok | | 1.5 | 7.3 | 45.2 | 6 | XGBoost F_service |
| Ander Barrenetxea | M | 66 | 0 | 0.13 | 0.04 | 231% | 2 | 0.7 | 7.0 | 45.0 | 6 | FBref Prior M_service |
| Beñat Turrientes | M | 65 | 0 | 0.07 | 0.02 | 322% | 1 | 0.7 | 7.0 | 46.0 | 6 | FBref Prior M_service |
| Yangel Herrera | M | 59 | 0 | 0.06 | 0.02 | 141% | 1 | 0.7 | 6.8 | 57.2 | 6 | FBref Prior M_service |
| Job Ochieng | M | 77 | 0 | 0.13 | 0.05 | 191% | 1 | 0.8 | 6.8 | 24.2 | 6 | XGBoost M_service |
| Team totals | 0 | 1.02 | | 10 | | 6.9 | | |
Bench Candidates (7 modeled substitutes with empirical playing time)
| Player | Pos | Exp Min | App Prob | Substitute Modeling Details |
|---|
| Unai Marrero | G | 4.4m | 10.9% | Market-value substitute prior |
| Jon Martin | D | 14.5m | 23.6% | Market-value substitute prior |
| Hector Fort | D | 14.0m | 31.7% | Market-value substitute prior |
| Jon Gorrotxategi | M | 19.1m | 71.3% | Market-value substitute prior |
| Takefusa Kubo | M | 17.4m | 42.0% | Market-value substitute prior |
| Gorka Carrera Zarranz | F | 25.3m | 56.1% | Market-value substitute prior |
Atl. Madrid — projected line vs season baseline (4-4-2)
| Player | Pos | Min | Proj G | Proj xG | Season xG/g | | Proj Sh | Season Sh/g | Proj Rtg | Season Rtg | Rec | Model / FBref Prior |
|---|
| Alex Baena | M | 42 | 0 | 0.15 | 0.10 | 60% | 2 | 1.4 | 7.6 | 59.2 | 5 | FBref Prior M_service |
| Julián Alvarez | F | 31 | 0 | 0.17 | 0.09 | 82% | 2 | 0.7 | 7.7 | 43.7 | 3 | FBref Prior F_service |
| Ademola Lookman | F | 92 | 1 | 0.51 | 0.09 | 497% | 2 | 1.4 | 7.4 | 54.8 | 5 | FBref Prior F_service |
| Morten Hjulmand | M | 67 | 0 | 0.07 | 0.00 | 1736% | 2 | 0.2 | 7.6 | 55.4 | 5 | FBref Prior M_service |
| Koke | M | 48 | 0 | 0.05 | 0.01 | 613% | 2 | 0.2 | 6.8 | 57.0 | 5 | FBref Prior M_service |
| Giuliano Simeone | M | 87 | 0 | 0.12 | 0.06 | 94% | 2 | 1.2 | 7.4 | 52.6 | 5 | FBref Prior M_service |
| Dávid Hancko | D | 95 | 0 | 0.00 | 0.01 | 100% | 1 | 0.6 | 7.4 | 58.8 | 5 | FBref Prior D_service |
| Cristian Romero | D | 34 | 0 | 0.27 | 0.06 | 357% | 1 | 0.3 | 7.3 | 22.3 | 3 | FBref Prior D_service |
| Marc Pubill | D | 88 | 0 | 0.00 | 0.03 | 100% | 1 | 0.4 | 7.2 | 42.0 | 5 | FBref Prior D_service |
| Marcos Llorente | D | 81 | 0 | 0.00 | 0.02 | 100% | 1 | 0.6 | 7.0 | 55.2 | 5 | FBref Prior D_service |
| Jan Oblak | G | 92 | 0 | 0.00 | - | n/a | 0 | 0.0 | 7.1 | 54.4 | 5 | FBref Prior G_service |
| Team totals | 1 | 1.59 | | 18 | | 7.3 | | |
Bench Candidates (7 modeled substitutes with empirical playing time)
| Player | Pos | Exp Min | App Prob | Substitute Modeling Details |
|---|
| Juan Musso | G | 4.4m | 10.9% | Market-value substitute prior |
| Robin Le Normand | D | 14.3m | 23.2% | Market-value substitute prior |
| José María Giménez | D | 16.7m | 60.7% | Market-value substitute prior |
| Pablo Barrios | M | 24.4m | 46.9% | Market-value substitute prior |
| Kang-in Lee | M | 23.1m | 54.9% | Market-value substitute prior |
| Jonathan David | F | 26.4m | 60.8% | Market-value substitute prior |
Antagonist / tactical summary
Away midfield dictates play with ~52% possession expected (Ander Barrenetxea vs Morten Hjulmand). Total antagonist xG — Home: 0.33, Away: 0.36. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Left Mid vs Right Back — Wide Channel.
Midfield — possession 47.6% / 52.4%, dominance ratio 0.42, key duel: Ander Barrenetxea vs Morten Hjulmand
Mikel Oyarzabal xG 0.16Beñat Turrientes xG 0.05Job Ochieng xG 0.09Julián Alvarez xG 0.12Alex Baena xG 0.14Koke xG 0.04
Summary
Antagonist matchup analysis for Real Sociedad vs Atl. Madrid. Most likely scoreline: 0-1. Lambdas derived from 4 home and 4 away individual player duels.
Tactical parity between 4-4-2 and 4-4-2. Match outcome driven primarily by individual player quality. Away midfield dictates play with ~52% possession expected (Ander Barrenetxea vs Morten Hjulmand). Total antagonist xG — Home: 0.33, Away: 0.36. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Left Mid vs Right Back — Wide Channel.
Levante vs Barcelona
LaLiga · 20260913 · 4-1-4-1 vs 4-2-3-1
Predicted 1-3Winner: BarcelonaConfidence 0.77λ home/away 1.27 / 3.84Most-likely P 7.3%
Actual vs predicted: result not available yet (upcoming fixture).
XGBoost Reasoning & FBref Dataset TracingAntagonist + XGBoost match analysis for Levante vs Barcelona. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 4-1-4-1 vs 4-2-3-1 (duel lambdas: Home 1.353, Away 3.248). Final calibrated Poisson lambdas: Home 1.271, Away 3.838 (most likely score 1-3).
