ACCOUNT / @aravind PUBLIC POSTS ONLY PUBLIC SOURCE RESEARCH ↗

How he does it

Six patterns behind the hits

Most of his well-known hits are real, and some are impressive. Our data suggests they come mostly from a few repeatable habits rather than inside knowledge. Here is what we found, in plain terms.

The findings

01

The hits get resurfaced, the misses don't

When something happens, he quotes his old post. Readers see the hit; the misses stay buried. His Venezuela call (2 Apr 2025) got 2 quotes in its first three days. Its quote list fills up from 4 Jan 2026, the day after Maduro was captured. In our curated set, calls he later quoted hit 59% of the time. Calls he never mentioned again hit 18%.

The best call we found, Assad gone by the end of 2024, is real and was against consensus. We only know about it because he quoted it on the day Assad fell (1865780878242185641).

02

Extending what was already announced

Many hits extend a policy that was already in motion. The Venezuela post came nine days after Trump ordered tariffs on buyers of Venezuelan oil. "Trump will soon hit India's IT industry" (1953841130132677094) came two days after the 25% Russian-oil penalty on India; the H-1B fee came 42 days later. Calls on ideas already in the news hit 62%; ideas not yet in the news hit 27%.

03

Vague timing and conditions

"Soon" stretches. "Trump will be tamed soon" on tariffs (1950786078358565055) was followed by tariffs doubling four weeks later; the cut came about six months on. Conditions drop away too: a 2024 gold post said $5,000 "if" a recession came (1854071662481981644); when gold got there he restated it without the condition (2015639350273605699). Calls with a stated deadline hit 31%. Calls with only an implied "soon" hit 75%.

04

Repetition

Themes come back again and again: unrest he links to foreign plots, US pressure on India, regime change in the region. When unrest is frequent, a repeated forecast of unrest will look right some of the time. Of 21 specific India protest calls made before anything started, 1 was a clear hit and 3 were partial.

He celebrates early, too. He said a RAF C-17 would carry the stranded F-35 out of Kerala (1939412918107271268) and posted "how did you know?" (1940702525466071578) before the jet flew out on its own. To his credit, he does admit some misses: "My bold prediction went VERY WRONG" on Larijani (2033867975196037355).

05

Sources and timing

He rarely links sources: 44 external links in about 8,080 posts. He names outlets often, mostly to criticise them or to read them as a signal (1856147231424032826). On 12 of 12 major events we checked, his first post came after the first public report, with a median lag of 2.9 hours. In 9 of them, that first post quoted an older call of his. Across 904 quotes of other accounts, the median gap was 5.7 hours.

EventLagSelf-quoteLinks
06

The growth story

Median views per post went from about 16K in Oct 2024 to about 210K in Oct 2026, and median likes from 429 to 8,804. Growth came in steps: Venezuela in Apr 2025, the India tariff and H-1B news in Aug and Sep 2025, Maduro's capture in Jan 2026, and the Sept 2026 protests, when "What did I say?" (2096674761543934290) got 24,448 likes. The people quoting him are mostly small accounts (median 299 followers), not big amplifiers. Told-you-so posts don't do better on a typical day, but some go viral when they ride a big news event.

What this means

He reads the news closely, extends what powerful people say they will do, and repeats a few big themes. That gets many things right. In our sample, the record looks stronger than it is, partly because the hits get resurfaced and the misses don't.