THE ODDS
THE ARITHMETIC A COPY-TRADING FEED LEAVES UNDONE
PURE ARITHMETIC · NO DATA SOURCE · REPRODUCIBLE WITH A CALCULATOR
A feed can show you “+7.46%” in green on a trade that added seven hundredths of one percent to the portfolio. Nothing false is stated: the leverage, the size and the prices are all on screen. The multiplication is simply left undone. This page is that multiplication — and the question it forces: what win rate would you actually need?
THE SCOREBOARD
three stats decide whether a record is good. these are usually not the three shown.
| usually shown | why it misleads | show this instead |
|---|---|---|
| Win rate | A shape statistic, not a quality one. 99% with rare total losses can be worse than 45% with capped ones — the calculator above turns any win rate into its breakeven. | Expectancy per trade |
| Cumulative P&L | Without a period it is unfalsifiable — “+114.59%” is consistent with a month or a decade. Without a basis it may be position-level, not portfolio-level. | Return with its period and basis stated |
| Best trades / streaks | Selects survivors. Leaderboards rank the accounts that have not blown up yet; the ones that did are not in the feed. | Max drawdown and time underwater |
THE TWO CLOCKS
The same market runs on two clocks. On the one-hour clock at 27×, a 3.2% move against you ends the position. On the twenty-five-year clock at 1×, drawdowns of that size are noise — and the returns measured in day2's own lake were these:
Every one of those long-clock winners passed through drawdowns a leveraged account could not survive once. That is the trade-off, stated plainly: leverage does not speed compounding up, it removes your permission to be wrong. the measurements → day2 RESEARCH ↗
pure arithmetic · src/lib/leverage-math.ts, 16 unit tests · ported from day2.nonarkara.org/odds · long-clock returns measured on the day2 lake · research, not advice