That version is also obviously wrong. It is possible for individual people to more correctly analyze the value of an asset than the stock market. People who looked carefully into the mortgages underlying mortgage-backed securities during the housing bubble (e.g., Michael Burry) are a good example of this. They clearly were not right "by accident" - they were right because they carefully analyzed the underlying assets, at a time when most people were not doing so.

In order to make the EMH plausible, you have to restrict it so much that it's essentially meaningless: something on the order of, "If you take the ensemble of market players, they do not beat the market on average." But it's clear that individual market players do often better understand the value of particular assets than the market, and not just because they're lucky, but because they just do better analysis than the crowd does.

This doesn't prove the EMH wrong. In order to prove EMH is wrong you need to show someone like Michael Burry can consistently outperform the market using only publically available information. Making one good trade (no matter how large) isn't that.

In fact, following your reasoning every trade would invalidate the EMH. If the price goes up afterward then the buyer was "right" and if the price goes down afterward then the seller was "right".

One trade is enough, if the reason for that trade is something other than dumb luck. The idea behind demanding consistency is that you can get lucky with one trade, but that over the long run, your luck will average out, and your true ability will show. The EMH says that your true ability to judge value won't be greater than the market's ability to judge value. But there are cases where it's clear that the trade wasn't merely luck. That's why I gave the example of the housing bubble: people using only public information were able to better judge the value of MBSs, and it wasn't just luck. They were objectively smarter than the market, because they did something that most people weren't doing: they analyzed the underlying assets. Put another way, the market can behave irrationally, and can seriously mis-price assets. In fact, it often does so, and the reason is not simply that there's missing information. The reason is that people are irrational, lazy, hubristic, etc.

> In fact, following your reasoning every trade would invalidate the EMH.

No, only trades in which the person making the right call was right by something other than luck.

The strongest version of the EMH that actually holds is probably something on the order of, "If you pick a random investor, their expected returns are the same as those of the broader market." That is, you have to weaken it until it's trivially correct.