Transparent model tracking

Navurex model performance.

These results use frozen pre-match prediction snapshots. Once kickoff passes, the model probabilities are not rewritten using the final score or future matches.

Integrity-verified history

129 historical settled rows are excluded by the integrity gate and do not affect these metrics.

Settled predictions

1684

441 pending snapshots

Pick accuracy

43.3%

730 correct top-outcome picks

High confidence

49.0%

1077 settled at ≥70 confidence

Brier score

0.6566

Lower is better · evaluates all 1X2 probabilities

Log loss

1.0999

Lower is better · strongly penalizes confident wrong probabilities

Recent window

43.0%

Last 30 days pick accuracy

1627 matches

How to read this

Pick accuracy measures the top 1X2 choice. Brier score evaluates all three probabilities, while log loss penalizes confident wrong probabilities most strongly. Calibration checks whether events predicted at a given probability happen at roughly that rate.

V1 vs V2 benchmark

Is Contextual V2 beating the baseline?

Only settled, integrity-verified matches with both V1 shadow and V2 pre-news probabilities are included. News adjustments are excluded so this comparison isolates the contextual engine.

498 compared matches

Compared matches

498

217 improved · 185 worsened · 96 neutral

Brier · V1 → V2

0.6578 → 0.6547

+0.0031 · +0.5%

Log loss · V1 → V2

1.1069 → 1.1013

+0.0055

Pick accuracy · V1 → V2

44.8% → 45.6%

Top 1X2 outcome only; probability scores remain the primary benchmark.

Average probability movement

1.80 pp

Raw V2 stored

498 / 498

Diagnostics stored

498 / 498

V2 snapshot write health

Are new Raw V2 diagnostics being frozen correctly?

Healthy

Recent contextual snapshots

latest 20 window

Raw V2 coverage

100.0%

20 / 20

Diagnostics coverage

100.0%

20 / 20

Complete V2 snapshots

100.0%

20 / 20

Latest diagnostics capture: 19 Oct 2026

Keep the current blend unchanged until the sample grows. Navurex reports the benchmark without automatically tuning the model from a small sample.

League comparison

CompetitionSampleV1 BrierV2 BrierV2 Brier delta

Championship

0.6287

0.6423

-0.0136

3. Liga

0.8085

0.8007

+0.0078

UEFA Champions League

0.7496

0.7367

+0.0129

UEFA Europa League

0.6426

0.6350

+0.0076

2. Lig

0.5373

0.5353

+0.0020

UEFA Conference League

0.6775

0.6791

-0.0016

1. Division

0.6017

0.6101

-0.0084

La Liga 2

0.5635

0.5668

-0.0032

Serie B

0.6734

0.6906

-0.0171

Campeonato Brasileiro Série A

0.5581

0.5463

+0.0118

V2 vs V3 shadow benchmark

Is V3 improving the model?

Only settled, integrity-verified matches with both V2 pre-news and V3 shadow probabilities are included.

498 compared matches

Compared matches

498

256 improved · 229 worsened · 13 neutral

Brier · V2 → V3

0.6547 → 0.6225

+0.0323 · +4.9%

Log loss · V2 → V3

1.1013 → 1.0390

+0.0624

Pick accuracy · V2 → V3

45.6% → 47.4%

Top 1X2 outcome only; probability scores remain the primary benchmark.

Market coverage

414 / 498

Average probability movement

6.58 pp

V3 promotion gate

Keep V2 production until the V3 sample is large enough and probability scores are consistently better.

League comparison

CompetitionSampleV2 BrierV3 BrierV3 Brier delta

Championship

0.6423

0.6072

+0.0351

3. Liga

0.8007

0.7495

+0.0512

UEFA Conference League

0.6791

0.5880

+0.0911

UEFA Champions League

0.7367

0.6615

+0.0752

UEFA Europa League

0.6350

0.5669

+0.0681

2. Lig

0.5353

0.5228

+0.0125

1. Division

0.6101

0.6065

+0.0036

La Liga 2

0.5668

0.5549

+0.0119

Serie B

0.6906

0.6946

-0.0041

Campeonato Brasileiro Série A

0.5463

0.5168

+0.0295

V3 vs market benchmark

Does V3 add information beyond the bookmaker consensus?

