Sunday's Prediction Performance Analysis

Sunday, 9 August 2026 delivered a full programme of football action across 161 fixtures, presenting predictors with considerable challenges and opportunities in equal measure. The day served as a comprehensive test of forecast accuracy across multiple betting markets, with varied results emerging from the breadth of matches on offer. The volume of fixtures ensured that patterns and trends could be observed across different competitions and regions throughout the course of the day.

Across the day's predictions, the 1X2 market returned a 56% success rate, translating to 90 correct outcomes from 161 attempts. The Over/Under market performed at an identical 56% clip, with 86 successful predictions from 153 applicable fixtures. The BTTS market proved more elusive, with a 47% success rate marking 76 correct predictions from the same 161 fixtures. These figures collectively indicate a day where match outcome predictions aligned closely with goal-based markets, while the more specific both-teams-scoring market presented greater difficulty for those navigating the predictions landscape.

Prediction Accuracy: A Candid Review of Our 161 Selections

The numbers tell a story of consistency with room for meaningful improvement. Across 161 matches in the sample period, our highest-confidence picks delivered a 56% success rate on both full-time result predictions and the Over/Under market. At face value, this represents solid performance relative to typical bookmaker vig, which typically requires above 52-54% accuracy just to break even. However, the Both Teams To Score market tells a different tale, with our selections converting at just 47% across 161 attempts, suggesting this particular market poses greater challenges when narrowing down to a single recommended option.

The disparity between the 56% figures and the BTTS performance warrants examination. When forcing a single definitive pick rather than offering range-based probabilities, the margin for error shrinks considerably. Full-time results benefit from clearer structural patterns, while BTTS outcomes often hinge on fine margins such as a late goal being scored or a defensive error proving decisive. The 56% accuracy across the majority market suggests our core methodology holds value, though the gap to the 47% BTTS rate indicates certain prediction categories deserve more nuanced approaches or perhaps broader coverage rather than singular recommendations.

What stands out is the sample size itself: 161 matches provides statistical significance that lends weight to these percentages. The data confirms that no prediction methodology achieves perfection, and honest evaluation must acknowledge both the demonstrated competence and the demonstrated limitations. Moving forward, the BTTS category clearly demands methodological refinement or adjusted confidence thresholds to bring performance in line with the stronger 1X2 and Over/Under metrics.

Five Predictions That Hit the Mark

The standout success of the weekend came from FC Porto's comfortable 2-0 victory over Alverca. The model assigned a 79% probability to a Home Win, making it one of the strongest single-outcome calls across the full slate. The gulf in class between the two sides was always likely to assert itself, and Porto delivered exactly that outcome in a match where they controlled proceedings from the opening whistle. At such a high confidence rating, this represented textbook value extraction from a lopsided matchup.

Two other high-confidence calls reinforced the model's ability to identify dominant away performances. Celtic's 5-1 demolition of Kilmarnock carried a 69% Away Win probability, and the scoreline reflected genuine attacking superiority rather than fortunate variance. Similarly, Red Bull Salzburg's 1-1 draw against Wolfsberger AC aligned with a 51% Away Win reading — a result that proved the model correctly sensed Salzburg's competitive edge even when they were unable to convert dominance into all three points. These matches demonstrated solid calibration across different confidence tiers.

The most analytically interesting call was the 0-0 draw between Feirense and Felgueiras 1932, which the model backed at only 29% — the lowest probability among the correct calls listed here. This was not a confident prediction by any measure, but it reflected sound reasoning: two attacking-minded sides canceling each other out is a recurring pattern when defensive organisation holds firm. Gimnasia M.'s 1-0 defeat to Instituto Cordoba also registered cleanly at 60%, suggesting the model picked up contextual factors — likely squad availability or home advantage — that tipped the balance correctly. Across all five matches, the common thread was disciplined adherence to structural evaluation, resisting the temptation to overreach on lower-confidence options.

Where the Model Stumbled: Analyzing the Largest Prediction Errors

The most striking failures this round came from matches where probability models placed overwhelming faith in the home side, only to see them stumble. Benfica's 2-2 draw against Academico Viseu stands out as the most consequential miss, with the model assigning an 86% chance of a home victory. The data clearly underestimated the visiting side's resilience, producing a scoreline that defied the heavy favorite status. Similarly, Zenit's home loss to Rodina Moskva at an 82% predicted probability represents a significant calibration failure that warrants deeper examination of how domestic cup competitions may introduce volatility not captured by league-based models.

The Scandinavian fixtures proved equally problematic. Malmo FF's home defeat to Degerfors IF, despite a 65% favorability rating, suggests potential fatigue or squad rotation effects that the model failed to price in. Heerenveen's victory over Twente presents an interesting case where the away side was marginally preferred at just 44%, yet lost 1-0 at home — a reminder that low-confidence predictions carry inherent uncertainty but still require accountability when they miss. FBC Melgar's 4-2 home loss to FC Cajamarca, despite 73% implied probability, indicates that high-scoring encounters in South American competitions may possess characteristics that resist straightforward probability estimation.

Collectively, these misses share a common thread: the model overweighted historical reputation and league positioning while undervaluing opponent-specific motivation and situational factors. Cup competitions appear particularly vulnerable to these blind spots, as teams often field rotated lineups or approach matches with different strategic priorities than league fixtures. The data suggests a need to incorporate competition-specific parameters and situational variables — such as fixture congestion or squad depth — to improve calibration on these high-variance matchups where upsets prove more prevalent than standard models predict.

Domestic Leagues Across Europe Produce Mixed Outcomes for 1X2 Selections

The Portuguese Primeira Liga delivered mixed fortunes, with FC Porto and GIL Vicente justifying favouritism with convincing home victories over Alverca and Rio Ave respectively. However, the round produced two notable upsets, as Moreirense and Benfica were both held to draws against SC Braga and Academico Viseu respectively, costing predictions that had backed the home sides. Over in the Netherlands, the Eredivisie proved a reliable hunting ground for correct predictions, with away wins for Feyenoord at Sparta Rotterdam and Ajax at PEC Zwolle landing comfortably, while Groningen upset Utrecht on home soil. The sole disappointment came at the Abe Lenstra Stadion, where Heerenveen defeated Twente against the odds.

The Scottish Premiership witnessed another dominant display from Celtic, who swept aside Kilmarnock with a commanding five-goal victory. Hearts also recorded a thumping home win against Dundee United, but Rangers suffered a surprise home defeat to Hibernian. Motherwell's goalless draw with Falkirk was another result that eluded those backing the home side. In Belgium's Pro League, Gent, Charleroi and Anderlecht all secured home victories as expected, though Genk's defeat at Zulte Waregem demonstrated the unpredictability that can undermine even well-researched selections. The Portuguese Segunda Liga rounded out the action, with Vizela and Feirense delivering for their backers, while Farense and AVS both defied predictions to claim victories.

Final Thoughts on Sunday's Performance

Sunday, 9 August 2026 brought 161 matches under the microscope, testing the predictive models across various leagues and competitions. The 1X2 accuracy settled at 56%, a result that highlights both the inherent unpredictability of football and the value of statistical analysis in navigating complex matchdays.

As new fixtures emerge, these metrics serve as a reminder that consistency in prediction markets requires patience, discipline, and continuous refinement of methodology. The foundation built on Sunday's data will inform approach and strategy moving forward.