Football predictions weekly review: when the draw was enough
Agreement reached 76.5%. But Radomlje, Hereford and Koper showed why the exact tip matters more than the size of its backing.
Read the week’s findingsSupport was broad. The evidence was more selective.
Our football predictions weekly review for 26 October–1 November 2026 follows the votes on 187 predictions, then checks the eligible tips against recorded final scores.
The clearest success story was a draw. Radomlje’s 1–1 finish against Mura was enough for the home-or-draw recommendation, which held seven Agree votes against one Disagree. Elsewhere, Hereford’s straight home-win tip failed despite four votes in favour, while Koper’s home-or-draw selection succeeded with only one of its three votes backing it. Those results capture the week’s central distinction: agreement describes an opinion about a specific market, and that market sets the test.
Across the export, 231 of 302 retained vote positions were Agree, a 76.5% share. A stricter outcome sample produced 100 correct tips from 160 assessed, or 62.5%. These figures answer different questions. One measures support; the other measures how often the recorded selections matched the final results. Their numerical gap is not a measure of how many voters were wrong.
Reading this edition: voting activity covers 26 October–1 November in UTC. Agreement figures are the latest retained positions in the export, not a reconstructed Sunday-night or pre-kickoff balance. Most samples are small: 120 of the 187 predictions have just one vote. See the calculation rules.
Radomlje did not have to win
The most-voted prediction in the screened outcome sample showed exactly what draw protection means.
A 1–1 draw was a successful selection
Radomlje’s tip covered two outcomes. The recorded probabilities were 46% for a home win and 28% for a draw, adding to the 74% confidence shown for 1X. Calling this simply a “Radomlje win prediction” would change what readers were asked to judge.
THE MARKET DEFINES THE RESULTThe seven-to-one split gives Radomlje the strongest voting support by volume among the records retained for our outcome analysis. It does not give us eight independent forecasts about a home victory. The vote concerned home or draw, and the final score fell within that selection. The supplied predicted score was also 1–1, but exact-score accuracy is a separate test and is not used in this report’s tip success rate.
The most-discussed record overall was Nea Salamis–Pafos, with nine Agree votes and one Disagree. Its original probability metadata was later enriched, so it remains part of the participation figures but is excluded from the stricter outcome comparison. Keeping those two populations separate prevents the biggest voting story from automatically becoming the strongest performance evidence.
100 correct tips, with a clear denominator
Of the 187 voted predictions, 160 met this edition’s result and metadata checks. The other 27 remain in participation totals but do not enter the screened success rate.
Ten assessed predictions had tied voting totals. They count in the overall tip result, but not in the majority comparison alongside it.
How retained votes lined up
150 clear majoritiesCounts of predictions, not votes. Bar lengths are relative to the largest group. Ten tied cases are excluded here.
Tips with more Agree than Disagree positions succeeded in 76 of 116 cases: 65.5%. Among tips with more Disagree positions, 20 of 34 still succeeded: 58.8%. The difference was 6.7 percentage points, but the underlying groups were unequal and often rested on a single vote. This is a description of the exported records, not evidence that following a voting majority would improve future predictions.
The tied group adds four correct and six incorrect tips, bringing the overall result back to 100 out of 160. A rejected tip that fails can support a voter’s objection, but it does not identify the alternative they expected. Equally, an accepted tip that succeeds does not tell us whether the person voting correctly anticipated the precise score.
Vote volume adds another wrinkle. The 20 assessed predictions with at least three retained votes produced seven correct tips, or 35.0%. That small subset performed below the overall sample, even though it includes Radomlje’s successful selection. Attention and success did not move together consistently this week; these 20 cases are too few, and too selectively voted on, to establish a general relationship.
Home-or-draw led this week’s market breakdown
The overall percentage combines straight-win selections with double-chance selections. Reading them separately makes the outcome analysis more useful.
| Recorded selection | Correct / assessed | Success rate |
|---|---|---|
| Home win (1) | 32 / 60 | 53.3% |
| Away win (2) | 12 / 18 | 66.7% |
| Home or draw (1X) | 41 / 52 | 78.8% |
| Away or draw (X2) | 15 / 30 | 50.0% |
Home-or-draw selections accounted for 41 correct results from 52 assessed tips, or 78.8%. Away-or-draw selections finished at 15 from 30, or 50.0%. Both cover a draw, but the fixtures and teams behind them were different. The contrast describes this week’s sample; it does not show that one side of double chance is inherently superior.