The market baseline uses frozen pre-kickoff no-vig consensus from eligible bookmaker snapshots. V2, V3 and market metrics are calculated on the exact same settled matches.

408 eligible market snapshots

Market benchmark sample

408

408 strict V3.1 quality verified · 0 legacy eligible · 6 reference-only excluded

Brier · V3 vs market

0.6177 → 0.5762

V3 vs market Brier edge -0.0415 · -7.2%

Positive means V3 better; negative means market better.

Log loss · V3 vs market

1.0304 → 0.9702

-0.0602

Pick accuracy · V3 vs market

47.3% → 54.4%

V3-market pick agreement 69.1%

1X2 calibration error

3.06 pp → 3.18 pp

same verified market sample

Average bookmakers

8.6

Average market blend

34.4%

V3-market pick agreement

69.1%

Average disagreement

12.30 pp

Strict market quality split

Does the market advantage persist after V3.1 strict quality validation?

Strict V3.1 snapshots and older legacy-eligible snapshots are reported separately so market-quality changes cannot hide inside one combined benchmark.

Strict V3.1 cohort

Only snapshots with an explicit V3.1 market quality gate are included.

n=408

Brier · V3 → market

0.6177 → 0.5762

-0.0415

Log loss · V3 → market

1.0304 → 0.9702

-0.0602

Pick accuracy · V3 → market

47.3% → 54.4%

Agreement 69.1%

Calibration error · V3 → market

3.06 pp → 3.18 pp

Bookmakers

8.6

Market blend

34.4%

Disagreement

12.30 pp

Structural → final → market Brier: 0.6481 → 0.6177 → 0.5762 · n=408

Positive means V3 better; negative means market better.

Strict disagreement buckets

DisagreementSampleV3 BrierMarket BrierBrier edge

0–5 pp

n=55

0.4306

0.4279

-0.0028

5–10 pp

n=136

0.6177

0.5953

-0.0224

10–20 pp

n=170

0.6477

0.6302

-0.0175

20+ pp

n=47

0.7282

0.4992

-0.2290

V3.4 dynamic market anchor shadow

Does stronger market anchoring improve V3 when disagreement is large?

This is a frozen pre-match shadow A/B test. Current V3 probabilities remain unchanged. No historical backfill is used.

1 A/B sample

Pending V3.4 snapshots 60Settled V3.4 payloads 323Strict A/B payloads 268

A/B sample

1 anchor active

Brier · current V3 → V3.4 → market

0.8906 → 0.8906 → 0.8834

Current → V3.4 Brier improvement 0.0000

Log loss · current V3 → V3.4 → market

1.4697 → 1.4697 → 1.4222

0.0000

Pick accuracy · current V3 → V3.4 → market

0.0% → 0.0% → 0.0%

Calibration · current V3 → V3.4 → market

51.33 → 51.33 → 31.90 pp

Current market blend

34.0%

V3.4 target blend

35.0%

Average disagreement

8.00 pp

Disagreement bands · V3.4

5–10 pp

n=1

V3 0.8906

V3.4 0.8906

MKT 0.8834

0.0000

Independent V3 signal check

Structural V3 is measured before bookmaker blending. Comparing structural V3, final V3 and the market helps show whether the model adds information independently or mostly follows the market.

Structural sample

408

Brier · structural → final → market

0.6481 → 0.6177 → 0.5762

Market-blend effect

+0.0304

Structural V3 vs market edge

-0.0719

Where does V3 disagree with the market?

DisagreementSampleV3 BrierMarket BrierBrier edgeV3 pickMarket pick

0–5 pp

0.4306

0.4279

-0.0028

76.4%

81.8%

5–10 pp

136

0.6177

0.5953

-0.0224

50.7%

10–20 pp

170

0.6477

0.6302

-0.0175

41.8%

46.5%

20+ pp

0.7282

0.4992

-0.2290

23.4%

61.7%

Market calibration audit

V3 and market calibration on the same settled benchmark sample.

