Reporters nail some beats and miss others.
That's why we score per topic.
Look up something specific — "Manchester United's next transfer target", "Apple's foldable iPhone window", "who Marvel recasts as Iron Man", "Lakers' trade-deadline plan". Rumora pulls every rumor on that exact beat and ranks the journalists by who's actually been right about it. Aggregate accuracy scores are useless. Per-topic accuracy is the only thing that matters.
Currently in closed testing on Google Play · iOS coming soon
In the app
Type any topic. Get a feed of rumors graded by who broke them. Tap any rumor for the full scoring breakdown.
Any topic, any beat
Type Mbappé, iPhone leaks, Marvel casting — anything. We build the feed.
Your rumours
A graded feed for every topic you follow. Medals show how strong each rumour is.
Why this medal
Every rumour shows the journalist's accuracy on that exact topic, plus the full reasoning behind its score.
Real data, across every beat
Sample scores from across the database. Sports, tech, entertainment, politics, finance — same scoring model, different journalists, different beats.
- 91% on Real Madrid transfers n=200
- 72% on Arsenal transfers n=85
- Untested on UCL predictions
- 89% on iPhone leaks n=120
- 76% on Vision roadmap n=40
- Provisional on Apple AI n=8
- 87% on NBA trades n=140
- 73% on Lakers roster moves n=55
- 58% on free agency predictions n=30
- 84% on studio greenlights n=92
- 70% on Marvel casting n=38
- Provisional on Netflix M&A n=6
- 88% on White House staffing n=110
- 67% on cabinet picks n=42
- Untested on policy timelines
- 86% on Big Tech M&A n=78
- 71% on private equity deals n=44
- 54% on IPO timing n=25
Sample data shown above. Real scores update automatically as outcomes resolve — provisional ratings (sample size < 10) are marked so you know what we don't yet know.
How Rumora works
Ingest
We pull rumors continuously from journalists across RSS, Bluesky, Reddit, and verified social feeds. Every rumor is tagged to a topic and a journalist.
Score
Each rumor gets a gold, silver, or bronze medal based on five inputs: journalist track record on that topic, language strength, outlet tier, specificity, and recency.
Verify
Rumors older than 24 hours go through outcome verification — we check whether they actually came true. Confirmed, denied, or collapsed.
Update
Every verified outcome updates the journalist's accuracy score on that topic. Trends become visible over weeks. Reputations are earned, not assumed.
Know the moment it breaks on your beat
Gold-medal rumour on a topic you follow? You get a push and an email the moment we score it — no refresh required.
Push notification
The instant we ingest and score a gold-medal rumour on a topic you follow, your phone buzzes. Tap straight into the rumour and the journalist's track record.
Email digest
Prefer your inbox? Same gold-medal trigger, delivered as a clean email with the headline, the source, and the journalist's score on that exact beat.
Per-topic, not per-outlet
Follow "Real Madrid transfers" — only get pinged on that. Not every football story. Not every Romano tweet. Only the ones graded gold on the beat you actually care about.
Push + email are both opt-in. Free for everyone in the beta. Silver and bronze never page you — only verified gold-tier rumours from journalists with track record on the topic.
Why aggregate journalist ratings are wrong
What other sites tell you
Source: Muck Rack reach score · NewsGuard outlet tier
Doesn't tell you whether he's right on the topic you care about.
What Rumora tells you
- 91% on Real Madrid transfers n=200
- 72% on Arsenal transfers n=85
- 54% on Champions League predictions n=18
- Provisional on Asian football n=4
Pick the beat you care about. We tell you whether to trust him on that specific one.
The methodology, in plain English
Per-topic Beta-binomial scoring
Each journalist starts with a topic-specific prior — usually 0.55 for unknowns, higher for established beats. As outcomes resolve, we blend the empirical accuracy with the prior using weight = n / (n + 10). Twenty outcomes is roughly two-thirds data, one-third prior.
Recency-weighted outcomes
Outcomes from this month count more than outcomes from 2024. A journalist who was great five years ago and quiet since shouldn't outrank an active beat reporter today.
Verifiability classifier
Some topics resolve cleanly (transfers, product launches, M&A). Some don't (vague political predictions, "interest" without action). We only update scores on topics with hard outcomes.