Home Starters 11/11 XGBoost (8 FBref) Away Starters 10/11 XGBoost (11 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 1.35 / 3.25 Final λ (H/A) 1.27 / 3.84
FBref Team Priors: Levante UD (Gls/90: 1.6666666666666667, GA/90: 1.67 → E[λ]: 3.5) vs FC Barcelona (Gls/90: 3.6666666666666665, GA/90: 4.0 → E[λ]: 3.5) | Blend Weight: 0.228
Modeling Pipeline Stages (5 layers)
- 1. XGBoost player regressions for volume & event rates with market-value quality scaling
- 2. Multi-source prior blending (FBref 2026-2027 tables + Sofascore season history)
- 3. Antagonist formation duel matrix resolving 1v1 and zonal matchups
- 4. Team-level prior calibration (FBref possession and attack/defense efficiencies)
- 5. Bivariate Poisson matrix with Dixon-Coles low-score correction
Standing & form context
| Metric | Levante | Barcelona |
|---|
| League position | 11.00 | 1.00 |
| Points per game | 1.25 | 3.00 |
| Attack index | 1.04 | 3.54 |
| Defense (concede) index | 0.42 | 1.04 |
| Form multiplier | 0.94 | 1.06 |
| Standing λ multiplier | 0.72 | 1.65 |
Lambda decomposition (how the goal rates were built)
| Stage | Levante λ | Barcelona λ |
|---|
| FBref Squad Prior E[λ] (weight 0.228) | 3.500 | 3.500 |
| Pre-mismatch | 1.353 | 3.248 |
| Pre-calibration (Micro-duels) | 1.353 | 3.248 |
| Final λ | 1.271 | 3.838 |
competition calibration scale 1.0competition observed total goals 0.0quality gap applied Truemismatch score pressure applied Truecompetitive underdog floor applied Falsestage discount 1.0
Scoreline distribution & markets
| Top 5 scorelines | P |
|---|
| 1-3 | 7.3% |
| 1-4 | 7.0% |
| 0-3 | 5.7% |
| 1-2 | 5.7% |
| 0-4 | 5.5% |
| Home win | 7.8% |
| Draw | 9.9% |
| Away win | 82.3% |
| Over 2.5 | 88.3% |
| BTTS yes | 70.4% |
Pregame fair line — favorite Barcelona (82.3%), fair total 5.11, fair spread -2.57
Projected players vs season baseline
Proj = projected for this match · Season = this season's per-game average from match data · deviation badge compares projected xG to the player's season xG/game.
Levante — projected line vs season baseline (4-1-4-1)
| Player | Pos | Min | Proj G | Proj xG | Season xG/g | | Proj Sh | Season Sh/g | Proj Rtg | Season Rtg | Rec | Model / FBref Prior |
|---|
| Iván Romero | F | 71 | 1 | 0.22 | 0.33 | 32% | 2 | 2.2 | 6.8 | 51.4 | 5 | FBref Prior F_service |
| Thiago Fernández | M | 75 | 0 | 0.23 | 0.11 | 101% | 1 | 1.2 | 6.8 | 51.5 | 4 | FBref Prior M_service |
| Jon Ander Olasagasti | M | 87 | 0 | 0.23 | 0.12 | 89% | 2 | 1.0 | 6.8 | 57.8 | 5 | XGBoost M_service |
| Enzo Bardeli | M | 69 | 0 | 0.07 | 0.01 | 487% | 1 | 0.4 | 6.8 | 54.0 | 5 | FBref Prior M_service |
| Roger Brugué | M | 86 | 0 | 0.14 | 0.50 | 73% | | 2.0 | 6.7 | 52.0 | 4 | XGBoost M_service |
| Oriol Rey | M | 88 | 0 | 0.06 | 0.00 | 1603% | 1 | 0.2 | 6.7 | 53.6 | 5 | FBref Prior M_service |
| Manuel Sánchez | D | 89 | 0 | 0.00 | 0.00 | 100% | 0 | 0.2 | 6.6 | 54.8 | 5 | XGBoost D_service |
| Aïssa Mandi | D | 88 | 0 | 0.00 | 0.01 | 100% | 0 | 0.2 | 6.7 | 56.8 | 5 | FBref Prior D_service |
| Adrián de la Fuente | D | 89 | 0 | 0.17 | 0.09 | 96% | 1 | 1.2 | 6.7 | 59.2 | 5 | FBref Prior D_service |
| Nacho Pérez | D | 68 | 0 | 0.00 | 0.02 | 100% | 0 | 0.2 | 6.6 | 53.8 | 5 | FBref Prior D_service |
| Mathew Ryan | G | 88 | 0 | 0.00 | - | n/a | 0 | 0.0 | 7.2 | 61.8 | 5 | FBref Prior G_service |
| Team totals | 1 | 1.27 | | 9 | | 6.8 | | |
Bench Candidates (7 modeled substitutes with empirical playing time)
| Player | Pos | Exp Min | App Prob | Substitute Modeling Details |
|---|
| Pablo Cunat Campos | G | 4.4m | 10.9% | Market-value substitute prior |
| Axel Tape | D | 13.8m | 31.1% | Market-value substitute prior |
| Ifeanyi Ndukwe | D | 13.6m | 29.9% | Market-value substitute prior |
| Carlos Álvarez | M | 22.6m | 31.9% | Market-value substitute prior |
| Hugo Sotelo | M | 22.1m | 50.3% | Market-value substitute prior |
| Etta Eyong | F | 25.9m | 58.5% | Market-value substitute prior |
Barcelona — projected line vs season baseline (4-2-3-1)
| Player | Pos | Min | Proj G | Proj xG | Season xG/g | | Proj Sh | Season Sh/g | Proj Rtg | Season Rtg | Rec | Model / FBref Prior |
|---|
| Marc Bernal | M | 62 | 0 | 0.13 | 0.13 | ok | 1 | 0.8 | 7.4 | 70.8 | 4 | FBref Prior M_service |
| Pedri | M | 81 | 0 | 0.42 | 0.09 | 346% | 2 | 1.0 | 8.0 | 74.0 | 4 | FBref Prior M_service |
| Xavi Espart | D | 89 | 0 | 0.00 | 0.02 | 100% | 1 | 0.8 | 6.6 | 70.2 | 4 | FBref Prior D_service |
| Gerard Martín | D | 95 | 0 | 0.00 | 0.03 | 100% | 0 | 0.2 | 7.2 | 55.2 | 4 | FBref Prior D_service |
| Andreas Christensen | D | 69 | 0 | 0.00 | 0.05 | 100% | 0 | 0.2 | 6.7 | 52.8 | 4 | FBref Prior D_service |
| João Cancelo | D | 11 | 0 | 0.00 | 0.01 | 100% | 0 | 0.2 | 6.7 | 49.8 | 4 | FBref Prior D_service |
| Joan García | G | 95 | 0 | 0.00 | - | n/a | 0 | 0.0 | 7.7 | 72.5 | 4 | FBref Prior G_service |
| Lamine Yamal | M | 87 | 1 | 0.77 | 0.64 | ok | 2 | 5.8 | 7.8 | 83.0 | 4 | FBref Prior M_service |
| Fermín López | M | 79 | 0 | 0.72 | 0.68 | ok | 3 | 2.5 | 7.8 | 80.8 | 4 | FBref Prior M_service |
| Raphinha | M | 90 | 0 | 0.49 | 1.08 | 55% | 2 | 2.6 | 7.5 | 68.6 | 5 | FBref Prior M_service |
| Anthony Gordon | F | 62 | 2 | 1.08 | 0.30 | 258% | 3 | 2.2 | 7.6 | 71.8 | 4 | FBref Prior F_service |
| Team totals | 3 | 3.84 | | 14 | | 7.4 | | |
Bench Candidates (7 modeled substitutes with empirical playing time)
| Player | Pos | Exp Min | App Prob | Substitute Modeling Details |
|---|
| Dominik Livaković | G | 4.5m | 11.1% | Market-value substitute prior |
| Pau Cubarsí | D | 15.4m | 38.0% | Market-value substitute prior |
| Jules Koundé | D | 15.2m | 37.5% | Market-value substitute prior |
| Dani Olmo | M | 23.6m | 57.3% | Market-value substitute prior |
| Rodri | M | 23.5m | 56.8% | Market-value substitute prior |
| Karim Adeyemi | F | 26.7m | 62.1% | Market-value substitute prior |
Antagonist / tactical summary
Away midfield dictates play with ~54% possession expected (Enzo Bardeli vs Lamine Yamal). Total antagonist xG — Home: 0.50, Away: 1.07. Home danger zone: Right Mid vs Left Back — Wide Channel. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.