BandV3 predicted / observedMarket predicted / observed

40–50%

43.6% / 48.6% n=148

44.7% / 49.7% n=163

50–60%

53.8% / 53.5% n=71

54.3% / 56.6% n=83

60–70%

64.2% / 61.7% n=47

64.4% / 68.0% n=50

70–80%

74.5% / 75.8% n=33

75.0% / 89.7% n=29

80–90%

80.5% / 0.0% n=2

84.5% / 50.0% n=8

90–100%

—

91.5% / 100.0% n=1

League comparison

CompetitionSampleV3 BrierMarket BrierBrier edge

Championship

0.6072

0.6320

+0.0249

3. Liga

0.7680

0.7161

-0.0518

UEFA Champions League

0.6615

0.5674

-0.0941

UEFA Europa League

0.5669

0.4563

-0.1106

UEFA Conference League

0.5880

0.3948

-0.1932

1. Division

0.6065

0.5459

-0.0606

Serie B

0.6946

0.7004

+0.0058

La Liga 2

0.5576

0.5399

-0.0177

Campeonato Brasileiro Série A

0.5041

0.4285

-0.0756

Serie B

0.6812

0.7431

+0.0619

Market baseline is a benchmark, not ground truth. Reference-only or explicitly unusable markets are excluded.

Probability calibration

Do Navurex probabilities behave like probabilities?

Each settled 1X2 probability is grouped into a 10-point band. A well-calibrated model should have predicted probability and observed frequency close together.

Calibration error

7.43 pp

5052 probability observations

V5.5 calibration monitor

Monitor overconfidence before changing probabilities.

This monitor is diagnostic only. It does not recalibrate published probabilities. Navurex waits for enough V2 comparisons and enough observations inside the watched probability bands before considering a calibration change.

Calibration candidate

V2 comparison sample

498

Review at 50 · calibration candidate at 100

50–60% watch band

+3.3 pp

n=314

53.6% predicted · 50.3% observed

60–70% watch band

+5.9 pp

n=212

63.9% predicted · 58.0% observed

The monitoring thresholds are met. This only means a calibration experiment can be evaluated; it does not mean Navurex should automatically change production probabilities.

BandPredictedObservedGapSample

0–10%

8.0%

0.0%

+8.0 pp

1 observations

10–20%

13.8%

24.9%

-11.1 pp

1170 observations

20–30%

24.5%

30.1%

-5.6 pp

1030 observations

30–40%

34.6%

31.6%

+3.0 pp

966 observations

40–50%

42.1%

32.5%

+9.7 pp

1177 observations

50–60%

53.6%

50.3%

+3.3 pp

314 observations

60–70%

63.9%

58.0%

+5.9 pp

212 observations

70–80%

75.3%

63.2%

+12.1 pp

182 observations

News Intelligence impact

Did verified match news improve the prediction?

For settled matches where verified model-eligible news changed the probabilities, Navurex compares the base model with the final news-adjusted prediction against the actual result.

Settled with news

Improved

Worsened

Brier improvement

-1.3%

Average Brier · before news

0.6118

Average Brier · after news

0.6197

By competition

Where the model has been tested.

Sample size is shown beside every league so a small run of results is not mistaken for long-term evidence.

