"Fair launch" is used the way "open source" was used a decade ago. As a marker word. It does work for the people saying it whether or not it's accurate.
The loose meaning is: anyone could have bought, at the moment everyone else could have bought, at the same place. That's the broadcast version of fairness — equal access to the surface. It's necessary. It's not enough.
The structural version is narrower. At t=0, the seller side has no information advantage. That one sentence does most of the work. It's the test the broadcast version doesn't run.
Most launches that call themselves fair fail it in one of three places.
The first place is timing. Someone always knows the deploy block before the public does. The team running the contract knows. The people the team told know. Anyone watching the team's wallet activity knows. A launch where the person pressing the button has been lining up positions for an hour is not fair on the seller side, regardless of how the buy is opened to the public on the buyer side.
The second place is allocation. The moment any of the supply moves before it goes on sale — to a team wallet, to advisors, to "marketing partners," to the deployer's second address — the seller-side information advantage shows up as a position advantage. Fairness in pricing requires that no seller has a position before the buyers do. Once that's broken, calling the rest fair is doing rhetorical work the structure isn't doing.
The third place is infrastructure. Some launch venues sell the ability to act on launch information faster than the broadcast travels. Snipers, priority lanes, observability tooling pointed at mempools. The surface looks fair — anyone could have bought — but the people positioned to act on the deploy event existed because the venue made room for them. If the venue's design has a class of buyer it favors, the launch on that venue inherits the favoritism whether or not the launch team chose it.
The broadcast version of fairness is easy to claim. The structural version is something you have to design for, before the day in question. Every place a launch chooses convenience over structural fairness, the convenience accrues to whoever knew about it first.
The useful question to ask, of any launch using the phrase, is not whether it was fair. That's the framing the phrase is trying to short-circuit. The useful question is: at t=0, was there a person with information about this launch who could act on it, that the public didn't have? If yes, which person, how long had they had it, what did they do with it. If the answers are clean, the phrase fits. If the answers are murky, the phrase is doing the work the structure can't.
Words like this drift. The right response isn't to abandon them — they're useful, they describe something real — but to use them as questions rather than as descriptions. Fair launch as a question is sharp. As a label it's almost always overstated.