In May 2026 we asked an Instagram audience to find a partner in the comments. The mechanism was two words. Comment OPEN to offer yourself as a partner. Comment TAKEN to claim someone who had. One OPEN equals one TAKEN. No matching system, no list, no host. The round would pass at 100 active links.

Of 3,694 comments on the pairing post, 2,543 were OPEN and 956 were TAKEN. Five people offered a slot for every two who took one. The round passed its bar anyway, because two in five was enough to make 100 links.

This paper is about the gap between offering and claiming. It is the same crowd, the same thread, the same cost in keystrokes. The difference is what the word binds you to.

The round before this one showed a crowd could converge on a number with no channel. This one asked whether a crowd could do the harder thing: form pairs with no channel. Convergence needs everyone to pick the same answer. Pairing needs two strangers to pick each other.

• Instruction: "Find a partner. One OPEN equals one TAKEN."
• Where: Instagram. A pre-attempt post on 11 May 2026 explained the mechanism, the pairing post went up on 12 May, and the pass was announced on 14 May.
• Pass condition: 100 active links, declared in the thread.
• How we counted: every comment on the pairing post. OPEN and TAKEN were counted as written. A comment that tagged another player was counted as a third category, since tagging is a pairing move that uses neither word.

Economists have a name for the first kind of word. Cheap talk is a message that costs the sender nothing and binds them to nothing, and the classic result is that people produce a lot of it. OPEN is cheap talk. It signals willingness without choosing anyone. TAKEN is the opposite: it names a person, it can be seen by that person, and it can be refused or ignored in front of everyone. That is a small social risk, and 62 percent of the people who offered never took it.

The 615 tags are the crowd routing around the mechanism. Tagging a friend is pairing with someone you already trust, which removes the risk the round was built to measure. It is also the move the platform rewards, since a tag pulls a new person into the thread.

The round passed. That is worth saying plainly. Two in five was enough, and 100 links formed between strangers with no host. The crowd can do the harder thing. It just does the cheap half of it first.

• A claim is not a link. TAKEN comments were counted as written. Whether each claim matched a live offer, and whether the pair did anything after, is not in this count. The 100-link pass was declared in the thread, not derived from these numbers.
• Order and visibility were not controlled. People who arrived late saw a wall of OPEN and may have read the round as already full or already failed.
• One round, one audience. The ratio is from one thread on one account in May 2026.

The same mechanism with the cost moved. Make OPEN the binding word and TAKEN the free one, and see whether the ratio follows the cost or the word. If it follows the cost, the finding holds. If it follows the word, something about claiming itself is the problem, and that is a different paper.

The round this paper reads is 3 posts, not one. Every post is below with the platform's numbers and what the comment archive holds of each.

The pairing round has no Chronicle entry of its own yet; the round record on The Game holds it. The convergence it built on is filed as THE NUMBER, ⟡035.

The instruction as it went out: the reel frame beside the caption, captured from the operation's own signed-in account. What the paper counts is the answer to exactly this.

Native captures of the comments this paper leans on, taken from the operation's own signed-in account on 31 Aug 2026. Handles and like counts are as they stood that day; the numbers in the paper are the round-time count.