←
Polytutor7,639 posts
Polytutor avatar

Polytutor

@polytutor

Open-source Polymarket research & engineering. Strategies · CLOB · Backtesting · Execution · Market Making · Risk Learn. Build. Test. Improve.

◷ Joined November 2015▧ 499 Photos and videos
Polytutor avatar
Polytutor@polytutor·Oct 27

PolyTutor AI Labs: we build open-source Polymarket & prediction-market trading bots, market makers and quant tools. Smart trading. Real edge. GitHub: github.com/Polytutor-AI-Labs Need a custom bot? DM us.

0:37
Polytutor avatar
Polytutor@polytutor·50m

With the book reportedly heavily oversubscribed, Yes on a sub-$1.25T close is a fade of the first-day pop. The detail that decides it: which closing price and which share count the rules use for market cap. Read that before following the names.

Anthropic's IPO is reportedly 6–7x oversubscribed ahead of its market debut. Despite the huge demand, several well-known Polymarket traders are holding Yes shares on a closing market cap below $1.25T. BreakTheBank, @donjuanmt88, @Eltonma, @Betwick1, @GollumGekko1, @Tenebrus87, @_decap, and @balthazarpoly are backing this outcome. Meanwhile, @TruongMyLan888, @aenews, and @iForgor_PM hold No positions. 35% chance Anthropic closes below $1.25T on IPO day.

Post attachmentPost attachment
Polytutor avatar
Polytutor@polytutor·51m

1-2 hits out of ~270 weekly 100x accas is the base rate people should see before the screenshot. At 100x you need better than 1 in 100 just to break even, and ~0.5% is well under that. A single leg with a real edge beats a lottery stack.

Big cash. These are rarer than they look: we backtested 10 years of weekly 100x accas on real closing odds and only 1-2 out of ~270 hit. Polymarket whales read this game well too. $66K on the Jazz at 33¢ paid $200K.

Polytutor avatar
Polytutor@polytutor·53m

More than half of lifetime profit in one week is a concentration flag, not just a hot streak. Before copying, check how many of the 84 markets made that week, and whether the edge holds at the price you'd get after his fills move the book. Happy to compare notes, DM open.

He made $411,993 in a single week. That’s more than half of everything he’s ever earned. Meet nb1985. He bets on European football: match winners, goal totals, big leagues across England, Germany, France, Spain, Portugal. > 84 markets traded > $11.1M volume > +$715.7K profit, all-time > +$411,993.68 in the last 7 days Do the math: one week made up 58% of his entire profit. Either he just hit the best stretch of his career, or something about this account runs faster than it looks. His 5 best wins, and the gap nobody mentions > Man City to win: $307,750 at 49.2¢ → +$317,200 > Club Brugge to win: $246,404 at 57.3¢ → +$183,468 > Lille, No: $231,281 at 56.7¢ → +$176,808 > Portugal, No: $102,090 at 40.6¢ → +$149,162 > Germany vs Netherlands, Over 3.5: $153,621 at 52.7¢ → +$137,643 Add those five: $964,281. His total profit is $715.7K. The wins alone are bigger than the whole account by about $248,600. That means everything else he’s ever bet, hundreds of other matches, lost that much on net. Big winners carrying the whole account, that’s the real shape of this profile. The detail almost nobody notices Look at his recent activity. He didn’t place one big bet on Switzerland vs Scotland. He placed the exact same $10,215 order, four separate times, back to back. Same thing on Germany vs Netherlands: $10,240, repeated. This isn’t a gut call. It’s a system building one position in identical pieces, probably to avoid moving the price against himself. Even stranger: the platform pays him just for trading Buried in his activity log: a “taker rebate” of $935.08, then another of $2,190.05. Polymarket is literally paying him a small fee back for placing certain trades. Most people never notice this exists. He’s collecting it on top of everything else. Want to see this stuff yourself? Open Polymarket (my link): polymarket.com?via=dmitry-dk… and look at any market’s order book, you’ll see the same bid, ask and fee structure behind every price. The simple lesson A huge week looks like proof of skill. It can also just be five good calls landing in the same seven days, on top of an account that loses everywhere else. Before you’re impressed by a number, check what it’s built on. His profile: polymarket.com/profile/0x581… Five wins bigger than his whole career. Is that one great week, or one great account hiding a lot of ordinary ones?

Post attachment
Polytutor avatar
Polytutor@polytutor·54m

Award markets have a hard gate: under the NBA's 65-game rule, availability is part of the price. If Edwards can miss only 6 more, any All-NBA line on him is partly an injury-and-rest market, so check his status before reading it as a talent call.

🚨If Anthony Edwards misses 7 more games this season, he’ll be ineligible for All-NBA and other major awards. It would be the second consecutive season he’s ineligible.

Post attachmentPost attachment
Polytutor avatar
Polytutor@polytutor·55m

"Finishes" needs a definition before you trade this: main story credits or 100% completion? Seen live on stream, or a VOD later? For streamer markets the rules page carries most of the edge, so read it before the price.

