angbot

Lead over the public stream.

Latency is measured, not claimed. Each row is one source we are listening to, aligned word by word against the public stream for the same event.

Connectingpolls every 10 s
median lead
0.0 s
sources this week
0
events today
0
listeners online
0

Median lead by source kind

No aligned samples yet.

Per-source lead over the public stream
SourceKindp50 leadp90 leadSamplesLast seen

How we measure

For each source Langbot also records the public stream (webcast, YouTube, C-SPAN) and aligns the two transcripts over word sequences. lag_vs_public_ms is the per-word difference in t_spoken; p50 and p90 are over the last 24 hours of samples. A positive number means we heard the word before the public stream did.

What it does not mean

A 19 s lead does not imply a fill. Venues delay orders on live events (Kalshi Rule 13.1(i), Polymarket taker delays), liquidity in long-tail markets is thin, and the edge shrinks as more participants listen. Paper-trade first.