CompetitionSampleAccuracyBrierLog loss

UEFA Nations League

UNL

Sample96

Accuracy42.7%

Brier0.6648

Log loss1.0961

2. Lig

TR2

Sample65

Accuracy55.4%

Brier0.5993

Log loss1.0089

League Two

EL2

Sample54

Accuracy33.3%

Brier0.6875

Log loss1.1426

La Liga 2

Sample52

Accuracy42.3%

Brier0.6571

Log loss1.1088

Championship

ELC

Sample48

Accuracy37.5%

Brier0.6825

Log loss1.1336

National League

ENL

Sample47

Accuracy61.7%

Brier0.5628

Log loss0.9696

Serie B

BRB

Sample45

Accuracy37.8%

Brier0.6622

Log loss1.1074

League One

EL1

Sample43

Accuracy41.9%

Brier0.6692

Log loss1.1248

Campeonato Brasileiro Série A

BSA

Sample40

Accuracy45.0%

Brier0.6182

Log loss1.0363

J3 League

Sample40

Accuracy45.0%

Brier0.6266

Log loss1.0503

J2 League

Sample40

Accuracy30.0%

Brier0.7235

Log loss1.2182

Serie C - Girone B

SCB

Sample36

Accuracy50.0%

Brier0.6252

Log loss1.0657

National

Sample35

Accuracy42.9%

Brier0.6647

Log loss1.0978

National League - South

ENLS

Sample35

Accuracy57.1%

Brier0.5596

Log loss0.9396

La Liga

Sample35

Accuracy51.4%

Brier0.5322

Log loss0.9127

Serie C - Girone A

Sample34

Accuracy44.1%

Brier0.6693

Log loss1.0976

Serie C - Girone C

SCC

Sample33

Accuracy42.4%

Brier0.6678

Log loss1.1275

Serie D - Girone C

SDC

Sample32

Accuracy31.3%

Brier0.7423

Log loss1.2442

Serie D - Girone H

SDH

Sample31

Accuracy35.5%

Brier0.7107

Log loss1.1927

Serie D - Girone A

SDA

Sample31

Accuracy51.6%

Brier0.6403

Log loss1.0760

National League - North

ENLN

Sample31

Accuracy22.6%

Brier0.7720

Log loss1.2762

3. Liga

BL3

Sample31

Accuracy38.7%

Brier0.7482

Log loss1.2685

UEFA Europa League

UEL

Sample30

Accuracy53.3%

Brier0.6255

Log loss1.0646

Serie D - Girone E

SDE

Sample30

Accuracy46.7%

Brier0.6241

Log loss1.0355

Paulista - U20

BRU20

Sample29

Accuracy62.1%

Brier0.5477

Log loss0.9422

Serie A

Sample29

Accuracy55.2%

Brier0.6088

Log loss1.0676

Serie B

Sample29

Accuracy44.8%

Brier0.6290

Log loss1.0430

Serie D - Girone D

SDD

Sample29

Accuracy51.7%

Brier0.6312

Log loss1.0568

Serie D - Girone F

SDF

Sample29

Accuracy31.0%

Brier0.6837

Log loss1.1475

Serie D - Girone G

SDG

Sample29

Accuracy31.0%

Brier0.7385

Log loss1.2160

Serie D - Girone I

SDI

Sample28

Accuracy39.3%

Brier0.6754

Log loss1.1331

Premier League

Sample27

Accuracy29.6%

Brier0.6921

Log loss1.1634

2. Bundesliga

BL2

Sample27

Accuracy37.0%

Brier0.7354

Log loss1.2366

Ligue 2

FL2

Sample26

Accuracy46.2%

Brier0.6487

Log loss1.0618

Serie D - Girone B

SDB

Sample26

Accuracy34.6%

Brier0.7407

Log loss1.2055

Bundesliga

BL1

Sample25

Accuracy36.0%

Brier0.7828

Log loss1.3221

Ligue 1

FL1

Sample24

Accuracy45.8%

Brier0.7318

Log loss1.2221

2. Division

DEN2

Sample24

Accuracy33.3%

Brier0.7311

Log loss1.1898

1. Division

NOR2

Sample23

Accuracy52.2%

Brier0.5471

Log loss0.9293

Primera División Femenina

ESPW

Sample22

Accuracy54.5%

Brier0.5554

Log loss0.9288

UEFA Conference League

UECL

Sample18

Accuracy50.0%

Brier0.6791

Log loss1.1345

Ykkönen

FIN3

Sample18

Accuracy61.1%

Brier0.5837

Log loss0.9781

Liga Portugal Betclic

PPL

Sample18

Accuracy44.4%

Brier0.5964

Log loss1.0114

UEFA Champions League

UCL

Sample18

Accuracy38.9%

Brier0.7367

