Prediction claims,
with receipts.
Influencer predictions separated by evidence quality, timing and market definition. Qualifying retrospective embedded-media claims are included with their provenance label, while independently verified pre-event captures remain easy to distinguish.
Can an influencer manipulate an old-looking prediction?
Yes. Captions can be edited, posts can be archived or made public later, losing calls can be deleted, and multiple outcomes can be posted. The uploaded video is not normally replaceable inside the same post—but an old Instagram timestamp alone still cannot prove it was public before kickoff.
Want the frozen research packet behind the public ledger?
Shimoverse Signals offers three signed, immutable World Cup forecast packets through an isolated x402 storefront. It is currently a Base Sepolia testnet validation—test tokens have no monetary value and real-money sales remain disabled.
A model that rewards proven records.
Astrology keeps a fixed 50% method allocation. The other 50% is shared equally by eligible non-astrology creators at 75% or better; if every eligible record is exactly 75%, their equal historical-accuracy share is used.
France vs Spain
3 sources received 0% See why
Big SharkzCurrent call is Grade X / conflicted.
Jacky the TurtleNo comparable historical accuracy is available.
MeshmeshHistorical accuracy is below 75%.
England vs Argentina
3 sources received 0% See why
Jacky the TurtleNo comparable historical accuracy is available.
MeshmeshHistorical accuracy is below 75%.
Sujay Nadkarni / The Nad CoderNo comparable historical accuracy is available.
World Cup champion
2 sources received 0% See why
Dr. Shweta SharmaOnly Grade A/B current calls enter the weighted model.
Sujay Nadkarni / The Nad CoderNo comparable historical accuracy is available.
Directional research blend only. Exact-score and settlement-phase claims remain separately scored in the source ledger. This is an experimental research blend, not a probability estimate or wagering recommendation.
What the monitored creators are saying
Direct picks and implied bracket calls are labeled separately. Conflicts stay visible and are excluded.
France vs Spain
England vs Argentina
World Cup champion
Historical results with provenance attached
Retrospective embedded-media evidence counts
Grade C claims are included in historical accuracy when the prediction is embedded in pre-event media and passes integrity checks. Grades A/B remain visibly stronger; Grade D/X stays unscored.
Captured before event
Public URL observed before kickoff; prediction and fingerprints saved.
Trusted forward recordIndependent pre-event proof
Archive, index, cross-post or third-party snapshot proves earlier public visibility.
Trusted forward recordRetrospective media evidence
Media timestamp predates the event and the prediction is embedded in the media, but public visibility was not observed then.
Included historical recordEdited or uncertain
Caption-only, ambiguous market, edited without recoverable original, or weak post-event evidence.
Context onlyExcluded
Post-result revision, incompatible duplicate, impossible bracket repair or integrity failure.
Never counted as a hitCaption and metadataYes
Detection: Instagram edit flag, caption timestamp, caption hash and observation history.
Limit: Instagram does not expose the original caption publicly.
Uploaded video or imageNot normally replaceable in-place
Detection: Media SHA-256, perceptual frame hash, OCR, transcript and extracted-prediction hash.
Limit: A creator can delete and repost, which creates a new post identity.
Public, private or archived stateYes
Detection: Public-access snapshots, direct-URL checks, disappearance and reappearance history.
Limit: A retrospective Instagram timestamp cannot prove the post was public at kickoff.
Deleting losing predictionsYes
Detection: Saved shortcode inventory, search indexes, archives, cross-posts and missing-post alerts.
Limit: Unknown deleted posts may remain undiscoverable if nobody captured or indexed them.
Multiple contradictory picksYes
Detection: Group every call by creator, fixture and market; incompatible calls are Grade X.
Limit: Different legitimate markets remain separate rather than being treated as conflicts.
Started per-creator source pages that preserve video context, exact prediction, outcome state, evidence grade and verdict. Added Jacky’s current semifinal calls and The Nad Coder’s frozen model claims.
Added four Grade C resolved quarterfinal calls and four Grade A pending calls. Advancement was 4/4 but matched a baseline that selected each DraftKings 90-minute three-way moneyline favorite to advance; phase-aware exact scores were 1/4.
Qualifying retrospective embedded-media predictions now count in historical accuracy with provenance attached; Grade D/X remains unscored.
Captured active semifinal/final forecasts with caption, prediction and media fingerprints.
France over Spain and England over Argentina recorded at public URLs before kickoff.
Found incompatible France 2–1 and Spain 2–1 calls. Both are excluded from clean scoring.
The watcher recorded a Sharkz France–Spain reel before kickoff and began change monitoring.
A later revision never erases the earlier miss.
- A/BTrusted forward record
- CIncluded historical scorecard
- DVisible context, not headline accuracy
- XPreserved, excluded, never counted as a hit