Midfield — possession 45.6% / 54.4%, dominance ratio 0.35, key duel: Enzo Bardeli vs Lamine Yamal
Iván Romero xG 0.14Thiago Fernández xG 0.14Jon Ander Olasagasti xG 0.15Pedri xG 0.16Fermín López xG 0.29Lamine Yamal xG 0.31
Summary
Antagonist matchup analysis for Levante vs Barcelona. Most likely scoreline: 1-3. Lambdas derived from 5 home and 4 away individual player duels.
Tactical parity between 4-1-4-1 and 4-2-3-1. Match outcome driven primarily by individual player quality. Away midfield dictates play with ~54% possession expected (Enzo Bardeli vs Lamine Yamal). Total antagonist xG — Home: 0.50, Away: 1.07. Home danger zone: Right Mid vs Left Back — Wide Channel. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.
Coventry vs Brighton
Premier League · 20260913 · 3-4-3 vs 4-2-3-1
Predicted 0-1Winner: BrightonConfidence 0.66λ home/away 1.05 / 1.66Most-likely P 11.0%
Actual vs predicted: result not available yet (upcoming fixture).
XGBoost Reasoning & FBref Dataset TracingAntagonist + XGBoost match analysis for Coventry vs Brighton. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 3-4-3 vs 4-2-3-1 (duel lambdas: Home 0.892, Away 1.657). Final calibrated Poisson lambdas: Home 1.050, Away 1.657 (most likely score 0-1).
Home Starters 11/11 XGBoost (9 FBref) Away Starters 10/11 XGBoost (10 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 0.89 / 1.66 Final λ (H/A) 1.05 / 1.66
Modeling Pipeline Stages (5 layers)
- 1. XGBoost player regressions for volume & event rates with market-value quality scaling
- 2. Multi-source prior blending (FBref 2026-2027 tables + Sofascore season history)
- 3. Antagonist formation duel matrix resolving 1v1 and zonal matchups
- 4. Team-level prior calibration (FBref possession and attack/defense efficiencies)
- 5. Bivariate Poisson matrix with Dixon-Coles low-score correction
Standing & form context
| Metric | Coventry | Brighton |
|---|
| League position | 20.00 | 10.00 |
| Points per game | 0.00 | 1.33 |
| Attack index | 1.00 | 1.67 |
| Defense (concede) index | 1.04 | 2.08 |
| Form multiplier | 0.95 | 1.05 |
| Standing λ multiplier | 1.01 | 1.65 |
Lambda decomposition (how the goal rates were built)
| Stage | Coventry λ | Brighton λ |
|---|
| Pre-competitive-floor | 0.892 | 1.657 |
| Pre-calibration (Micro-duels) | 0.892 | 1.657 |
| Final λ | 1.050 | 1.657 |
competition calibration scale 1.0competition observed total goals 0.0quality gap applied Truemismatch score pressure applied Falsecompetitive underdog floor applied Truestage discount 1.0
Scoreline distribution & markets
| Top 5 scorelines | P |
|---|
| 0-1 | 11.0% |
| 1-1 | 11.7% |
| 1-2 | 9.6% |
| 0-2 | 9.2% |
| 1-0 | 6.9% |
| Home win | 23.9% |
| Draw | 24.6% |
| Away win | 51.5% |
| Over 2.5 | 50.8% |
| BTTS yes | 52.7% |
Pregame fair line — favorite Brighton (51.5%), fair total 2.71, fair spread -0.61
Projected players vs season baseline
Proj = projected for this match · Season = this season's per-game average from match data · deviation badge compares projected xG to the player's season xG/game.
Coventry — projected line vs season baseline (3-4-3)
| Player | Pos | Min | Proj G | Proj xG | Season xG/g | | Proj Sh | Season Sh/g | Proj Rtg | Season Rtg | Rec | Model / FBref Prior |
|---|
| Ephron Mason-Clark | M | 60 | 0 | 0.13 | - | n/a | 2 | 0.0 | 7.1 | 42.7 | 3 | FBref Prior M_service |
| Taiwo Awoniyi | F | 60 | 0 | 0.17 | 0.31 | 44% | 2 | 1.2 | 6.6 | 49.5 | 4 | FBref Prior F_service |
| Jack Rudoni | M | 47 | 0 | 0.09 | 0.12 | ok | 2 | 0.8 | 7.3 | 48.5 | 4 | XGBoost M_service |
| Jay DaSilva | D | 90 | 0 | 0.17 | 0.05 | 267% | 1 | 0.2 | 6.7 | 49.8 | 4 | FBref Prior D_service |
| Matt Grimes | M | 89 | 0 | 0.05 | 0.01 | 379% | 1 | 0.5 | 6.8 | 51.0 | 4 | FBref Prior M_service |
| Milan van Ewijk | D | 94 | 0 | 0.00 | - | n/a | 0 | 0.0 | 6.9 | 47.2 | 4 | FBref Prior D_service |
| Ethan Pinnock | D | 50 | 0 | 0.10 | 0.03 | 271% | 1 | 0.2 | 6.6 | 33.2 | 4 | XGBoost D_service |
| Bobby Thomas | D | 88 | 0 | 0.12 | 0.06 | 93% | 2 | 1.5 | 7.0 | 56.8 | 4 | FBref Prior D_service |
| Aurèle Amenda | D | 91 | 0 | 0.02 | 0.00 | 388% | 1 | 0.2 | 6.7 | 52.0 | 4 | FBref Prior D_service |
| Carl Rushworth | G | 92 | 0 | 0.00 | - | n/a | 0 | 0.0 | 7.4 | 51.5 | 4 | FBref Prior G_service |
| Frank Onyeka | M | 76 | 0 | 0.03 | - | n/a | 1 | 0.0 | 6.8 | 42.3 | 3 | FBref Prior M_service |
| Team totals | 0 | 1.05 | | 15 | | 6.9 | | |
Bench Candidates (7 modeled substitutes with empirical playing time)
| Player | Pos | Exp Min | App Prob | Substitute Modeling Details |
|---|
| Daniel Bentley | G | 4.4m | 10.9% | Market-value substitute prior |
| Liam Kitching | D | 14.1m | 32.5% | Market-value substitute prior |
| Joel Latibeaudiere | D | 13.6m | 29.9% | Market-value substitute prior |
| Caleb Yirenkyi | M | 22.7m | 32.2% | Market-value substitute prior |
| Gustavo Hamer | M | 18.4m | 70.3% | Market-value substitute prior |
| Sidiki Cherif | F | 25.9m | 58.8% | Market-value substitute prior |
Brighton — projected line vs season baseline (4-2-3-1)
| Player | Pos | Min | Proj G | Proj xG | Season xG/g | | Proj Sh | Season Sh/g | Proj Rtg | Season Rtg | Rec | Model / FBref Prior |
