Transparency
Model Performance
Every published prediction, tracked from publication to final result. Sample history — demo data
Tracking policy: every prediction Vextiq publishes is logged when it’s published and tracked to its eventual result: pending, win, loss or push. Losing predictions are never deleted, hidden or excluded from these numbers, and graded results can’t be edited. Predictions for cancelled games are marked void and shown separately. Hit rate = wins ÷ (wins + losses).
Showing: College Basketball · Season · Prop · 60–69% confidence
Published
2
2 sample
Wins
1
Losses
0
All included
Pushes
0
Pending
1
Awaiting final result
Hit rate
100%
Wins ÷ (wins + losses)
Daily results & cumulative hit rate
SampleBy prediction type
SamplePending: published, awaiting result (1)
Sample| Published | Game time | Sport | Matchup | Type | Confidence | Status |
|---|---|---|---|---|---|---|
| Sep 27, 2026 | Mon, Sep 28, 5:00 PM CDT | CBB | Duke Blue Devils vs UConn Huskies | PROP | 61% | PENDING |
Picks stay hidden here until game time so they don’t undercut members’ access. Each one is graded here once its game is final.
Graded predictions (most recent)
Sample| Graded | Sport | Matchup | Type | Pick | Confidence | Result |
|---|---|---|---|---|---|---|
| Aug 21, 2026 | CBB | Duke Blue Devils vs Kansas Jayhawks | PROP | Player prop (sample) — Points | 66% | WIN |
Past model performance does not guarantee future performance. Sports outcomes are unpredictable. The seeded history is SAMPLE data for demonstration and doesn’t reflect real-world results. Grading from a live data provider is prepared (lib/grading.ts) but not yet connected.
ⓘ Confidence percentages are model confidence and do not guarantee an outcome.