polyrag

← Decided

BTC price on Sep 15, 2026 at 2pm EDT? — $76,500 or above

kalshi · open the market ↗ · closes 2026-09-15T18:00

undecided high confidence not the named source buy YES price moved 9 → 93 in the last 48h bid 0.930 / ask 0.940 6.0c residual

Was there ever a window?

Unknown. the cited sources carry no publication date.

What the contract requires

Condition. The simple average of the 60 seconds of CF Benchmarks' Bitcoin Real-Time Index (BRTI) immediately before 2:00 PM EDT on September 15, 2026 is greater than 76,499.99.

Named authority. CF Benchmarks (Bitcoin Real-Time Index, BRTI)

Form the evidence must take. CF Benchmarks' official Bitcoin Real-Time Index (BRTI) data. The rules state: 'the price used to determine this market is based on CF Benchmarks' corresponding Real Time Index (RTI). At the last minute before expiration, 60 RTI prices are collected. The official and final value is the average of these prices.' No specific publication medium is required beyond the index itself.

Does not count.

Deadline. 2026-09-15T14:00:00-04:00 (2:00 PM EDT)

This venue published no resolution text for this contract, so the condition above is inferred from the question alone.

What the sources establish

The market resolves based on a future event (2:00 PM EDT on September 15, 2026), and the current time is before that deadline. No evidence of the required BRTI average exists in the provided sources.

The evidence is not in the form this contract requires. The condition requires the simple average of 60 one-second BRTI readings from the last minute before 2:00 PM EDT on September 15, 2026, as published by CF Benchmarks. The supplied sources contain no such data; the deadline has not yet passed.

The event may well have happened. This contract does not resolve on that alone.

The named authority has not established this. The sources below report the event; the contract resolves on what its named source says.

The official CF Benchmarks Bitcoin Real-Time Index (BRTI) data for the 60 seconds ending at 2:00 PM EDT on September 15, 2026, showing the simple average.

Sources it was read from

8 retrieved but not cited
XRP-Preismarkt bei 7pm EDT: So wird die Abrechnung an CF Benchmarks RTI gemessen ↗
LONDON (IT BOLTWISE) – Ein neuer Prognosemarkt koppelt den XRP-Preis direkt an den CF Benchmarks Real-Time Index (RTI). Entscheidend ist nicht ein Durchschnitt über Tage, sondern ein einfacher Mittelwert der RTI-Werte in den 60 Sekunden vor dem festgelegten Zeitpunkt. Der Kontrakt zahlt pro richtige
XRP-Prognosemarkt klärt Preis von Ripple am 11. September 2026 um 14:00 EDT ↗
LONDON (IT BOLTWISE) – Ein neuer Prognosemarkt fragt nach dem XRP-Preis von Ripple am 11. September 2026 um 14:00 EDT. Entscheidend ist dabei ein regelbasiertes Abwicklungsverfahren auf Basis des CF Benchmarks Real-Time Index, der 60 Sekunden vor dem Stichtag gemittelt wird. Trifft die Preisbedingun
Artificial Analysis defends its AI benchmarking methodology ↗
A Reddit post defends Artificial Analysis, arguing that claims of the benchmarking service being "broken" or "bought out" are unfounded. The author explains that Artificial Analysis uses its own funding for independent benchmarks and publishes its methodology, allowing for transparency. The post hig
Bitcoin Mining in September 2026: How to Mine Bitcoin and Earn $7,700 in Profits ↗
Over the past week, CryptoQuant’s Bull Score Index surged from 30 to 80, reaching its highest level since October 2025. Eight of the index’s 10 underlying indicators are now flashing bullish signals. The post Bitcoin Mining in September 2026: How to Mine Bitcoin and Earn $7,700 in Profits first appe
The WiseMatch Index Sets a New Standard for Behavioral Health Provider Quality ↗
Reviewed by Motive Medical Intelligence, the WiseMatch Index brings quality evaluation to enhance patient–therapist matchingSAN FRANCISCO--(BUSINESS WIRE)--Octave today introduced the WiseMatch index, a proprietary methodology for matching patients wi... L'articolo The WiseMatch Index Sets a New Sta
Review maps 173 LLM mental health benchmarks, finds data limitations ↗
A systematic scoping review titled "Mental Health Benchmarks for Large Language Models" has identified and analyzed 173 benchmarks. The review found that the majority of these benchmarks utilize data from social media or are generated by LLMs themselves. Notably, only a small fraction, specifically

Adjudicated 28d ago · retrieval: newsrag. This is a reading of retrieved documents by a language model, not a settlement. The venue's named authority is the only thing that resolves a market, and it is often slower than the news.