|---|
| Pascal Groß | M | 88 | 0 | 0.22 | 0.28 | ok | 5 | 2.0 | 6.8 | 68.7 | 3 | FBref Prior M_service |
| Olivier Boscagli | D | 91 | 0 | 0.00 | 0.02 | 100% | 0 | 0.3 | 7.0 | 73.3 | 3 | FBref Prior D_service |
| Lewis Dunk | D | 89 | 0 | 0.00 | - | n/a | 0 | 0.0 | 6.8 | 72.3 | 3 | FBref Prior D_service |
| Luka Vušković | D | 95 | 0 | 0.19 | 0.12 | 68% | 3 | 2.0 | 7.6 | 75.0 | 3 | FBref Prior D_service |
| Ferdi Kadıoğlu | D | 91 | 0 | 0.06 | 0.04 | 58% | 2 | 1.5 | 7.1 | 60.0 | 2 | XGBoost D_service |
| Bart Verbruggen | G | 95 | 0 | 0.00 | - | n/a | 0 | 0.0 | 7.3 | 68.0 | 3 | FBref Prior G_service |
| Yasin Ayari | M | 83 | 0 | 0.10 | 0.14 | 31% | 2 | 2.0 | 7.4 | 66.0 | 2 | FBref Prior M_service |
| Diego Gómez | M | 94 | 0 | 0.31 | 0.22 | 44% | 4 | 4.0 | 7.3 | 69.7 | 3 | FBref Prior M_service |
| Malick Yalcouyé | M | 59 | 1 | 0.18 | 0.25 | ok | 3 | 1.0 | 7.1 | 71.3 | 3 | FBref Prior M_service |
| Maxim De Cuyper | D | 90 | 0 | 0.03 | 0.50 | 94% | 1 | 1.7 | 7.1 | 70.7 | 3 | FBref Prior D_service |
| Charalampos Kostoulas | F | 69 | 0 | 0.25 | 0.26 | ok | 3 | 2.0 | 7.2 | 65.7 | 3 | FBref Prior F_service |
| Team totals | 1 | 1.66 | | 25 | | 7.2 | | |
Bench Candidates (7 modeled substitutes with empirical playing time)
| Player | Pos | Exp Min | App Prob | Substitute Modeling Details |
|---|
| Tom McGill | G | 4.4m | 10.9% | Market-value substitute prior |
| Mats Wieffer | D | 14.5m | 34.0% | Market-value substitute prior |
| Pascal Struijk | D | 14.4m | 33.9% | Market-value substitute prior |
| Yankuba Minteh | M | 23.4m | 56.2% | Market-value substitute prior |
| Georginio Rutter | M | 23.1m | 55.0% | Market-value substitute prior |
| Evan Ferguson | F | 26.1m | 59.5% | Market-value substitute prior |
Antagonist / tactical summary
Midfield is evenly contested — Jack Rudoni vs Yasin Ayari will be the pivotal duel. Total antagonist xG — Home: 0.50, Away: 0.38. Home danger zone: Shadow Striker vs Centre Back — Half-Space Threat. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.
Midfield — possession 49.2% / 50.8%, dominance ratio 0.48, key duel: Jack Rudoni vs Yasin Ayari
Matt Grimes xG 0.05Jack Rudoni xG 0.10Ephron Mason-Clark xG 0.13Pascal Groß xG 0.14Yasin Ayari xG 0.07Charalampos Kostoulas xG 0.17
Summary
Antagonist matchup analysis for Coventry vs Brighton. Most likely scoreline: 0-1. Lambdas derived from 5 home and 3 away individual player duels.
Tactical parity between 3-4-3 and 4-2-3-1. Match outcome driven primarily by individual player quality. Midfield is evenly contested — Jack Rudoni vs Yasin Ayari will be the pivotal duel. Total antagonist xG — Home: 0.50, Away: 0.38. Home danger zone: Shadow Striker vs Centre Back — Half-Space Threat. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.
Getafe vs Deportivo
LaLiga · 20260913 · 5-3-2 vs 4-4-2
Predicted 0-1Winner: DeportivoConfidence 0.63λ home/away 1.02 / 1.13Most-likely P 13.0%
Actual vs predicted: result not available yet (upcoming fixture).
XGBoost Reasoning & FBref Dataset TracingAntagonist + XGBoost match analysis for Getafe vs Deportivo. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 5-3-2 vs 4-4-2 (duel lambdas: Home 0.857, Away 1.134). Final calibrated Poisson lambdas: Home 1.020, Away 1.134 (most likely score 0-1).
Home Starters 11/11 XGBoost (11 FBref) Away Starters 11/11 XGBoost (9 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 0.86 / 1.13 Final λ (H/A) 1.02 / 1.13
Modeling Pipeline Stages (5 layers)
- 1. XGBoost player regressions for volume & event rates with market-value quality scaling
- 2. Multi-source prior blending (FBref 2026-2027 tables + Sofascore season history)
- 3. Antagonist formation duel matrix resolving 1v1 and zonal matchups
- 4. Team-level prior calibration (FBref possession and attack/defense efficiencies)
- 5. Bivariate Poisson matrix with Dixon-Coles low-score correction
Standing & form context
| Metric | Getafe | Deportivo |
|---|
| League position | 15.00 | 5.00 |
| Points per game | 1.00 | 2.00 |
| Attack index | 0.42 | 1.67 |
| Defense (concede) index | 1.04 | 1.04 |
| Form multiplier | 0.95 | 1.05 |
| Standing λ multiplier | 0.72 | 1.26 |
Lambda decomposition (how the goal rates were built)
| Stage | Getafe λ | Deportivo λ |
|---|
| Pre-competitive-floor | 0.857 | 1.134 |
| Pre-calibration (Micro-duels) | 0.857 | 1.134 |
| Final λ | 1.020 | 1.134 |
competition calibration scale 1.0competition observed total goals 0.0quality gap applied Falsemismatch score pressure applied Falsecompetitive underdog floor applied Truestage discount 1.0
Scoreline distribution & markets
| Top 5 scorelines | P |
|---|
| 0-1 | 13.0% |
| 1-1 | 13.6% |
| 0-0 | 11.7% |
| 1-0 | 11.7% |
| 1-2 | 7.6% |
| Home win | 32.3% |
| Draw | 29.7% |
| Away win | 38.0% |
| Over 2.5 | 36.5% |
| BTTS yes | 43.5% |
Pregame fair line — favorite Deportivo (38.0%), fair total 2.15, fair spread -0.11
Projected players vs season baseline
Proj = projected for this match · Season = this season's per-game average from match data · deviation badge compares projected xG to the player's season xG/game.
