DEV Community

ABROWNFOX001
ABROWNFOX001

Posted on

Why I Built a BTC 5-Minute Specialist

GitHub: https://github.com/abrownfox0/abrownfox001-twap60-prediction-trigger-system

YouTube walkthrough: https://www.youtube.com/watch?v=XzhugRL6BV4

This is part 1 of an updated 5-part series on the live abrownfox001 engine.

  • Part 1 — Why specialize in BTC 5-minute markets (this post)
  • Part 2 — TWAP60 as the settlement target
  • Part 3 — The private signal stack
  • Part 4 — Slot lifecycle: enter, scratch, redeem
  • Part 5 — Public evidence, risks, and what stays closed

Live profile: @abrownfox001

Wallet: 0x12b712029efeee7efc4066e98e4e1b6beb8b1b0b

Active since: Jan 12, 2026

The Market Everyone Treats as Noise

Polymarket BTC 5-minute Up/Down books sit near 50¢ because short-horizon direction looks almost random.

That is why most builders do one of three things:

  1. Skip 5-minute markets
  2. Chase the last CEX tick
  3. Leave the mid-band and farm 94–99¢ tickets instead

I took the opposite route.

The engine is a single-market specialist:

  • BTC
  • 5-minute clock
  • coin-flip band
  • official Chainlink TWAP as the target
  • scratch decaying tickets
  • redeem only the high-conviction book

Public sample from the README: ~99% BTC, ~96% 5-minute, ~90.6% of fills in 45–55¢.

Why 5 Minutes, Not Everything

A 5-minute slot is a complete trading problem that repeats 288 times a day.

That clock gives you:

  • enough independent trials to measure process
  • a short enough horizon that last-tick folklore used to dominate
  • a long enough horizon, after TWAP60, that a path forecast can exist
  • capital recycling if you scratch instead of marrying every fill

Politics, long-dated books, and multi-asset spray are different businesses. This repo is not trying to be those businesses.

Why Not 99¢ Farming

Convergence farm abrownfox001
Zone 94–99¢ 45–55¢
Edge Near-certainty × volume TWAP-path forecast + culling
Exit Hold almost everything Scratch weak / redeem strong
Binding constraint Fees and inventory Signal, latency, fill quality

Resolved sample in the docs (150 markets near 50¢, Jul 2026):

  • Held-side win rate 60.7% (91W / 59L)
  • ROI on capital at resolution ≈ +24.5%
  • Average held entry ~0.50

Those numbers describe the filtered book that reached settlement. They are not a promise that every fill is a 61% event.

The Three Jobs That Have to Work Together

  1. Calibrated 5m signal — live TWAP versus slot-open reference
  2. Fair-value entry — mean ~0.503, clips $50–175, median buy ~47s after open
  3. Conviction cull — ~930 sampled markets bought and sold near the same price

Miss any one and 50¢ trading is coin-flipping minus costs.

What Changed the Product

Before August 2026, a last-second wick could decide a 5-minute market.

Now those markets settle on a 60-second Chainlink TWAP. The repo is named for that world: twap60-prediction-trigger-system.

The bot is not built to fight the new rule. It is built to forecast the number the venue actually pays.

What Is Public

  • Live profile and wallet
  • Session walkthrough video
  • Windows binary: Abrownfox001.exe (2-day trial)
  • TypeScript shell: engine.ts, twapFeed.ts, signal.ts, clob.ts, markets.ts

What stays closed: the calibrated probability. The repo is the architecture, not a drop-in replica of the live read.

What Part 2 Covers

The settlement target in detail:

  • why 60s TWAP replaced the snapshot
  • why open pinning must use the same series
  • why CEX last price is only lead-lag

If that layer is wrong, nothing in parts 3–5 can save the book.

Links

Not financial advice. Short-horizon markets are high-risk. Paper or micro-size first.

If you have more questions, please feel free to contact me at any time: https://t.me/abrownfox001

My Polymarket Activity: https://polymarket.com/@abrownfox001?tab=activity

Top comments (0)