01 · The control
Scout
Scans for opening-range breakouts and follows a fixed checklist. Trades a real paper account at Alpaca. The yardstick every new idea is measured against.
Meet ScoutThree AI systems. One experiment. Total transparency.
Different
approaches.
Same reality.
I’m paper trading three automated systems, Scout, Quant and Atlas, in the same market under the same safety limits. Every trade, every loss and every failed idea gets published.
The race
Loading published dataFresh start with $5,000, Tuesday Nov 3 · 6:30 am Pacific / 9:30 am New York
–days
–hours
–minutes
–seconds
Standings from the latest published snapshot
ScoutAlpaca paper account
––
QuantVirtual fills · 21 bots in testing
––
AtlasHolds for months · in final testing
SoonJoins once its rules pass
Published snapshot · Not a live market feed
Updates
Big moves only. A post when something real changes.
The Experiment
Same limits. Separate brains. A guard that can always say no.
Research
Costs ate the edge. Every test, failures included.
Scout · Quant · Atlas
Fixed rules. The yardstick.
Data-driven. Must prove itself.
Patient. Holds for months.
Three minds.
One market.
Scout never changes how it picks trades. Quant researches alternatives and has to prove they work on data it has never seen before anything changes. Atlas plays the long game, holding stocks for months instead of minutes. Same safety limits.
02 · The challenger
Quant
Collects years of market data, tests ideas on unseen periods and logs every signal, even the ones it skips. Shadow trades with virtual fills until it earns more.
Meet Quant03 · The long game
Atlas
Ignores the daily noise. Buys a handful of stocks with lasting reasons behind them and holds for months. Measured against simply owning the whole market.
Meet AtlasTested before
trusted.
These come from historical simulations in the research database, not from a brokerage account. They include the failures.
–Simulated trades studied
–Strategy versions tried
–Stocks and funds tracked
–Versions that beat costs convincingly
–Years of daily price history
–Ideas given the one-time final exam on hidden data
–Ideas adopted after the exam
–Live forward test running
This is a public experiment
No cherry picking.
No hidden trades.
I publish the losses and the failed ideas because that is the only way to learn what actually works. Updates come when something big changes, not every day.
Read the evidence