Log loss1.2764

Serie C

Sample18

Accuracy55.6%

Brier0.5975

Log loss0.9688

Frauen-Bundesliga

FBL

Sample18

Accuracy38.9%

Brier0.6871

Log loss1.1791

Women's Super League

WSL

Sample17

Accuracy41.2%

Brier0.5235

Log loss0.8760

Veikkausliiga

FIN1

Sample15

Accuracy40.0%

Brier0.6662

Log loss1.1148

Süper Lig

TRSL

Sample15

Accuracy66.7%

Brier0.4661

Log loss0.7874

Ykkösliiga

FIN2

Sample15

Accuracy13.3%

Brier0.8784

Log loss1.4317

Eliteserien

NOR1

Sample14

Accuracy64.3%

Brier0.4651

Log loss0.7962

Serie C

BRC

Sample13

Accuracy23.1%

Brier0.8007

Log loss1.3052

Super League Women

BEW

Sample12

Accuracy41.7%

Brier0.7273

Log loss1.1712

1. Division

DEN1

Sample12

Accuracy58.3%

Brier0.5538

Log loss0.9289

Superliga

DENSL

Sample10

Accuracy30.0%

Brier0.8196

Log loss1.3395

Jupiler Pro League

BE1

Sample9

Accuracy66.7%

Brier0.5490

Log loss0.9607

Kakkonen - Lohko C

FIKC

Sample6

Accuracy33.3%

Brier0.6931

Log loss1.1423

Kakkonen - Lohko B

FIKB

Sample6

Accuracy0.0%

Brier0.9129

Log loss1.6094

Super League

CHSL

Sample6

Accuracy33.3%

Brier0.6488

Log loss1.1057

Kakkonen - Lohko A

FIKA

Sample6

Accuracy16.7%

Brier0.6535

Log loss1.0544

Audit trail

Latest settled predictions.

Each row links back to the match page and uses the frozen pre-kickoff snapshot.

IncorrectLa Liga 2 · 18 Oct 2026

Burgos 1–0 Sabadell

Pick

Draw

Pick %

35%

Confidence

61%

Model

navurex-v1.0-contextual-form

IncorrectSerie B · 18 Oct 2026

São Bernardo 0–1 Criciuma

Pick

Draw

Pick %

36%

Confidence

63%

Model

navurex-v1.0-contextual-form

CorrectCampeonato Brasileiro Série A · 18 Oct 2026

Gremio 2–4 Cruzeiro

Pick

Cruzeiro

Pick %

47%

Confidence

78%

Model

navurex-v1.0-contextual-form

CorrectSerie A · 18 Oct 2026

Juventus 0–3 Lazio

Pick

Lazio

Pick %

53%

Confidence

69%

Model

navurex-v1.0-contextual-form

CorrectLigue 1 · 18 Oct 2026

Lyon 2–1 Nice

Pick

Lyon

Pick %

61%

Confidence

84%

Model

navurex-v1.0-contextual-form

CorrectSerie B · 18 Oct 2026

Sampdoria 1–2 Arezzo

Pick

Arezzo

Pick %

45%

Confidence

61%

Model

navurex-v1.0-contextual-form

Correct3. Liga · 18 Oct 2026

SV Meppen 1–1 Havelse

Pick

Draw

Pick %

40%

Confidence

72%

Model

navurex-v1.0-contextual-form

CorrectEliteserien · 18 Oct 2026

Tromso 1–1 KFUM Oslo

Pick

Draw

Pick %

38%

Confidence

66%

Model

navurex-v1.0-contextual-form

CorrectLa Liga 2 · 18 Oct 2026

Girona 2–0 FC Andorra

Pick

Girona

Pick %

47%

Confidence

75%

Model

navurex-v1.0-contextual-form

CorrectLa Liga · 18 Oct 2026

Malaga 2–5 Real Sociedad

Pick

Real Sociedad

Pick %

52%

Confidence

85%

Model

navurex-v1.0-contextual-form

IncorrectSuperliga · 18 Oct 2026

FC Copenhagen 1–1 Aarhus

Pick

FC Copenhagen

Pick %

77%

Confidence

88%

Model

navurex-v1.0-contextual-form

Correct2. Lig · 18 Oct 2026

Muş Sport Klübü 4–2 Ankaraspor

Pick

Muş Sport Klübü

Pick %

78%

Confidence

88%

Model

navurex-v1.0-contextual-form

Methodology note

Performance is only meaningful when history cannot be rewritten.

Navurex counts only prediction rows that existed before kickoff and were later settled against the final result. Old matches with no saved pre-match snapshot are excluded instead of being reconstructed after the result is known.