Getafe — projected line vs season baseline (5-3-2)
| Player | Pos | Min | Proj G | Proj xG | Season xG/g | | Proj Sh | Season Sh/g | Proj Rtg | Season Rtg | Rec | Model / FBref Prior |
|---|
| Martín Satriano | F | 83 | 0 | 0.28 | 0.15 | 80% | 2 | 2.0 | 6.9 | 51.5 | 4 | FBref Prior F_service |
| Enes Ünal | F | 90 | 0 | 0.22 | 0.27 | ok | 2 | 1.0 | 6.7 | 50.5 | 4 | FBref Prior F_service |
| Mario Martín | M | 91 | 0 | 0.11 | 0.03 | 207% | 2 | 1.0 | 6.8 | 33.0 | 2 | FBref Prior M_service |
| Orel Mangala | M | 89 | 0 | 0.04 | 0.03 | 55% | 2 | 0.5 | 6.9 | 49.5 | 4 | FBref Prior M_service |
| Ramón Terrats | M | 88 | 0 | 0.19 | 0.14 | 42% | 3 | 2.5 | 6.8 | 52.2 | 4 | FBref Prior M_service |
| Johan Mojica | D | 88 | 0 | 0.00 | 0.00 | 100% | 0 | 0.2 | 6.5 | 51.2 | 4 | FBref Prior D_service |
| Sebastián Boselli | D | 6 | 0 | 0.00 | - | n/a | 0 | 0.0 | 5.8 | 16.8 | 4 | FBref Prior D_service |
| Nemanja Gudelj | M | 14 | 0 | 0.07 | 0.01 | 525% | 2 | 0.5 | 6.7 | 32.5 | 4 | FBref Prior M_service |
| Djené | D | 88 | 0 | 0.00 | - | n/a | 0 | 0.0 | 6.4 | 48.2 | 4 | FBref Prior D_service |
| Andrés García | D | 82 | 0 | 0.00 | 0.03 | 100% | 0 | 0.8 | 6.7 | 53.0 | 4 | FBref Prior D_service |
| David Soria | G | 88 | 0 | 0.00 | - | n/a | 0 | 0.0 | 7.0 | 56.2 | 4 | FBref Prior G_service |
| Team totals | 0 | 1.02 | | 14 | | 6.7 | | |
Bench Candidates (7 modeled substitutes with empirical playing time)
| Player | Pos | Exp Min | App Prob | Substitute Modeling Details |
|---|
| Jiří Letáček | G | 4.4m | 10.9% | Market-value substitute prior |
| Abdel Abqar | D | 6.7m | 41.0% | Market-value substitute prior |
| Zaid Romero | D | 13.6m | 21.7% | Market-value substitute prior |
| Francho Serrano | M | 22.0m | 30.7% | Market-value substitute prior |
| Oscar Lopez | M | 22.0m | 49.9% | Market-value substitute prior |
| Christantus Uche | F | 25.8m | 58.1% | Market-value substitute prior |
Deportivo — projected line vs season baseline (4-4-2)
| Player | Pos | Min | Proj G | Proj xG | Season xG/g | | Proj Sh | Season Sh/g | Proj Rtg | Season Rtg | Rec | Model / FBref Prior |
|---|
| Pierre-Emerick Aubameyang | F | 87 | 1 | 0.52 | 0.52 | ok | 2 | 2.6 | 7.2 | 63.6 | 5 | FBref Prior F_service |
| Bil Nsongo | F | 52 | 0 | 0.23 | 0.06 | 265% | 2 | 0.8 | 6.6 | 54.6 | 5 | FBref Prior F_service |
| Marc Casadó | M | 25 | 0 | 0.08 | - | n/a | 2 | 0.0 | 6.8 | 34.0 | 2 | XGBoost M_service |
| Lorenzo Amatucci | M | 90 | 0 | 0.05 | - | n/a | 2 | 0.0 | 6.8 | 56.4 | 5 | FBref Prior M_service |
| Mario Soriano | M | 73 | 0 | 0.06 | 0.02 | 226% | 1 | 0.6 | 6.8 | 54.8 | 5 | FBref Prior M_service |
| Luismi Cruz | M | 73 | 0 | 0.10 | 0.06 | 62% | 1 | 0.6 | 6.8 | 55.4 | 5 | XGBoost M_service |
| Giacomo Quagliata | D | 88 | 0 | 0.00 | 0.00 | 100% | 0 | 0.2 | 6.5 | 51.6 | 5 | FBref Prior D_service |
| José María Giménez | D | 34 | 0 | 0.00 | - | n/a | 0 | 0.0 | 6.7 | 40.8 | 5 | FBref Prior D_service |
| Lucas Noubi | D | 88 | 0 | 0.00 | - | n/a | 0 | 0.0 | 6.6 | 53.2 | 5 | FBref Prior D_service |
| Ximo Navarro | D | 87 | 0 | 0.00 | 0.01 | 100% | 1 | 0.4 | 6.4 | 51.8 | 5 | FBref Prior D_service |
| Leo Román | G | 89 | 0 | 0.00 | - | n/a | 0 | 0.0 | 7.0 | 58.4 | 5 | FBref Prior G_service |
| Team totals | 1 | 1.13 | | 11 | | 6.8 | | |
Bench Candidates (7 modeled substitutes with empirical playing time)
| Player | Pos | Exp Min | App Prob | Substitute Modeling Details |
|---|
| Álvaro Ferllo | G | 4.4m | 10.9% | Market-value substitute prior |
| Adrià Altimira | D | 13.6m | 29.9% | Market-value substitute prior |
| Dani Barcia | D | 13.6m | 29.9% | Market-value substitute prior |
| David Mella | M | 22.1m | 50.2% | Market-value substitute prior |
| Angeliño | M | 22.0m | 50.1% | Market-value substitute prior |
| Yeremay Hernández | F | 26.2m | 59.9% | Market-value substitute prior |
Antagonist / tactical summary
Midfield is evenly contested — Mario Martín vs Marc Casadó will be the pivotal duel. Total antagonist xG — Home: 0.75, Away: 0.62. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.
Midfield — possession 49.4% / 50.6%, dominance ratio 0.49, key duel: Mario Martín vs Marc Casadó
Johan Mojica xG 0.07Orel Mangala xG 0.03Ramón Terrats xG 0.16Luismi Cruz xG 0.06Marc Casadó xG 0.05Lorenzo Amatucci xG 0.03
Summary
Antagonist matchup analysis for Getafe vs Deportivo. Most likely scoreline: 0-1. Lambdas derived from 6 home and 6 away individual player duels.
