Second entry: Same logic, plus order flow analysis gave even more confluence.

First entry: Price was respecting the EMA, we had a nice downtrend, plus old targets below. The 1h FVG got respected, which kept our short idea valid. Then I waited for an inverse of the EMA 20 and for the small FVGs to get inversed.

$XAUUSD Two shorts, both hit their targets. Two entries, one idea. 6.4R in the bag. Logic in the comments

Agreed, it's a bot. The play isn't mirroring every fill, it's using its history to build your own: feed the trades to an AI, find the entry timing and prices. And with fast enough execution, copying can still work.

Wallet link: polymarket.com/uk/@asdaasda

One wallet worth watching for copytrading on Polymarket. +$47.5K in October alone Active since August. 34,334 trades in just over two months, all on one market: Bitcoin 5m Up or Down. The numbers: → $1.3M volume → $77.5K profit → about $2.30 per trade on a ~$38 average entry → 53.1% win rate → 1,610 markets traded → +$47.5K in October alone The best single trade made $4,113, but that’s an outlier. The profit comes from thousands of small entries with a small edge. A 53% win rate doesn’t look like much. At this volume, it turns into a steady upward curve. Wallet in the first reply.

We are here!

Rest in peace, @arkham 💔 $1,500/month for access + slashed alert limits. If you were tracking wallets there, time to migrate.

And additional 2.4R on Gold 4.4R totally today

DXY chart

Nice 2R trade on EURUSD Logic: Frankfurt high was swept, DXY low was swept and on a low timeframe there was a good possibility to go up. Then you look for an entry and here you go.

Block Blast made around $127M between January and May 2026. $127M from a block puzzle. Someone reverse-engineered the code, and according to them, your skill barely matters. What matters is how much you’re worth in ad revenue. What the game reportedly reads: → your phone model → your country → how much the ads you’ve seen have earned them → whether vibration is turned on A model (XGBoost) runs right on your phone and predicts how much ad money you’ll still bring in. The game gets easier or harder based on that. The rules found in the code: “MultiClearBoostThenKillTrait” You feel like a genius until your session reaches the usual length. If you play past it while an expensive ad is ready, you lose right away. “RewardHighValueAdHarvestTrait” If you watch ads, you get easier pieces so you can almost beat your record. Then you lose, and an ad pops up to “continue your run.” Keep going and your score quietly gets a hidden +10% so you come back. “MABXiBaiTrait” A/B tests which minute to make you lose, to see which one gets you to watch an ad. If a block puzzle can engineer this much, think about what casinos and slot apps are doing. “It’s warmed up, it’s about to pay” is just the system working as designed. The game isn’t testing you. It’s pricing you.

Where did the idea that trading is easy come from? That it beats a regular job? Here's my own experience. 1. Hours
Job: a 40-hour week.
Trading: I sit at the chart from 8:00 to 17:00 and skip 12:00 to 14:00.
It's basically the same time. 2. Exhaustion
Job: I came home physically drained.
Trading: I finish the day mentally drained.
Either way, all I want is rest. 3. Health
Job: physical work wears your body down over time.
Trading: constant stress does the same, just from the inside. 4. Pay
Job: you get paid at the end of every month.
Trading: nothing protects you, and you can lose what you already have. The dream of easy money and a beautiful life is tempting, and it's fake. Once you quit your job and go full-time, you see the picture isn't that pretty after all.

A wallet worth watching No name, just an address. Since March it has traded 14,360 markets, pushed $29M in volume and made $60.1K in profit. It only trades crypto Up or Down markets: Bitcoin, Solana, BNB, XRP. It enters at 98–99c when the outcome is almost decided and waits for the $1 payout. Position sizes run from $2,000 to $39,600, and everything closes the same day. The key number isn’t $60K. If it had never lost, it would be sitting near $290K. So it loses regularly and still ends up in profit. That’s far stronger proof than a smooth PnL line with no red at all. polymarket.com/uk/@0x9ad6c53…

Next test: feed it a real incident and see if the modeled spread lines up with the actual perimeter. That’s where prototypes like this either earn trust or don’t.

Haiku for research, Sonnet for building, Opus for verifying is a smart cost split. $1.30 for a full run with a verifier step beats most people’s single-model setups.

Cool workflow, but “rebuild any app from its compiled code” is going to run into licenses and ToS long before it runs into technical limits.

Robot gets its first paycheck and buys a flying car. I got my first paycheck and bought groceries. Who’s the real machine here?

The scary part isn’t that Vitalik needed Tor + cover traffic. It’s that the default setup leaks enough that someone like him thought it was necessary.