Crypto Signal Provider Subscriber Counts: Distribution and Source Bias

Measurement rules

Version finite-nonnegative-integer-nearest-rank-v1 accepts only JavaScript integer values that are at least zero. Missing values remain missing. The frozen snapshot contains 0 invalid provided values, 0 negative values, and 0 zeros. Quantiles use the nearest-rank rule: sort observed values ascending and select the value at the ceiling of the requested proportion multiplied by the observed count. No interpolation, imputation, deduplication of audiences, or freshness adjustment is applied.

One provider slug contributes at most one stored observation. The manifest does not retain a per-value capture timestamp, active-member count, view count, paid-member count, or methodology supplied by each directory. Repeated people can belong to multiple channels, and related records can describe overlapping audiences. The values must never be summed into a total CSR audience or a total crypto-signal audience.

Observed distribution

There are 1,198 unique values among 1,244 observations. The maximum is 18,074.5 times the median. That ratio and the distance between the median and mean show why a headline average would be easy to misread as a typical channel size.

Size buckets among observed rows

The two largest buckets by row count are 10,000-99,999 and 1,000-9,999. Bucket shares use only the 1,244 observed rows. Applying these percentages to all 3,200 records would erase the 1,956 missing observations and overstate what the inventory knows.

Coverage and distribution by source family

TGStat group-rating rows have a median of 6,638, while TGStat category rows have a median of 41,581. Cornix rows have a median of 287. These differences cannot be attributed to provider quality: the source families use different discovery frames and may capture different kinds of public object at different times.

The source-bias finding

RealTelegram supplies 1,870 records, the largest source family, but no subscriber observations in this manifest. Manual and WhatsApp seeds also contribute none. By contrast, both TGStat families and Cornix have complete subscriber fields, and all 6 Discord seeds have observations. The resulting distribution is therefore a distribution of values exposed by particular source pipelines, not a neutral census of audience size across the entire candidate inventory.

Even inside one source family, a public count may refer to a channel, group, bot, marketplace profile, or another platform object. The report does not reconcile object type, dormant accounts, purchased members, duplicate accounts, regional access, message views, or private-room membership. Source completeness is useful for locating measurement bias; it does not repair the bias.

Workflow status does not change with count size

How to cite a subscriber observation safely

A bounded provider-level statement names the source, says the value was observed rather than audited, and gives the capture date when one exists. This snapshot can support an aggregate sentence such as: “CSR stored subscriber observations for 1,244 of 3,200 candidate records, with a nearest-rank median of 11,210 among observed rows.” It cannot support a sentence that calls the median typical of the market, treats the maximum as verified reach, or says large channels are more reliable.

For due diligence, replace audience shortcuts with evidence that matches the decision: independently resolve the provider route, preserve the offer and payment identity, request a complete signal archive, reconstruct losses and open positions, and test follower-side execution with fees and slippage. A count can describe a public interface at one point in time. It cannot answer whether a reader should trust, pay, connect, or trade.

Official context sources

These sources explain why identity, disclosure, complete performance evidence, and resistance to urgency matter. They do not validate any record in the CSR dataset.

• ESMA: finfluencer factsheet

  Transparency, accuracy, paid-promotion disclosure, and recommendation boundaries.

• Investor.gov: social media and investment fraud

  Identity verification, impersonation, testimonials, urgency, and social-media limitations.

Continue the transparency research

• Where the 3,200 Crypto Provider Records Came From

  Source-family and field-completeness analysis for the 3,200-record CSR provider inventory, including TGStat, RealTelegram, Cornix, manual seeds, and gaps.

• Which Crypto Signal Provider Metadata Fields Appear Together?

  Cross-tab analysis of direct routes, aliases, subscriber observations, and explicit keys across 3,200 provider records, with seven observed field patterns.

• Crypto Signal Provider Country and Language Metadata Audit

  Audit country and language labels across 3,200 crypto signal provider records, including source defaults, non-global coverage, and jurisdiction limits.

• Cornix Crypto Signal Marketplace Metrics: Review Sample Size Audit

  Original audit of 85 frozen Cornix marketplace records: reviewer denominators, star summaries, derived weekly activity, offer flags, and exchange compatibility limits.

• Crypto Signal Provider Transparency Report 2026

  Original analysis of 3,200 crypto signal provider candidate records: platform concentration, verification gaps, quarantine rates, sources, and missing evidence.