Tactical parity between 5-3-2 and 4-4-2. Match outcome driven primarily by individual player quality. Midfield is evenly contested — Mario Martín vs Marc Casadó will be the pivotal duel. Total antagonist xG — Home: 0.75, Away: 0.62. Home danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.
Celta vs Málaga
LaLiga · 20260913 · 3-4-3 vs 4-3-3
Predicted 0-1Winner: MálagaConfidence 0.64λ home/away 0.82 / 1.02Most-likely P 16.0%
Actual vs predicted: result not available yet (upcoming fixture).
XGBoost Reasoning & FBref Dataset TracingAntagonist + XGBoost match analysis for Celta vs Málaga. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 3-4-3 vs 4-3-3 (duel lambdas: Home 0.823, Away 0.600). Final calibrated Poisson lambdas: Home 0.823, Away 1.020 (most likely score 0-1).
Home Starters 10/11 XGBoost (10 FBref) Away Starters 11/11 XGBoost (8 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 0.82 / 0.60 Final λ (H/A) 0.82 / 1.02
FBref Team Priors: Celta Vigo (Gls/90: 0.25, GA/90: 0.25 → E[λ]: 0.2) vs Málaga CF (Gls/90: 0.3333333333333333, GA/90: 0.33 → E[λ]: 0.2) | Blend Weight: 0.228
Modeling Pipeline Stages (5 layers)
- 1. XGBoost player regressions for volume & event rates with market-value quality scaling
- 2. Multi-source prior blending (FBref 2026-2027 tables + Sofascore season history)
- 3. Antagonist formation duel matrix resolving 1v1 and zonal matchups
- 4. Team-level prior calibration (FBref possession and attack/defense efficiencies)
- 5. Bivariate Poisson matrix with Dixon-Coles low-score correction
Standing & form context
| Metric | Celta | Málaga |
|---|
| League position | 16.00 | 18.00 |
| Points per game | 0.60 | 0.50 |
| Attack index | 0.42 | 0.21 |
| Defense (concede) index | 1.46 | 1.04 |
| Form multiplier | 1.01 | 0.99 |
| Standing λ multiplier | 0.82 | 0.64 |
Lambda decomposition (how the goal rates were built)
| Stage | Celta λ | Málaga λ |
|---|
| FBref Squad Prior E[λ] (weight 0.228) | 0.200 | 0.200 |
| Pre-competitive-floor | 0.823 | 0.650 |
| Pre-calibration (Micro-duels) | 0.823 | 0.600 |
| Final λ | 0.823 | 1.020 |
competition calibration scale 1.0competition observed total goals 0.0quality gap applied Truemismatch score pressure applied Falsecompetitive underdog floor applied Truestage discount 1.0
Scoreline distribution & markets
| Top 5 scorelines | P |
|---|
| 0-1 | 16.0% |
| 0-0 | 16.0% |
| 1-1 | 13.4% |
| 1-0 | 12.9% |
| 0-2 | 8.2% |
| Home win | 28.4% |
| Draw | 32.5% |
| Away win | 39.1% |
| Over 2.5 | 28.1% |
| BTTS yes | 36.0% |
Pregame fair line — favorite Málaga (39.1%), fair total 1.84, fair spread -0.20
Projected players vs season baseline
Proj = projected for this match · Season = this season's per-game average from match data · deviation badge compares projected xG to the player's season xG/game.
Celta — projected line vs season baseline (3-4-3)
| Player | Pos | Min | Proj G | Proj xG | Season xG/g | | Proj Sh | Season Sh/g | Proj Rtg | Season Rtg | Rec | Model / FBref Prior |
|---|
| Williot Swedberg | M | 34 | 0 | 0.08 | 0.08 | ok | 2 | 0.6 | 7.0 | 54.4 | 5 | FBref Prior M_service |
| Ferrán Jutglà | F | 50 | 0 | 0.16 | 0.01 | 2885% | 3 | 0.2 | 6.8 | 65.2 | 5 | FBref Prior F_service |
| Hugo González | M | 90 | 0 | 0.07 | 0.07 | ok | 1 | 0.8 | 6.9 | 27.4 | 5 | FBref Prior M_service |
| Sergio Carreira | M | 91 | 0 | 0.06 | 0.03 | 79% | 2 | 0.5 | 7.0 | 58.7 | 6 | FBref Prior M_service |
| Miguel Román | M | 83 | 0 | 0.05 | 0.01 | 508% | 2 | 0.6 | 7.1 | 69.6 | 5 | FBref Prior M_service |
| Ilaix Moriba | M | 67 | 0 | 0.03 | 0.01 | 160% | 2 | 0.6 | 7.0 | 66.0 | 5 | FBref Prior M_service |
| Javier Rueda | M | 74 | 0 | 0.04 | 0.01 | 274% | 1 | 0.4 | 6.7 | 66.6 | 5 | XGBoost M_service |
| Marcos Alonso | D | 79 | 0 | 0.19 | 0.06 | 220% | 2 | 0.5 | 6.6 | 70.0 | 4 | FBref Prior D_service |
| Carl Starfelt | D | 72 | 0 | 0.08 | 0.06 | 35% | 2 | 0.8 | 6.8 | 42.6 | 5 | FBref Prior D_service |
| Yoel Lago | D | 80 | 0 | 0.00 | - | n/a | 0 | 0.0 | 6.7 | 68.2 | 5 | FBref Prior D_service |
| Ionuț Radu | G | 90 | 0 | 0.00 | - | n/a | 0 | 0.0 | 6.8 | 72.4 | 5 | FBref Prior G_service |
| Team totals | 0 | 0.82 | | 17 | | 6.9 | | |
Bench Candidates (7 modeled substitutes with empirical playing time)
| Player | Pos | Exp Min | App Prob | Substitute Modeling Details |
|---|
| Altay Bayındır | G | 4.5m | 11.3% | Market-value substitute prior |
| Javi Rodríguez | D | 10.3m | 32.7% | Market-value substitute prior |
| Abdoulaye Faye | D | 13.6m | 21.8% | Market-value substitute prior |
| Hugo Álvarez | M | 26.5m | 56.8% | Market-value substitute prior |
| Aleix Febas | M | 22.1m | 30.9% | Market-value substitute prior |
| Jones El-Abdellaoui | F | 25.8m | 53.2% | Market-value substitute prior |
Málaga — projected line vs season baseline (4-3-3)
| Player | Pos | Min | Proj G | Proj xG | Season xG/g | | Proj Sh | Season Sh/g | Proj Rtg | Season Rtg | Rec | Model / FBref Prior |
|---|
| Joaquín Muñoz | M | 66 | 0 | 0.12 | 0.01 | 1322% | 1 | 0.3 | 6.8 | 44.0 | 6 | FBref Prior M_service |
| Chupete | F | 91 | 1 | 0.38 | 0.58 | 36% | 3 | 2.4 | 6.8 | 51.2 | 5 | FBref Prior F_service |