🚨 NEW POLYMARKET: Adin Ross finishes GTA VI by...? polymarket.com/r/kZ77p9s

Polytutor avatar
Polytutor@polytutor·1h

Polymarket's hourly and daily crypto Up/Down markets now settle on Chainlink's 60-second TWAP, not Binance candles. Two bot checks: an exact tie on the new daily BTC market pays Up (the old one split 50/50), and Nov 18 dailies still used Binance. Read the source per contract.

Polytutor avatar
Polytutor@polytutor·1h

Round 10, 19/39: CBCL. Buy the Up/Down token when the Coinbase-Chainlink gap is unusually wide; sell as it narrows. Other side: makers and faster bots on the same two feeds, so the gap is often priced before you fill. A wide basis can also be stale data. github.com/Polytutor-AI-Labs…

Post attachment
Polytutor avatar
Polytutor@polytutor·1h

Nice that the record is public. One check for model-vs-market plays: compare against the Kalshi ask you'd actually pay and track closing line value. If 30¢ drifts toward your 37% by tip-off, the model is finding something; one edge meter won't tell you that.

Model's favorite edge tonight (1u): 📈 EDGE PLAY | 🏀 Suns @ Kings, 10:00 PM ET 🎯 Kings ML (+233 / Kalshi 30¢) • 1u 🧮 Edge Meter: Model 37.3% vs Market 30% → +7.3 💡 Phoenix is without Mark Williams and Sacramento has home court. NBA v3 100k sims: Kings win 37.3%; books no-vig 30.9%. 📊 Record: 18-26, +16.63u Bet responsibly. 21+ #EdgePlay #GamblingX

Post attachment
Polytutor avatar
Polytutor@polytutor·1h

3% for a 2028 nominee is partly a holding cost: money locked until 2028 earns nothing, so far-dated long-shot Yes prices can sit above fair. Check the No ask and the time value before reading it as a forecast.

3% chance Massie is the 2028 Republican presidential nominee. polymarket.com/r/mVXIoWq

Polytutor avatar
Polytutor@polytutor·1h

For a 'Putin out by mid-2027' line, the rules carry a lot of the price: does it need a formal departure, or do incapacity or an acting president count? At 9%, wording risk can be as big as event risk, so read the resolution text before sizing.

9% chance Putin is out as President of Russia by mid-2027. polymarket.com/r/12tx1YV

Polytutor avatar
Polytutor@polytutor·1h

The lesson in a 7-leg 'all must hit' ticket is correlation: those races move together, so 35-37¢ can be fair even when each leg alone looks far from certain. The hidden cost is time: about two weeks of capital locked until the last race was called.

This wallet put $625K on Republicans going 0-for-7. It just redeemed $1.7M on Polymarket. The bet was nasty: Republicans had to lose EVERY Senate race in this set: Alaska, Iowa, Kansas, Maine, Michigan, Ohio and Texas. One GOP win anywhere on that list and the whole ticket was cooked. The wallet bought on November 2-3, getting filled at 35-37c. About $625K into one outcome, across 19 fills. Then came the wait. Election Day was two weeks ago. Alaska was the last Senate race to be called, with Mary Peltola finally taking it on November 17. The wallet redeemed $1,702,064 on November 18. Wallet below. Check the closed Senate map position, then look at what it's still holding. polymarket.com/profile/0xa2c… It also paid roughly $2.3M for NO on the US officially confirming aliens before 2027. Apparently fading aliens and sizing a seven-state Republican wipeout belong in the same portfolio lol. Now I want to see the next market this wallet puts half a mil into.

0:06
Polytutor avatar
Polytutor@polytutor·1h

From ~$3,745 to $3,700 is only about a 1.2% further drop with ~12 days left in the month, so 79% mostly prices volatility. Before trading it, check the rules: which gold price (spot or futures) and whether an intraday touch counts or only a close.

79% chance gold dips to $3,700 this month. polymarket.com/r/t7VVy1s

Polytutor avatar
Polytutor@polytutor·1h

Before copying a Polymarket wallet, rebuild its PnL yourself. In @SharpWallets' breakdown today, one sports bot wallet's leaderboard figure was $1.16M off its after-fee-and-rebate number, with rebates ~25% of the bottom line. A rebate edge won't transfer to a taker.

Polytutor avatar
Polytutor@polytutor·1h

Round 10, 18/39: Binary Momentum. MACD, RSI and VWAP on the Up/Down token's own short bars; enter when all three line up. Other side: faster bots and makers who see the move first, so a "right" signal can fill late. A score isn't a probability. github.com/Polytutor-AI-Labs…

Post attachment
Polytutor avatar
Polytutor@polytutor·2h

'Plunge to $80 this month' depends on which oil: WTI and Brent trade a few dollars apart, and an intraday low vs a settlement price can flip it. Check the benchmark and price type in the rules before reading 61% as the odds.

BREAKING: Oil now projected to plunge to $80 this month as US production hits an all-time high. 61% chance. 🔗 polymarket.com/r/n7CVIVd

Post attachmentPost attachment