Written by: Mahe, Foresight News
Polymarket hides many top smart minds capable of making money. A user with the nickname noovd (0x063aeee10fbfd55b6def10da28e87a601e7deb4b) has earned hundreds of thousands of dollars simply by predicting the number of Elon Musk’s tweets on Earth.
According to the official homepage, he only entered the market in July 2024, and now has a net profit of $345,000. This user has made a total of 1,281 predictions, with his largest single win exceeding $136,000, and he still holds a position worth 6,877 tokens.
Looking at his profit curve, the first few months are almost flat, but starting from late 2025, several nearly vertical steep slopes suddenly appear. Each surge comes from the same source: betting on how many tweets Elon Musk will post each week.
As the world’s wealthiest billionaire, Musk naturally has no motivation to manipulate the market.
He doesn’t guess NBA scores, nor does he bet on political events; he focuses solely on this track. Opening his settled records, his top eight profits are all “Will Musk tweet between X and Y times this week?” Yes.
Polymarket weekly breaks down Musk’s tweet count into over a dozen mutually exclusive intervals, each spanning about 20 tweets—200-219, 220-239, 420-439, and even extreme ranges like 850-874. Retail traders see so many options and casually pick a middle range, buying at 35-50 cents for some psychological comfort. Those edge intervals? Their prices often drop to 2-24 cents, even though the true probability is clearly 15-30%.
noovd specializes in these undervalued cheap options. He only buys Yes, never No, never sells early, never hedges. When he places bets, his positions range from $4,000 to $30,000, with an average entry price between 2 and 39 cents. Once his model confirms a prediction, he goes all-in and waits for the Sunday settlement. His strength lies in a real-time rhythm model that others can’t see. He doesn’t rely on mystical “Musk might tweet more this week” guesses but feeds every day’s tweet data since 2022 into his model, updating daily.
The weights of catalysts like Musk’s Monday-to-Friday posting habits, weekend lulls, Starlink launches, Tesla earnings reports, DOGE news, and xAI new features are all quantified. By Wednesday or Thursday, he can accurately calculate how many tweets are needed each remaining day to land in a specific 20-tweet interval.
While the market still guesses based on historical averages, he has minimized the error to within ±10-15 tweets. If he finds that a 2-cent interval has a true probability of 18-25%, he directly takes a large position. The expected value of a single bet often reaches 8-20 times. A big win can wipe out previous small losses and generate substantial profits.
For example, in the prediction for December 9-16, 2025, Musk had already tweeted over 130 times by Wednesday night. To land precisely in the 260-279 range, he needed to tweet 32-35 times daily for the remaining four days. That week, there was no explosive news, but discussions about xAI and Tesla continued. The market priced this interval very low because everyone thought it would be either a dull 180-220 or a surge over 400.
noovd’s model perhaps captured Musk’s “intermittent activity” pattern. He bought in at an average price of 2 cents, investing about $1,569, and ultimately earned $67,686.33, a return of 4311.87%.
He entered when the odds were highly skewed, locking in this narrow profit zone at minimal cost. This is a typical “deep out-of-the-money option” approach, betting on the long tail of the probability distribution.
This case alone yielded an astonishing 43-fold return, and the next example became his most profitable bet.
From December 2 to 9, did Musk tweet between 420 and 439 times?
He bought in at an average price of 24 cents, investing $49,429, and ultimately earned over $130,000.
At that time, Musk’s tweet count had already exceeded 180 on Tuesday. In the following days, he needed to maintain 65-70 tweets daily to stay within the 420-439 range. That week, amid election aftermath and product hype, the market priced this at 24 cents. But he saw that others spread their positions across 400-419, 420-439, and 440-459 ranges, while the most likely outcome was the middle one. He bought all 208,694 shares at a cost of $49,000, settled for $186,000, nearly tripling his money. These two trades in December alone totaled over $255,000. The rest of the thousands of smaller bets were mostly testing or auxiliary bets on similar ranges, continuously fueling his model.
Why can this strategy keep earning, and why is it so hard for others to copy?
First, the information barrier is very high. You need to monitor the timeline daily, ideally have a complete historical dataset, and write a Poisson distribution model that updates every 6-12 hours. Most people can’t scrape data or are too lazy to tune parameters. The psychological hurdle is high. Intervals like 2 cents or 7 cents seem like throwing money into water. Those who haven’t experienced 100 weeks of drawdowns or seen 100 settlement results firsthand won’t dare to bet heavily. When you see screenshots of his settled losses, you might breathe a sigh of relief—he’s not 100% perfect.
Second, execution discipline must be extreme. Only when the advantage exceeds 10 times does he go big; small losses are paid as tuition, big wins are compounded for exponential growth. Most importantly, market inefficiencies still persist.
From noovd, ordinary traders might learn that instead of trying to sweep all markets, focus on a narrow, repeatable signal and master it. Depth always beats breadth, as long as your advantage is real. Abandon feelings entirely; let numbers and data speak. Make spreadsheets, scrape data, run regressions—these are a thousand times more reliable than “I feel Musk will tweet a lot this week.”
A 2-cent interval with a true probability of 20% is far better than a 50-cent interval with only a 5% edge. Most people do the opposite. In a casino shouting for 100x returns, the quietest and most ruthless wealth is earned by those who can count tweets better than everyone else.
Now, the weekly Musk tweet count prediction intervals are still being accurately updated. The only question is, how many can turn this into their own printing press?