| David Larrubia | M | 89 | 0 | 0.18 | 0.09 | 112% | 2 | 1.5 | 6.8 | 44.8 | 6 | FBref Prior M_service |
| Pablo Martínez | M | 51 | 0 | 0.04 | 0.00 | 1020% | 2 | 0.2 | 6.8 | 33.0 | 4 | FBref Prior M_service |
| Dani Lorenzo | M | 76 | 0 | 0.06 | 0.04 | 31% | 1 | 0.2 | 6.8 | 53.0 | 4 | FBref Prior M_service |
| Carlos Dotor | M | 74 | 0 | 0.08 | 0.02 | 229% | 2 | 0.4 | 6.8 | 55.8 | 5 | FBref Prior M_service |
| Jose Salinas | D | 45 | 0 | 0.00 | - | n/a | 0 | 0.0 | 6.5 | 39.2 | 5 | FBref Prior D_service |
| Einar Galilea | D | 82 | 0 | 0.00 | - | n/a | 0 | 0.0 | 6.6 | 53.6 | 5 | XGBoost D_service |
| Ángel Recio | D | 88 | 0 | 0.02 | 0.00 | 610% | 1 | 0.2 | 6.5 | 52.0 | 5 | XGBoost D_service |
| Rafa Garrido | D | 83 | 0 | 0.03 | 0.01 | 450% | 0 | 0.2 | 6.5 | 52.8 | 5 | XGBoost D_service |
| Alfonso Herrero | G | 88 | 0 | 0.00 | - | n/a | 0 | 0.0 | 7.1 | 59.8 | 5 | FBref Prior G_service |
| Team totals | 1 | 1.02 | | 13 | | 6.7 | | |
Bench Candidates (7 modeled substitutes with empirical playing time)
| Player | Pos | Exp Min | App Prob | Substitute Modeling Details |
|---|
| Carlos Lopez | G | 4.4m | 10.9% | Market-value substitute prior |
| Adam Aznou | D | 13.7m | 30.6% | Market-value substitute prior |
| Fernando Calero | D | 13.6m | 29.9% | Market-value substitute prior |
| Jens-Lys Cajuste | M | 22.1m | 50.7% | Market-value substitute prior |
| Izan Merino | M | 22.0m | 49.9% | Market-value substitute prior |
| Adrián Niño | F | 25.3m | 55.9% | Market-value substitute prior |
Antagonist / tactical summary
Midfield is evenly contested — Williot Swedberg vs David Larrubia will be the pivotal duel. Total antagonist xG — Home: 0.62, Away: 0.64. Home danger zone: Shadow Striker vs Centre Back — Half-Space Threat. Away danger zone: LW vs CB.
Midfield — possession 51.6% / 48.4%, dominance ratio 0.56, key duel: Williot Swedberg vs David Larrubia
Hugo González xG 0.08Ilaix Moriba xG 0.05Miguel Román xG 0.07Joaquín Muñoz xG 0.09David Larrubia xG 0.14Chupete xG 0.33
Summary
Antagonist matchup analysis for Celta vs Málaga. Most likely scoreline: 0-1. Lambdas derived from 7 home and 5 away individual player duels.
Tactical parity between 3-4-3 and 4-3-3. Match outcome driven primarily by individual player quality. Midfield is evenly contested — Williot Swedberg vs David Larrubia will be the pivotal duel. Total antagonist xG — Home: 0.62, Away: 0.64. Home danger zone: Shadow Striker vs Centre Back — Half-Space Threat. Away danger zone: LW vs CB.
Man Utd vs Man City
Premier League · 20260913 · 4-2-3-1 vs 4-2-3-1
Predicted 1-2Winner: Man CityConfidence 0.68λ home/away 1.49 / 2.42Most-likely P 8.7%
Actual vs predicted: result not available yet (upcoming fixture).
XGBoost Reasoning & FBref Dataset TracingAntagonist + XGBoost match analysis for Man Utd vs Man City. Player statistics regressed using position-specific XGBoost models (G, D, M, F) with 2026-2027 FBref priors. Micro-duels resolved across 4-2-3-1 vs 4-2-3-1 (duel lambdas: Home 1.489, Away 2.424). Final calibrated Poisson lambdas: Home 1.489, Away 2.424 (most likely score 1-2).
Home Starters 10/11 XGBoost (11 FBref) Away Starters 11/11 XGBoost (11 FBref) Bench Modeled H: 7 / A: 7 Duel λ (H/A) 1.49 / 2.42 Final λ (H/A) 1.49 / 2.42
FBref Team Priors: Manchester United (Gls/90: 2.0, GA/90: 2.5 → E[λ]: 3.5) vs Manchester City (Gls/90: 3.0, GA/90: 3.0 → E[λ]: 3.5) | Blend Weight: 0.192
Modeling Pipeline Stages (5 layers)
- 1. XGBoost player regressions for volume & event rates with market-value quality scaling
- 2. Multi-source prior blending (FBref 2026-2027 tables + Sofascore season history)
- 3. Antagonist formation duel matrix resolving 1v1 and zonal matchups
- 4. Team-level prior calibration (FBref possession and attack/defense efficiencies)
- 5. Bivariate Poisson matrix with Dixon-Coles low-score correction
Standing & form context
| Metric | Man Utd | Man City |
|---|
| League position | 11.00 | 1.00 |
| Points per game | 1.33 | 3.00 |
| Attack index | 2.92 | 1.46 |
| Defense (concede) index | 0.42 | 2.50 |
| Form multiplier | 0.94 | 1.06 |
| Standing λ multiplier | 1.07 | 1.65 |
Lambda decomposition (how the goal rates were built)
| Stage | Man Utd λ | Man City λ |
|---|
| FBref Squad Prior E[λ] (weight 0.192) | 3.500 | 3.500 |
| Pre-competitive-floor | 1.489 | 2.424 |
| Pre-calibration (Micro-duels) | 1.489 | 2.424 |
| Final λ | 1.489 | 2.424 |
competition calibration scale 1.0competition observed total goals 0.0quality gap applied Truemismatch score pressure applied Falsecompetitive underdog floor applied Falsestage discount 1.0
Scoreline distribution & markets
| Top 5 scorelines | P |
|---|
| 1-2 | 8.7% |
| 1-1 | 7.3% |
| 1-3 | 7.1% |
| 2-2 | 6.5% |
| 0-2 | 5.9% |
| Home win | 22.8% |
| Draw | 19.1% |
| Away win | 58.1% |
| Over 2.5 | 74.9% |
| BTTS yes | 70.6% |
Pregame fair line — favorite Man City (58.1%), fair total 3.91, fair spread -0.94
Projected players vs season baseline
Proj = projected for this match · Season = this season's per-game average from match data · deviation badge compares projected xG to the player's season xG/game.
