One ledger per voice, down to the timestamp.
A granular ledger on each influencer: every claim and prediction on file, dated, timestamped, sourced and scored.
The accountability index
Viral Influencers.Predictions Tracked
Vindict is the only platform that indexes and tracks the predictions of political influencers. See who's actually worth listening to.
Vindict LIVE INDEX · 12 JUL 2026Who we keep score on.
The loudest forecasters online: indexed, dated, and scored against the public record.
On the record.
In real time.
New claims come in around the clock, and every gradable one is logged the moment it's made.
INDEXED CLAIMS · ALL TIME12,841
ALL-TIME PREDICTIONS3,594
RESOLVED PREDICTIONS2,871
EXTRACTED PREDICTIONSLIVE
VOICEEXTRACTED PREDICTIONDEADLINESOURCE
37% of Americans under 30 get their news from influencers, and 52% of 18–24s worldwide say social media and video are their main source.
Pew Research Center · Reuters Institute Digital News Report 2026
Why these voices
- 01Nick Fuentes
- 02Hasan Piker
- 03Kyle Kulinski
- 04Glenn Greenwald
- 05Tucker Carlson
Average views per video
24-month window45% more attention per upload329K vs 227K average views per video
The five voices328K
CNN · Fox News · MS NOW155K
2.12×more attention
per upload
per upload
Current public views on videos published July 2024 – June 2026, measured July 2026, via the YouTube Data API and public Rumble data. Publication-cohort attention, not unique viewers.
Zoom in. Zoom out.
“Democrats retake the House in November.”
Left consensus72%
Right consensus31%
Centrist consensus55%
Top-decile credibility71% yes
Whole field, unweighted58% yes
30-day momentum+6 toward yes
Sample142 voices · 511 claims
Accuracy earns weight. Reach doesn't. Each voice counts in proportion to its verified record.
The whole field as one weighted signal.
Zoom out and the ledgers become one signal: credibility-weighted sentiment across every tracked voice, and how it moves.
0102030405060708
Top tracked human
Claude Fable 5
GPT-5.6 Sol
Grok 4.5
Claude Opus 4.8
Gemini 3.1 Pro
DeepSeek V4 Pro
Qwen 3.7 Plus
Questionsfrozen at ask
Toolsnone · no retrieval
Fair testresolves past cutoff
Runs5 per model · mean kept
Scored on Brier, not vibes. Confident and right beats a lucky coin toss. A pure coin-flipper scores .25.
Frontier models, same test as the humans.
Frontier models forecast the same questions the humans argue over. Each answer is locked in before the outcome, then scored once the result is in.
