Unlikely is not the same as impossible
Games are full of moments where the odds say no and the result says yes: a last-second comeback, a rare drop, a single die that saves a lost position. Anyone who has watched video game luck produce an absurdly rare item knows the feeling. These stories stick because they are rare. They also tempt players to treat a small chance as a plan.
The useful distinction is between three categories. An outcome can be impossible under the rules, very unlikely but possible, or merely uncommon. Most arguments about what “can happen” in a game are about the middle category. If the rules allow it, it has some probability; the real question is how much, and what you give up by chasing it.
How small is small?
People tend to be poor at feeling the difference between small probabilities. One in fifty and one in five hundred both register as “very unlikely”, yet one is ten times as common as the other. When you are weighing a long shot, put a rough number on it, even if it is a range.
Dice odds are a good place to build that sense, because the numbers are exact. Some anchors from common equipment:
- A single fair six-sided die showing a particular face: 1 in 6.
- Two fair dice both showing a particular face: 1 in 36, since 6 x 6 = 36.
- Three fair dice all showing the same particular face: 1 in 216, since 6 x 6 x 6 = 216.
These are exact for fair dice and give you a sense of how quickly chances fall away when several things must go right together. That last point matters: a comeback that needs a chain of favourable events is much less likely than any single link suggests, because the chances multiply.
Repeated attempts add up
A low chance is less hopeless if you can try many times. If something has a 5 percent chance per attempt and attempts are independent, the chance of at least one success in 10 attempts is 1 minus the chance of failing every time:
- Chance of failing once: 0.95
- Chance of failing 10 times in a row: 0.95 to the power 10, about 0.60
- Chance of at least one success: about 1 - 0.60 = 0.40, or roughly 40 percent
With 20 attempts, failing every time is about 0.36, so success at least once is about 64 percent. Even then it is far from guaranteed, and many players find that surprising in both directions: it is more likely than a gut feeling suggests, yet less certain than “I’ve tried so many times, it must come soon”.
That last thought is a trap. Each independent attempt starts fresh. A long run of failures does not make the next attempt more likely to work, a mistake sometimes called the gambler’s fallacy. The cumulative chance rises as you add attempts, but the chance on any particular attempt does not.
What matters is the price of each try. If attempts are cheap and the prize is large, a long shot can be sensible. If each try burns a limited resource, repeated tries may drain you before the odds have time to work, which is where good risk management comes in.
When a comeback is realistic
Being behind does not tell you how far behind. A useful way to think about a comeback is to separate the size of the gap from the amount of game left.
- A small gap with much time remaining is usually recoverable through ordinary good play, with no long shot required.
- A large gap with much time remaining may be recoverable if the game has mechanics that swing momentum, such as catch-up rules or scoring that rewards a single big play.
- A large gap with little time remaining is where a true long shot lives. Steady play will almost certainly lose, so a low-probability line may be the best available option even though it is likely to fail.
In that last case, a long shot does not have to be good in absolute terms; it only has to be better than the alternative. If the safe line has essentially no chance, a 5 percent chance is an improvement. This is the logic behind taking more variance when behind.
Before committing, it is worth checking your read of the position. Players often overestimate how lost they are or how winnable a swing is, so an honest look at the game state before declaring a situation hopeless (or promising) is time well spent.
Ways to improve a long shot
A long shot is not fixed. Sometimes you can raise its chance, or lower its cost, before you take it.
- Stack small improvements. If a bonus, a re-roll or a favourable position adds a few points of probability, those can double a very small chance.
- Reduce the cost of failure. Preparing a fallback means that if the long shot misses you are no worse off than before.
- Choose the moment. Some long shots are better attempted early, when you still have resources to recover, and some late, when there is nothing to protect.
- Seek information. Learning more about the opponent or the situation can turn a vague hope into a measured estimate. Strong opponent reading often reveals which risky lines are less risky than they appear.
Memorable is not likely
Dramatic upsets are remembered and shared, while the far more numerous failed attempts are quietly forgotten. This makes unlikely events seem more common than they are. A highlight reel of improbable wins says very little about how often they occur, and it is no substitute for looking at real win rates over many games, and the same is true of personal memory: you recall the time it worked.
If real money is at stake, this effect needs extra care. Stories of unlikely wins say nothing about your own chances on the next try, and a long shot with negative expected value stays that way however attractive the payoff looks. It is reasonable to enjoy a remote chance for entertainment, but not to rely on it.
Putting it together
When you face a long shot, a short routine helps:
- Estimate the real chance, using a range if you must.
- Work out what each attempt costs and how many you can afford.
- Compare it with your other options, not with perfection.
- Decide in advance what you will do if it fails.
Often the answer will be that the long shot is not worth it. Sometimes, particularly when the alternative is a near-certain loss, it is. Either way, you will have chosen based on the chances rather than on hope alone, which is the closest thing to “making the impossible possible” that probability allows.