Man Utd — projected line vs season baseline (4-2-3-1)
| Player | Pos | Min | Proj G | Proj xG | Season xG/g | | Proj Sh | Season Sh/g | Proj Rtg | Season Rtg | Rec | Model / FBref Prior |
|---|
| Matheus Cunha | F | 85 | 0 | 0.19 | 0.16 | ok | 2 | 2.3 | 7.5 | 63.7 | 3 | FBref Prior F_service |
| Marcus Rashford | F | 49 | 0 | 0.29 | 0.06 | 348% | 3 | 1.0 | 7.5 | 50.8 | 4 | FBref Prior F_service |
| Bruno Fernandes | M | 95 | 0 | 0.29 | 0.75 | 61% | 2 | 5.0 | 7.4 | 78.0 | 3 | FBref Prior M_service |
| Bryan Mbeumo | F | 90 | 1 | 0.38 | 0.58 | 34% | 4 | 2.8 | 7.5 | 53.0 | 4 | FBref Prior F_service |
| Kobbie Mainoo | M | 69 | 0 | 0.06 | 0.07 | ok | 1 | 1.3 | 7.9 | 73.7 | 3 | FBref Prior M_service |
| Youri Tielemans | M | 90 | 0 | 0.08 | 0.18 | 58% | 2 | 2.0 | 7.9 | 71.7 | 3 | FBref Prior M_service |
| Luke Shaw | D | 84 | 0 | 0.00 | 0.01 | 100% | 0 | 0.3 | 6.7 | 69.3 | 3 | FBref Prior D_service |
| Lisandro Martínez | D | 95 | 0 | 0.00 | 0.01 | 100% | 1 | 0.3 | 7.4 | 44.3 | 3 | FBref Prior D_service |
| Harry Maguire | D | 90 | 0 | 0.00 | 0.03 | 100% | 1 | 1.0 | 7.0 | 68.0 | 3 | FBref Prior D_service |
| Diogo Dalot | D | 59 | 0 | 0.00 | 0.05 | 100% | 2 | 1.7 | 7.0 | 65.3 | 3 | FBref Prior D_service |
| Senne Lammens | G | 95 | 0 | 0.00 | - | n/a | 0 | 0.0 | 7.4 | 68.0 | 3 | FBref Prior G_service |
| Team totals | 1 | 1.49 | | 20 | | 7.4 | | |
Bench Candidates (7 modeled substitutes with empirical playing time)
| Player | Pos | Exp Min | App Prob | Substitute Modeling Details |
|---|
| Karl Darlow | G | 4.4m | 10.9% | Market-value substitute prior |
| Leny Yoro | D | 15.0m | 36.6% | Market-value substitute prior |
| Matthijs de Ligt | D | 14.7m | 35.2% | Market-value substitute prior |
| Carlos Baleba | M | 23.5m | 57.0% | Market-value substitute prior |
| Amad Diallo | M | 23.4m | 56.4% | Market-value substitute prior |
| Benjamin Šeško | F | 27.1m | 63.8% | Market-value substitute prior |
Man City — projected line vs season baseline (4-2-3-1)
| Player | Pos | Min | Proj G | Proj xG | Season xG/g | | Proj Sh | Season Sh/g | Proj Rtg | Season Rtg | Rec | Model / FBref Prior |
|---|
| Enzo Fernández | M | 56 | 0 | 0.17 | 0.04 | 297% | 2 | 0.7 | 7.9 | 47.3 | 3 | FBref Prior M_service |
| Joško Gvardiol | D | 94 | 0 | 0.00 | 0.05 | 100% | 1 | 0.5 | 7.4 | 53.8 | 4 | FBref Prior D_service |
| Marc Guéhi | D | 95 | 0 | 0.00 | 0.22 | 100% | 0 | 0.5 | 7.8 | 58.5 | 4 | FBref Prior D_service |
| Rúben Dias | D | 95 | 0 | 0.00 | 0.02 | 100% | 0 | 0.2 | 7.7 | 57.2 | 4 | FBref Prior D_service |
| Abdukodir Khusanov | D | 95 | 0 | 0.00 | - | n/a | 0 | 0.0 | 7.4 | 52.8 | 4 | FBref Prior D_service |
| Gianluigi Donnarumma | G | 95 | 0 | 0.00 | - | n/a | 0 | 0.0 | 7.9 | 57.8 | 4 | FBref Prior G_service |
| Elliot Anderson | M | 88 | 0 | 0.20 | 0.02 | 793% | 1 | 0.8 | 8.5 | 57.8 | 4 | FBref Prior M_service |
| Antoine Semenyo | M | 95 | 0 | 0.32 | 0.04 | 641% | 2 | 1.2 | 7.7 | 50.2 | 4 | FBref Prior M_service |
| Iliman Ndiaye | M | 95 | 0 | 0.29 | 0.07 | 310% | 2 | 1.2 | 7.5 | 53.0 | 4 | FBref Prior M_service |
| Erling Haaland | F | 95 | 2 | 1.02 | 0.62 | 65% | 3 | 3.5 | 8.2 | 59.2 | 4 | FBref Prior F_service |
| Rayan Cherki | M | 65 | 0 | 0.28 | 0.14 | 103% | 2 | 1.5 | 7.9 | 58.0 | 4 | FBref Prior M_service |
| Team totals | 2 | 2.42 | | 13 | | 7.8 | | |
Bench Candidates (7 modeled substitutes with empirical playing time)
| Player | Pos | Exp Min | App Prob | Substitute Modeling Details |
|---|
| Gerónimo Rulli | G | 4.5m | 11.5% | Market-value substitute prior |
| Nico O'Reilly | D | 15.3m | 25.3% | Market-value substitute prior |
| Matheus Nunes | D | 25.5m | 63.3% | Market-value substitute prior |
| Ayyoub Bouaddi | M | 15.6m | 72.8% | Market-value substitute prior |
| Jérémy Doku | M | 23.7m | 57.7% | Market-value substitute prior |
| Omar Marmoush | F | 26.9m | 63.1% | Market-value substitute prior |
Antagonist / tactical summary
Midfield is evenly contested — Kobbie Mainoo vs Elliot Anderson will be the pivotal duel. Total antagonist xG — Home: 0.55, Away: 1.08. Home danger zone: Left Wing vs Right Back — Flank Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.
Midfield — possession 49.0% / 51.0%, dominance ratio 0.46, key duel: Kobbie Mainoo vs Elliot Anderson
Bruno Fernandes xG 0.29Matheus Cunha xG 0.20Kobbie Mainoo xG 0.06Antoine Semenyo xG 0.20Rayan Cherki xG 0.19Enzo Fernández xG 0.06
Summary
Antagonist matchup analysis for Man Utd vs Man City. Most likely scoreline: 1-2. Lambdas derived from 3 home and 5 away individual player duels.
Tactical parity between 4-2-3-1 and 4-2-3-1. Match outcome driven primarily by individual player quality. Midfield is evenly contested — Kobbie Mainoo vs Elliot Anderson will be the pivotal duel. Total antagonist xG — Home: 0.55, Away: 1.08. Home danger zone: Left Wing vs Right Back — Flank Battle. Away danger zone: Striker vs Two Centre Backs — 1v2 Aerial/Physical Battle.
Generated 2026-09-13 13:34:12 UTC · Soccer prediction reporting