Straight home-win tips succeeded in 32 of 60 cases, while straight away-win tips returned 12 successes from 18. Radomlje illustrates why these categories cannot be swapped casually: its draw was enough for 1X and would have failed a straight home-win selection. No odds or stakes are supplied, so these success rates cannot establish profit, return on investment or betting value.
Hereford and Koper pulled in opposite directions
Two small voting samples produced very different relationships between support and the final score.
A clean sweep of votes could not rescue Hereford
All four retained votes backed the home-win tip. Its recorded confidence was 80%, yet AFC Telford United won 1–0. Hereford’s own match report confirms the result, giving this selection a straightforward outcome: the predicted home win did not happen.
The 100% agreement figure is accurate for four positions, but it should be read with that denominator attached. This was unanimity in a small sample. It was neither a large consensus nor a guarantee that an 80%-confidence recommendation would succeed.
Koper delivered beyond the minimum requirement
Koper’s home-or-draw tip held one Agree vote against two Disagree votes. The 4–0 win over Maribor satisfied the selection through a home victory; it did not need the draw protection on this occasion.
This is the clearest multi-vote example in the screened sample of an opposed tip succeeding. The two dissenting positions do not tell us whether those voters expected a draw, an away victory or simply disliked the recommendation. Their recorded objection and the tip’s outcome can be compared without inventing that missing detail.
West Ham’s home-win tip divided four votes
A two-to-two balance leaves no majority to credit or blame. The result still gives the underlying prediction a clear outcome.
Two agreed. Two disagreed. West Brom won.
West Ham’s home-win recommendation carried 74% recorded confidence. The 1–3 result meant the tip failed, while the vote remained evenly divided. This fixture is an illustration of a split opinion, not a statistically established controversy.
The export’s formal controversy measure requires at least ten votes. Only Nea Salamis–Pafos reached that threshold, and its split was 9–1. There is no high-volume, evenly divided fixture to present as this week’s “most controversial” match.
Friday carried the most voting activity
The busiest day was 30 October, with 84 actions. That was 23.5% of the week’s total and 25 more actions than Saturday.
Daily voting actions
UTC · 26 Oct–1 NovIncludes new votes and switches. These bars do not count unique people.
Where votes accumulated
Top competition groups by retained vote positions. These are balances, not the action counts in the daily chart.
New votes accounted for 296 of the 357 actions; 61 were switches, or 17.1%. Those changes were almost balanced: 29 moved towards Agree and 32 towards Disagree. That is a net three-position movement towards Disagree from switches alone, not a collapse in support. The action count also cannot tell us how many distinct people changed their minds.
Saturday contained the busiest individual hour, 18:00–18:59 UTC, with 23 actions, even though Friday led across the full day. Saturday and Sunday together produced 92 actions, or 25.8% of the weekly total. The pattern supports a Friday peak in this export; without kickoff times, it cannot explain whether readers were reacting to team news, live matches or completed results.
Cyprus’s 1. Division led the retained competition balances with 21 votes across five predictions, followed by Slovenia’s 1. SNL with 15 across three. Nea Salamis–Pafos supplied ten of the Cypriot group’s 21 positions, while Radomlje–Mura supplied eight of Slovenia’s 15. A single fixture therefore accounts for a substantial share of each leading group. These totals measure engagement with the available predictions, not the relative popularity of whole leagues.
The same support can end in different scorelines
These fixtures broaden the review beyond the lead stories. Each card shows its actual vote count alongside the percentage.
Skanste vs Metta / LU
Four Agree votes against one Disagree supported the away-win tip. Metta / LU’s recorded 2–1 victory satisfied that straight-win selection. The 80% agreement share describes five positions, while the model’s separate confidence figure was 79%.
Arminia Bielefeld vs Eintracht Braunschweig
Arminia’s home-or-draw tip also held a four-to-one split. A 3–1 win over Eintracht Braunschweig made the draw cover unnecessary. Its 77% confidence and 80% vote agreement were close numerically, but they remained different measures.
