NRL | Round 22

alphr.com.au

SGI
Dragons
VS
CAN
Raiders
WIN STADIUM, WOLLONGONG • SATURDAY 2 AUG, 5:30 PM

Win Probability

AI Game Review

Dragons defied the model's 66% prediction for Raiders, a notable upset. The predicted margin of 4.1 was reasonable against the actual 6-point result. The game's 30 points came in 17 points lower than the predicted 47. A tough result for the model, all 3 picks missed on this one.

Model vs actual outcomes • Post-match analysis

🏁

AI Referee Insights

Todd Smith officiated this match (105 career games). The combined score of 30 points was 16 points below Todd Smith's career average of 46. Dragons bucked the trend, Raiders historically win 47% of games under Todd Smith, but couldn't convert that edge today.

Based on referee career statistics • Post-match analysis

Momentum Replay
Beta
80', Dragons firmly in control (99%)
DRA18
99%80'1%
12RAI
HT100%50%0%0'20'40'60'80'
Raiders momentumMomentum -3Dragons momentum →
Next Try (within 10 min)
AI Model
89% none
DRA 5%No try 89%RAI 6%
Biggest Swings

AI Win Probability

66%RaidersFavourite

Dragons

34%

Raiders

66%

AI Match Overview

Raiders are clear favourites here at 66%, with our model expecting a comfortable victory over Dragons. The model sees Raiders ahead on 3 of 7 key factors including ELO Difference, Halves Control and Recent Win Rate. Raiders carry a 337-point ELO rating advantage (1678 vs 1342). Recent form favours Raiders with 5 wins from their last 5 compared to 1 for Dragons. The margin model predicts Raiders by 4.1 points with a combined total of 47.

Generated from model features • Pre-kick-off analysis

Edge Analysis

1 ACTIVE EDGE

Each market is predicted by an independent model, H2H, margin, and totals may occasionally disagree.

H2H Recommendation

Raiders to Win @1.45

Lost ✗

Edge

-5.3%

Line / Spread

Raiders +6.5 @1.91

Lost ✗

Edge

-5.3%

Total Points

Over 43.5 @1.91

Lost ✗

Edge

+9.6%

Form & History

TeamLast 5Avg Pts
Dragons
W
L
L
L
L
26.4
Raiders
W
W
W
W
W
30.0

Avg Conceded

27.4

Dragons

17.6

Raiders

Avg Margin

-1.0

Dragons

12.4

Raiders

Run Metres

1705

Dragons

1813

Raiders

Line Breaks

4.8

Dragons

6.6

Raiders

Referee Indicator

Favours Raiders

Todd Smith

105 career games · since 2019

AI Analysis

Win rate when Todd Smith refs each team (vs any opponent)

Dragons
2W – 11L
15%
Raiders
7W – 8L
47%

When Todd Smith officiates, Raiders have won 7 of 15 games (47%), significantly stronger than Dragons's 2 from 13 (15%). Games average 46.1 pts, above the league norm, suggesting a free-flowing style.

Avg Total

46.1 pts

Home Win %

53%

Home Bias

Leans home

Penalty & Discipline

Pen / Game

10.7

Sin Bins / Gm

0.50

SB Away %

52%

Avg Penalties Per Game

vs Home Teams5.3
vs Away Teams5.4

Penalty Advantage Under This Ref

Positive = opponent penalised more than your team

Dragons
-0.4
Raiders
-0.5
H2H History (Last 5)Raiders lead 5-0
Jul 2025SGI 24 - 28 CAN
Sep 2024SGI 24 - 26 CAN
Jul 2023SGI 26 - 36 CAN
Apr 2023SGI 14 - 20 CAN
Aug 2022SGI 22 - 24 CAN
Prediction BreakdownPure Alpha Model

📊Team ELO Ratings

SGI
1342Overall1678
CAN
ELO difference: -337 in favour of Raiders

🏈Positional Matchups

Player ELO aggregated by position group, higher = stronger unit

1134Forwards1121
Best: 1281SGI +13Best: 1313
1157Backs1129
Best: 1197SGI +28Best: 1187
1109Halves1269
Best: 1109CAN +160Best: 1269
1060Hooker971
SGI +89

📈Recent Form (Last 5)

SGI
Stat
CAN
1.0
Wins (Last 5)
5.0
26.4pts
Avg Score
30.0pts
27.4pts
Avg Conceded
17.6pts
-1.0pts
Avg Margin
12.4pts
1704.8m
Run Metres
1812.8m
4.8
Line Breaks
6.6
332.6
Tackles
338.8
8.8
Errors
9.8

🔑Key Prediction Factors

What the model weighted most in this prediction

1
ELO Difference14.0%
Raiders
2
Forward Pack12.0%
3
Backline Quality10.0%
Dragons
4
Halves Control9.0%
Raiders
5
Recent Win Rate9.0%
Raiders
6
Referee Tendency7.0%
7
Venue Advantage7.0%
Dragons

Model Confidence

66%

Raiders predicted to win by 4 points

Predicted total: 47 · Line: -4.1

0/3 match predictions correct
Coming Soon

Try Scorer Predictions

AI-powered try scorer predictions and player prop markets, built on our 6-model player stats engine.

First / Anytime / Last ScorerPlayer Props