Buying is intuitive: you own a thing, you want it to be worth more. Shorting inverts that, and the inversion creates a risk profile most people never properly think through — one where the worst case isn't losing what you put in, but owing far more than you put in.
To short a stock, you borrow shares you do not own, sell them immediately at today's price, and hope to buy them back cheaper later. You return the borrowed shares and keep the difference.
Note what you owe: shares, not dollars. That single detail is the source of everything dangerous about shorting.
When you buy a stock at $50, the worst possible outcome is that it goes to zero and you lose $50 per share. Bad, but bounded — and you know the number in advance.
When you short at $50, your loss is whatever it costs to buy the shares back. There is no ceiling on a share price. If it goes to $150, you lose $100 per share: twice your original proceeds. If it goes to $500, you lose $450 per share.
Worse, the position moves against you in an accelerating way. As the price rises, your losing position gets larger in dollar terms, so each additional percentage move costs you more than the last. A losing long position shrinks as it falls; a losing short position grows as it rises. Risk that grows when you're wrong is the opposite of what you want.
A squeeze is what happens when that structural fragility becomes self-reinforcing.
Short sellers do not get to hold indefinitely. They post collateral, and when losses mount, their broker issues a margin call: post more money or close the position. Closing a short means buying the shares back. Buying pushes the price up. A higher price margin-calls the next short seller, who must also buy. That buying pushes the price higher still.
This is why squeezes look so unhinged relative to fundamentals. During one, the marginal buyer isn't an investor forming a view — it's a trapped short with no choice.
GameStop is the textbook example because it was so extreme. Measured from the actual daily closes used in this site's Mystery mode, over roughly three months from late November 2020:
| Measure | Value |
|---|---|
| Run-up from period start to peak | +1,998.6% |
| Peak-to-trough collapse afterwards | −88.3% |
| Net change over the whole window | +514.3% |
A roughly twentyfold increase, then an 88% collapse, then still up more than fivefold across the period as a whole. Consider what each of those numbers does to a short seller.
Someone short near the start faced losses approaching twenty times their initial proceeds at the peak. No ordinary collateral survives that; they were force-closed long before, their buying feeding the very spike that was destroying them.
But now look at the second row. Someone who shorted at the top — the trade that felt insane at the moment of maximum euphoria — would have been right by 88%. The squeeze punished the people who were arguably correct about valuation but early, and rewarded the people who arrived after the mechanical buying was exhausted.
Squeezes need fuel. The usual ingredients:
None of these predict a squeeze. Plenty of heavily-shorted stocks simply decline, which is why shorts were there in the first place. They describe a fragile configuration, not a forecast.
RugPull.trade lets you hold SHORT the same way you hold LONG, and the asymmetry is modelled where it matters: a short that goes wrong runs into the same forced-close logic real margin calls use. In Paper mode, a 📈 MOMENTUM SURGE warning is the squeeze telegraph — the moment to reconsider a short before the move happens rather than during it.
Mystery mode is arguably the better teacher here, because you trade a real historical chart without knowing which one it is. If GameStop's winter comes up, you will not recognise it from the price level — it is rebased — and you will have to decide, with no hindsight available, whether that vertical move is the start of something or the end of it. That is the actual problem shorts faced, and it is much harder than the story makes it sound.
Try shorting in the game →