In my opinion, the point where it all went wrong was when Melvin announced that they had closed their short position and the WSB hivemind decided that they were lying to the public. The conversations after that became increasingly conspiratorial and divorced from reality.
Although it of course seemed quite strange at the time that they would publicly announce that they had closed their positions, unless they wanted the price to drop... heh, it was a mess.
In a sense it was the perfect combination of "limited knowledge" (because only a select few had access to Short Interest estimates) and "limited understanding" (because people thought an unchanging level of Short Interest would necessarily imply that no shorts had closed their positions -- when in reality it was much more likely that new, quite reasonable shorts had instantly taken their place).
Ironically though, I think there could have still been a larger squeeze simply because there still was plenty of shorted stock and people continued to buy, not understanding that their main "opponent" had already tapped out.
And then what? How long is meme inflation good for? AMC managed to issue stock to cash in; how long would it have been before GME's management sold out and also did a share issue?
> variant of the prisoner's dilemma
It's an amazing one where you can meet someone on the virtual street, con them into working with you, defect against them, and leave them worse off.
And then, I suppose, the expectation was to sell most shares to desperate hedge funds and maybe keep one or two because you actually like GameStop. The idealistic and completely unrealistic narrative being that it would only be short sellers that pay for the redditor's profits, of course. I think GME wouldn't really have had the opportunity to capitalize on the situation, because them issuing shares would have angered basically 95% of reddit -- and if you're dependant on gamer's liking you, that would probably be a death sentence.
> con them into working with you
I'm not sure how much active misleading was going on, when the narrative that most people believed allowed for motivating people to buy/hold without any nefarious motives.
Just like in a pump-and-dump, some people were always going to be bag holders with GME, namely those that misunderstood what 'short interest over 100%' meant and thought they could set their price at $42069.
But I still think there was an important difference: The point at which the best move switches from cooperate to defect comes sooner in a regular PnD, because there is no 'outside money' to be gained when cooperating for long enough. So, in a PnD, you should probably sell as soon as the momentum has really picked up, whereas in this case, the momentum actually increased the chance of cooperating being a successful strategy -- up to a point that would have had to be determined.
This was, originally, not a 'pump and dump' that everyone with two working brain cells should immediately have recognized as one. Buying into a pump'n'dump when it's already rising is not a smart decision -- but buying into a short squeeze of potentially HF shattering proportions, even after the underlying stock has already gone up dramatically in value, is a quite different scenario, and rationally justifiable imho.
The point where it all went wrong was probably the buying restrictions for retail customers that killed the momentum completely. Without that element, ridiculously high meme prices like $1k per share could have been realistically achievable, driven by FOMO and further fueled by Melvin-copy-cat hedge funds that might actually have been margin called at those ridiculous prices.
So, I'm a bit disheartened by the view here that 'of course it was always going to go wrong, what did they expect'. It definitely was naive to believe everything would go 'smoothly' when billions are on the line, and markets are put under stress they might have never seen before. But the whole thing wasn't destined to fail like it did.
Looking back, I believe the only real mistake many made was not realizing the variant of the prisoner's dilemma that people were in had rapidly shifted on Thursday to favor defecting, even if just to let the dust settle and reevaluate their position. But when you're caught up in an echo chamber that still promotes 'holding' above all else, it becomes difficult to realize that, since holding would indeed have been the right decision if everybody else had done the same. In retrospect it's of course ridiculous to expect a huge group of ultimately self-interested people to coordinate on that level. Still, people held because they got caught up in the herd mentality. After that point though, it indeed turned into a serious of small pump-and-dumps that one should have better left alone.
All in all, I think people shouldn't beat themselves up too badly about what happened, because there _was_ no good template for any of this that one should have consulted.
It was at least a very interesting lesson in stock market mechanics and the kind of manipulation that people with tons of money will engage in when cornered. The decisions leading to retail being prevented from buying may or may not have been indirectly influenced by a motivation to stop the squeeze, but the 'WallstreetBets has moved on to Silver' was definitely a widespread disinformation campaign -- kinda cool to see one of those in action for once.