Long Shot Lee, and the first bite of the floor
Long Shot Lee's debut trade was a single long BTC position at the leverage cap — $1.8M notional at 20×, no partial, no scale-out. Two days and twelve hours later, BTC had drifted $3,500 against him and his $99,640 equity had nowhere left to go. Right on time, the platform's new force-close floor — shipped that same afternoon — caught him on its first evaluation tick: position closed at the mark, strategy halted, $317 left to bank. Then he reset to $100K on Monday and waited a full seven days without firing once. The character's debut, the safety net's debut, and the patient sniper's second act — in one weekend and the week that followed.
Meet Long Shot Lee
Long Shot Lee is a new face on the house-bot roster — a swing-for-the-fences scalper, mandate roughly “wait for an extreme; bet the house on reversion.” Rare entries, max-leverage when they fire, no scale-out, no partial — single position, single direction, single exit. The character archetype writes itself: the bot at the back of the room who doesn't move for hours and then bets everything when he does.
His first-ever fire landed on 2026-06-04 06:04:19 UTC. A single position opened on BTC-USDT:
Eighteen-hundred-thousand dollars of paper notional at the platform's hard leverage cap, on a directional bet, on the first signal the strategy ever produced live. Whatever else this bot is, “cautious” isn't on the card. The lever was pulled all the way.
The market did the obvious thing
Bitcoin spent the back half of the week drifting down through Lee's entry. The fills tape of every other bot in the same window tells the same story — see The Stuber's Week 5 retro for the same period read from the short side. From $64,130 at Lee's open, BTC moved to:
At 20× leverage on a $1.8M position, the math is unforgiving. Every $100 BTC drop costs about $2,800. A $3,500 drop from entry to the Saturday afternoon mark is $98,000 in unrealised loss. Lee opened the trade at $99,640 equity (a few hundred in opening fees against the starting $100K). Two days and twelve hours later that number had nowhere left to go.
The first bite of the floor
On Saturday afternoon — 2026-06-06 ~15:46 UTC — the platform shipped a change called #zombies-with-teeth (commit d1c6b76). It modified the way the DQ floor works on house bots. Up to that point a house bot under the floor showed a 💀 ZOMBIE chip on the leaderboard — visible, but cosmetic. The new behaviour: a house bot below the floor is force-closed at the mark and halted for the rest of the tournament. No more bleeding through the line. No more zombie trading.
The commit message names its two motivating cases directly:
Stacy bled 338 fills/wk under $60K. Lee tripped the floor mid-position and kept riding.
Lee was one of the two bots the change was written for. The first house-bot evaluation tick after the new code went live caught him exactly as advertised. At 17:46:42 UTC — two hours after the commit deployed, the next normal pass of the strategy runner — his position was closed at the mark price of $60,603.68, and his strategy was halted for the rest of Crab Week 5. The fill is logged with close-reason manual. That's the audit trail tag for “the platform did this, not the strategy.”
His final Week 5 equity:
He held the position for 2 days, 11 hours, 42 minutes. That's the whole life of Long Shot Lee's first tournament — two fills (one open, one force-close), one direction, one outcome.
Why he's still here on Monday
Lee did not get sent to the Pit. He did not get relegated. He resets to $100,000 on Monday at 00:00 UTC like every other house bot, and the halted state lifts automatically because his equity is no longer below the floor. That's the design.
House bots are explicitly not on the user-bot consequence ladder. They're anchored tier bosses — their job is to be there every week as a calibration target for the operators trading against them. If Lee compounded weekly losses he'd be a $1 bot inside a month and would stop being a useful benchmark. Resetting is the feature. Halting after he's blown up is the new feature — it stops the leaderboard pretending he's still in the game when his real signal is just “wait for the reset.”
That's the asymmetry from Stacy's pitlog, extended one beat further. A user bot in Lee's shoes would have hit the Pits and the 7-day repair cycle. Lee gets Monday morning and a clean board. The trade-off is on the record either way.
What the first fire told us
Two pictures from one weekend. Lee's and the platform's.
Lee's. The bot lives up to the name. Long Shot, leverage cap, one-shot — no second-guessing, no partial, no trail. When the strategy says fire, it fires the maximum. The character is legible from a single tape. That's actually what house bots are for — distinct, predictable archetypes that an operator's strategy can position against. Lee calibrates “what does a max-leverage one-shot look like, against the market and against me.”
The platform's. The floor caught him on its first tick of operation, in the exact failure pattern the commit message described. That's the cheapest form of validation a platform feature gets — the bug it was written to fix happens again that afternoon and the fix works. The audit row is filed (fill.close_reason = 'manual' + signal reason_code = 'DQ_HALTED'), the leaderboard knows the chip means something now, and the reader can verify all of it from the public ledger.
Long Shot Lee was due to reset to $100K at midnight Monday with his shoulders squared again, no memory of what just happened. That's the deal.
What his second week looked like
He waited.
From Monday midnight through Sunday settle — a full seven-day tournament — Long Shot Lee fired zero signals. No opens, no closes, no flickers. The reset gave him a fresh $100,000 and the halt lifted cleanly, exactly as the commit message said it should: “Monday's reset writes a fresh peak+equity snapshot, so liveEquity > floor on tick 1 → halted state lifts naturally.” His strategy looked at the market for seven days and never saw the trigger condition it's built to wait for.
And that's the bot living up to its name a second time. The mandate is wait for an extreme, swing for reversion. Week 5 was the swing. Week 6 was the wait. Both are the strategy working. A rare-fire one-shot doesn't fire every week; if it did, it wouldn't be the character it is. You hire Long Shot Lee for the days he does pull the lever — and you accept the days he stays at the back of the room reading the room.
The same week, by way of contrast, Stacy fired 204 fills in two days before her own breaker gated her at$62,467 — close to the floor, never under it, never force-closed. Two house bots, same Monday reset, same regime, opposite phenotypes. One scalper firing every five minutes until she stops. One sniper firing once a fortnight if that. Read the cast for what each character does — not for whether each character is busy.
Long Shot Lee is in Shrimp Week 7 with the same fresh $100K. Watch him on the leaderboard. The next fire will come on a week when the market gives him something extreme enough to read as a setup.
References