Hide 2 additional matches
Karlsruher SC vs 1. FC Heidenheim
Three of four retained votes backed Heidenheim to avoid defeat. Karlsruher’s 2–1 home win fell outside both outcomes covered by X2. This was a double-chance miss despite 78% recorded confidence and 75% agreement.
Hødd vs Bryne
Hødd’s home-or-draw tip also had three Agree votes and one Disagree. Bryne’s 3–0 away win made it incorrect. Compared with Radomlje, the lesson is precise: draw cover helps only when the result is a draw; it cannot protect against defeat.
Read the selection before reading the percentage.
This week’s useful distinction sits inside the tip itself. A draw delivered for Radomlje, a home win delivered for Koper, and neither Hereford’s confidence figure nor its unanimous four-vote backing changed the conditions needed for its selection to succeed. The 1X group’s 41 successes from 52 assessments is the strongest market result in this edition, but it belongs to this particular set of fixtures.
The broader voting picture is encouraging as participation: 357 actions spread across 187 predictions and 98 competition groups. It is still thin evidence about collective forecasting ability. Future editions will be more informative if larger vote samples and confirmed pre-kickoff balances become available. For this week, the honest takeaway is a well-defined comparison of recorded opinions and results, with the small samples left visible.
What do you make of the next tip?
Check the suggested market, compare its confidence with the community vote, and add your own Agree or Disagree position.
Explore today’s football predictionsMethodology, exclusions and result sources
Reporting window and voting units
The activity window is 26 October 2026 at 00:00 UTC to 2 November 2026 at 00:00 UTC, excluding the endpoint. The original export was generated on 2 November at 07:44:53 UTC; its result enrichment is timestamped 09:10:17 UTC. This is an activity-based selection of predictions, not a list restricted to matches played within the week. Three fixtures dated 2 November received votes during the window and remain in participation totals.
A retained position is the latest recorded vote per browser token on a prediction. One person may vote on multiple predictions or use more than one browser, so neither positions nor actions count unique people. The 302 positions are current retained balances; the 357 actions are submissions and switches recorded during the week. The two totals need not match, and retained balances are not guaranteed to represent the week-end cutoff.
How tip outcomes are assessed
We compare the supplied full-time scores with the recorded market: 1 requires a home win; 2 requires an away win; 1X accepts a home win or draw; X2 accepts an away win or draw. Predicted exact scores are not used for settlement. No other betting markets appear in this export. These are result-based tip checks, not bookmaker settlement or a calculation of returns.
The screened sample requires a finished result, complete original metadata, and no flagged date or identity mismatch. It contains 160 predictions: 100 correct and 60 incorrect. The 27 excluded records comprise 19 with later-enriched original metadata, two unresolved team identities, two postponed matches, three not yet completed, and one additional date-mismatched record. One of the 19 enriched records also has a date mismatch; it is counted only once.
For transparency, a looser calculation using all 180 records marked finished would return 111 correct tips, or 61.7%. We use 100/160 throughout the headline analysis because the looser set includes records with unresolved original-metadata or date questions. The export’s own outcome-evaluated flag is false: the outcome calculations in this article were performed for this report.
What the majority comparison can establish
Majorities use exact vote counts, not rounded percentages. Ten tied predictions are removed from the majority table but retained in the 160-tip result. All kickoff timestamps are absent, so no pre-match voting cutoff can be verified. The comparison is retrospective and must not be presented as a backtest of a crowd betting strategy. No odds, stakes or unique-person counts are available.
Sources and selection
Voting and prediction figures come from the supplied TopBettingSites.ng weekly export. Scores come from its external result enrichment. Links beside the eight featured fixtures identify the supplied result sources. Spot-checks confirmed the eight featured scorelines against accessible source material. This is not an independent re-verification of all 180 finished records.
The eight featured matches were selected to illustrate supported, opposed and tied votes, both successful and unsuccessful tips, and the difference between straight-win and double-chance markets. They are editorial examples, not a random sample; the aggregate figures use the full stated populations. Percentages are rounded to one decimal place. Competition groups combine country and tournament so similarly named leagues remain separate.
