Overview
Every strategy family on one line: its headline number, how many observations stand behind it, and what kind of evidence it is - live paper (real books, real prints, no real orders), a replay simulation of the journal, or a historical backtest. The status never outruns the data: below 30 observations it says exploratory; where a best variant is picked from several, its t must clear a higher bar.
- Loading…
Intra-candle simulator
Every journalled candle replayed against the order books recorded at each judgment. Per candle: never more than the budget out of pocket (so a candle can lose at most the budget), one side at a time, buys walk the asks and sells walk the bids, the taker fee on both, and whatever is held at the close settles at $1 or $0. Every fair-value source (the model, the stacked predictor, the model moved by the Binance lead, and a coin) is traded three ways: active (sell when the bid clears fair value - usually flipping to the other side on the same tick), flat (as active, but a flip needs a wider margin) and hold (the first buy held to settlement). Only candles the trader watched from open to cutoff are scored.
Sample
Rule parameters
Sensitivity only - choosing parameters on this same data overfits it. Changing these re-runs the simulation on the server to show how fragile a result is. The defaults are the rule under test; a parameter set picked because it looks good here has already used up this data.
Strategies
Luck band · the model against its own rule on borrowed paths
Table view
Paired comparisons
A − B per candle, over the same scored candlesCumulative P&L by scored candle
complete, settled candles in order · cut-short candles are not in these curves · click a key to hide or show a lineTable view
Per-candle P&L
every settled candle · click a bar to open that candleTable view
Candle
Tick table
Orders in this candle
All orders
Next-candle strategies
Buy the next 5-minute candle's market before it opens, once our probability for one side is 55% or more, then either sell
the whole position as soon as that side's bid reaches 85¢ while the candle runs (take profit) or keep it to settlement (hold),
for comparison. One entry per candle, at most $5, the taker fee on the buy and the sell. Replayed by src/nextsim.js from the
trader's nextbook samples of the next market's book. This is a fixed rule: the parameter form above does not re-run it.
The candle-pattern variants (next-pattern-*) bet the next candle from the
shape of the one closing now: the same colour again, the opposite colour after an inverted hammer, or the same colour except after an inverted hammer.
Each buys at the last sample before the open and holds to settlement, with flat staking ($5 on every bet) or parlay staking: $5, then
the whole bankroll after every win, cashed out near $80 or restarted at $5 after any loss. A parlay's stake per candle and its chains (started,
reached the target, busted) are shown below; these rules have no luck band.
Why the lean only appears in the last minute
Strategies
Take profit against holding, per candle
Next candles
Next candle
Sample table
Late lock
Buy the side the settlement is already locked towards, late in the candle, and hold it to settlement. From T+237 the settlement
average fixes one second at a time. A side counts as locked when the model gives it at least 99.9% and the projected
settlement leads the strike by at least $10 on that side. The rule then buys the locked side the first time its ask is at or under
the cap: one entry per market, $5 all-in, held to the official outcome. src/latesim.js replays it on the books the
trader journalled every second from T+180 to T+297 (latebook rows): real books, no real orders. It is a fixed rule, so the parameter form
above does not re-run it.
The payoff is lopsided: count reversals, not wins
Variants
Markets
Market
Sample table
Historical test
pendinghistorical test: pending. A pre-registered test of this rule on about 7,500 more historical markets goes here, once its result exists.
Nothing is loaded into this slot yet, so nothing on this page counts it for or against the rule. Until then, the verdict above rests on the journalled books alone.
Paper market-maker
Sample
Variants
Markout per share after each fill
Table view
Most recent fills
Live paper trader
The strategies in src/main.js taking paper positions on the live books, scored by src/scoreboard.js against the
official outcomes. Claimed edge is what a strategy believed each share was worth at entry, minus its all-in cost; realized edge is what the
share paid ($1 or $0) minus that cost. If the probability a strategy trades on is honest, the two converge over many trades - that gap, more than the
P&L, decides whether to trust it.
Strategies
Claimed against realized edge, per share
the horizontal gap is the number that mattersJev question lab
Every question in the Jev bank is a candidate feature. src/jevlab.js admits one only if adding it to the market-anchored
predictor (the book plus the exact settlement arithmetic) lowers log-loss on markets it never saw - leave one market out, each market weighted once,
Holm-adjusted across the whole bank. The expected result is that none does.
Sample
Questions, one at a time
change in out-of-sample log-loss per market when the question is added; positive helpsAll questions together
Backtests and verification
History the live journal cannot give yet, rebuilt from exchange data by npm run backtest, and checks that the
simulators on this page find an edge that is certainly there and invent none that is not.
The report