Outcome verification by web search
Every 24-hour-old rumor goes through automated verification. Our AI runs two independent web searches and only marks the outcome resolved when both agree.
Originator vs aggregator credit
When MacRumors cites Mark Gurman, only Gurman's score moves on outcome. The aggregator gets a separate "curator" score for whose work they choose to amplify.
Provisional flag
Scores with sample sizes under 10 are flagged as provisional. We'd rather show you "we don't know yet" than pretend.
What the database makes possible
- Football fans can finally answer "is this transfer rumor real?" by checking the breaking journalist's score on that exact club.
- Apple watchers can compare Gurman, Kuo, MacRumors, and 9to5Mac side-by-side on the specific product line they care about.
- Beat reporters can see their own track record per topic and benchmark against peers.
- News editors can decide whether to amplify a rumor based on the breaking journalist's predictive record on that beat.
- Researchers can cite per-topic accuracy data in media-trust studies.
Join the beta
Rumora is in closed testing on Google Play. Drop the Gmail you use on your Android phone — we'll add you to the tester list and send the install link within 24 hours.
iOS version coming after Android stabilizes. The same form gets you notified when TestFlight opens.
Frequently asked questions
How does Rumora score journalists?
We ingest rumors from journalists across RSS, Bluesky, and Reddit, then automatically verify outcomes through web search. A journalist's score on a topic is their confirmed-rumor percentage, blended with a prior using a Beta-binomial model so small samples don't swing wildly. As more outcomes resolve, the score converges on the empirical accuracy.
Why is per-topic accuracy better than overall accuracy?
A reporter who's 91% accurate on football transfers might be 60% accurate on Champions League predictions. Aggregating these into a single "overall" score hides the signal that actually matters to readers: which beat is this person credible on? Every other journalist-rating service we know of either averages across all topics (Muck Rack) or scores outlets instead of individual journalists (NewsGuard). Rumora is the first to track (journalist × beat) as the canonical scoring unit.
What does a gold, silver, or bronze medal mean?
Gold means a rumor scored 0.85 or higher on our composite scale (journalist track record × language strength × outlet tier × specificity × recency). Silver is 0.65 to 0.84. Bronze is 0.45 to 0.64. Speculative rumors below 0.45 are still visible in the feed but unranked. Tapping any rumor shows the full breakdown — why this medal, with the actual weights.
How is this different from Muck Rack or NewsGuard?
Muck Rack tracks journalist output and impressions — useful for PR pitching, not for predicting who's right. NewsGuard rates outlets (CNN vs InfoWars), not individual journalists. Neither scores journalists by predictive accuracy on specific beats. Rumora is the only database that gives you per-(journalist, topic) accuracy backed by verified outcomes — which is the only thing that matters when you're trying to decide whether to believe a breaking rumor.
How does outcome verification work?
Every rumor older than 24 hours goes through automated outcome verification. Our AI runs two independent web searches — one neutral, one adversarial ("argue this didn't happen") — and only marks an outcome resolved when both agree. Disagreements go into a manual review queue. This dramatically reduces false-positive resolutions.
Do I get notified when a gold rumour breaks on my beat?
Yes — push notification and email, both opt-in, both free in the beta. The trigger fires only when a rumour is graded gold on a topic you follow, scored by a journalist with track record on that exact beat. Silver and bronze never page you. You can turn each channel off independently in Settings → Notifications.
What topics does Rumora cover?
Anything where rumors break before facts. Football transfers, NBA trades, NFL QB carousel, F1 driver markets, iPhone leaks, Apple supply chain, Marvel casting, Hollywood greenlights, Netflix M&A, OpenAI acquisitions, Big Tech antitrust, White House staffing, election strategy, private equity deals, IPO timing — and any custom beat you type in. The app builds a topic-specific feed on demand and we start tracking which journalists are credible on that exact subject. The nichest beats often have the clearest accuracy signals because the journalist pool is smaller.
Can journalists dispute their scores?
Yes. Every journalist profile has a dispute submission form. We publish our methodology, our outcome verifications, and our scoring algorithm transparently — disputes go through a structured review process and resolved disputes show as annotations on the profile. We'd rather be corrected than wrong.
When is Rumora launching publicly?
We're in closed testing on Google Play right now. Public Android launch is targeted within the next 30 days, once we have 12+ active testers using the app for the 14-day Google qualification period. iOS launch follows shortly after.