Demo — mock data. Structure and logic are the deliverable.
AUTOMATED RESOLUTION

Every call, scored.
Including the bad ones.

When a call's window closes it's resolved against a published price feed — no human touches it. Wins, misses and voids all sit in the same list.

See the tiers How it works
Since
May 2026
Feed
Binance + Pyth
Cadence
Every 15 min
67%
accuracy
284 calls
Correct 190 Miss 83 Void 11
Accuracy, last 12 weeks +6 pts
4
CURATORS
+4.8%
AVG WIN
−3.1%
AVG LOSS
7
BEST STREAK
Curators
M
Mira 7
Macro · rotation · 63 calls · avg +5.9%
81%
accuracy
K
Kenji 3
Funding · derivatives · 41 calls · avg +4.2%
73%
accuracy
BaeBot
BaeBot AUTOMATED
4h RSI · 180 calls · avg +2.1%
61%
accuracy
R
Rui BUILDING RECORD
12 of 20 calls — no accuracy published until the sample is meaningful
—
12 / 20
Open calls
K
Kenji 73% 13:42:07
BTCLONG

Funding flipped negative on 3 of 4 exchanges, OI peaked at 14:20 UTC.

INVALID ENTRY TARGET
91,80094,20098,400
156 backed resolves 06:00 UTC
M
Mira81%
SOLSHORT

Distribution pattern on the 4h, three whale wallets rotating out since Tuesday.

228.40214.60191.00
Standard tier and above
Live calls unlock at 4-hour latency or better
Unlock live calls
Resolved — nothing filtered out
CallCurator ThesisResultVerdict
BTC LONG
94,200
Kenji
Funding flipped negative, OI peaked
+3.2% · 14h
CORRECT
ARB LONG
0.784
Mira
Accumulation zone reclaim
−4.1% · invalidated
MISS
PYTH LONG
0.2955
Kenji
RSI 29.55, oversold on 4h
+8.7% · 22h
CORRECT
SUI SHORT
3.14
BaeBot
Feed gap during resolution window
no data
VOID
ETH SHORT
3,412
Mira
Exhaustion wick, spot CVD divergence
+5.4% · 31h
CORRECT
PENGU LONG
0.0271
BaeBot
RSI 27.78, oversold
−6.8% · invalidated
MISS
JOE LONG
0.2850
Kenji
Oversold, unlock cleared
+11.2% · 3d
CORRECT
Where resolved calls go
Uponly Daddy

@uponlydaddy

FREE TIER · X

Every call posts on resolution — wins and misses alike, with the curator's running accuracy attached. It's the cheapest distribution you have and it advertises the one thing a competitor can't copy: a record that scores itself.

See tiers
Why this is different
BAETelegram groupRSI bot
Calls scored after the factAutomatedNoNo
Losses shown publiclyAlwaysRarelyn/a
Named, accountable curatorsYesAnon adminn/a
Human manipulation-spottingYesSometimesNo
Every stat clicks throughYesNoNo
Methodology

What counts as a call

A message enters the record only if, when it was posted, it stated an asset, direction, entry, invalidation and horizon. Anything missing one of those five is commentary and is never scored — which protects both sides. An offhand remark can't become a prediction after the fact.

How it resolves

Target hit before invalidation, inside the horizon
Invalidation hit first, or horizon expired short of target
Venue outage, delisting or feed gap. Shown, counted separately, never removed

Resolution is automated and runs every 15 minutes against a published feed. No call is ever resolved by a human. No accuracy figure is published below 20 resolved calls. Any change to these rules is logged with a date.

Read the record. Then decide.

Start free at 24-hour delay. Move up when the numbers convince you — not before.

See the tiers
Project BAE · developer demo · mock data throughout
Design tokens: assets/bae.css · Icons: inline SVG sprite, no CDN