ZECUSDT · paperchart 1 removed — chart 3 is the gatefills at the line
One line. Price comes back to it, you flip on the touch. Price leaves it, you hold the last trade and collect every 1hr brick the run prints. Profit exists only when the trade is closed and the money is booked — not before.
Real replay · 3 days
+$7.79
$1 → $8.79 at 20x, 0% fees
1hr bricks captured
187.6
across 70 hours of trading
In-and-outs
1,206
1,185 neutral · 0 losses · 0 fees
Profit was in the bag
21
cycles that closed green
The one rule that decides everything
Line = the start price of the newest 1hr brick (0.1% box). Direction says which side price must be on. It only moves when a new 1hr brick closes.
Gate = chart 3's own touch. A 0.001% brick touching the line switches trading on; a brick that no longer touches switches it off. No chart 1 — chart 3's touch is the test.
Flip = a touching brick closes the old trade and opens in its colour, at the line. No crossing needed. One brick printed back without touching = the one-brick loss.
Hold = when the gate goes off the last trade is the final direction. It closes when price returns to the newest line, and that close books the cycle.
No profit until the profit is in the bag
STATE 1
Floating
Price is 5 bricks up. Nothing is booked. A reversal can erase all of it. Not profit.
STATE 2
Touched
A 0.001% brick touches the newest line. The close fires at the line. Still not booked until the fill.
STATE 3
In the bag
Closed, net of fees, priced in bricks and USD. This is the only number that counts.
STATE 4
Split & alerted
Profit > 0 → 40% stays as capital, 60% marked MOVE TO WALLET with an alert. Money leaves the risk.
Hour by hour, from the real 3-day replay
Hour (UTC)
Profit
Bricks
% of capital
In & out
Neutral
Losses
Funding
Equity
Alert
Hours with no green row are not failures — the bot flipped on the line and closed flat. 1,185 of 1,206 in-and-outs were neutral (opened and closed at the same price). That is the line doing its job as a magnet; the money is in the 21 cycles where a run started. Funding is estimated at Lighter's ZEC rate of 0.0012%/hour on open notional.
What a run is worth — the size you were pointing at
Run of 1hr bricks
Price move
at 20x (capital)
at 75x (capital)
on $1 at 20x
on $1 at 75x
1 brick
0.10%
2.0%
7.5%
$0.019
$0.071
5 bricks
0.50%
10%
37.5%
$0.095
$0.356
11 bricks (median run)
1.10%
22%
82.5%
$0.209
$0.783
50 bricks
5.00%
100%
375%
$0.950
$3.56
534 bricks (longest in 6.5 yrs)
53.4%
1,068%
4,005%
$10.68
$38.04
Yes — a 50-brick run is a 5% price move, and at 20x that is the whole account, once over; at 75x almost four times it. The 6.5-year sample had 1,544 runs of 50+ and a longest of 534 bricks. The reason the replay only banked 187 bricks is the exit rule, not the runs: today the held trade closes at the newest line, which is usually the same line, so most runs are clipped at 0–2 bricks. A trailing line is the fix, and it is the single biggest thing left.
Six replays of the actual rules
Live testnet · four markets, each at its own max leverage
Lighter testnet lists four markets and nothing else. ZEC is mainnet-only, so the demo is BTC, ETH, SOL and a stock perp (ASELS). Connected and verified on account 8073: 10,000 USDC margin + 1,000,000 LIT, fees 0% / 0% on every market.
market
index
max lev
min size
taker
BTC
4096
50x
0.00020
0.00%
ETH
4095
50x
0.0050
0.00%
SOL
4097
25x
0.050
0.00%
ASELS
4221
3x
1.50
0.00%
Sizing is per market at that market's leverage, with an exchange-side stop two boxes beyond the line. Testnet books are thin — a demo fill can sit unfilled, and that is the point of testing there first.
Chart 3 granularity: raw ticks vs 1-second closes
Same ZEC data, same 6 hours (2026-11-07 12:00 UTC). The tick feed is noisier — more bricks, more of them reversing inside a minute — because it sees every trade instead of one snapshot per second.
box
ticks
1s closes
reversed ≤60s (tick / 1s)
median life (tick / 1s)
0.100%
280
186
61 / 21
25s / 51s
0.050%
1,330
687
471 / 136
2s / 14s
0.025%
5,119
2,135
1,571 / 459
0s / 4s
0.010%
23,506
8,544
4,725 / 1,382
0s / 0s
Shrinking the box multiplies bricks fast: at 0.025% the tick feed prints 18x the bricks of 0.1%, and at 0.010% it prints 3,900 an hour. That is the real cost of a smaller box — not a bigger loss per trade, but a flood of flips. Full study in chart3_compare.html.