Same average, different experience
Two options can offer the same expected value and still behave very differently. Consider a strategy game where you choose between two attacks. These numbers are an illustration.
- Steady attack: always deals 5 damage.
- Swingy attack: deals 10 damage with probability 1/2, and 0 otherwise.
The swingy attack averages 0.5 × 10 + 0.5 × 0 = 5 damage, the same as the steady one. Over a very large number of uses they would be hard to tell apart on average. In a single use they are not: one is certain, the other is a coin flip between a strong hit and nothing.
The gap between those experiences is variance, and how you feel about it is your risk tolerance. Both are legitimate. Choosing well means knowing which one your current situation rewards.
Measuring how swingy an option is
Variance is a measure of how far results typically fall from the average. For the swingy attack, each result is 5 away from the mean of 5, so the squared distance is 25 either way, and the variance is 25. The steady attack has a variance of 0. The standard deviation is the square root of the variance, which for the swingy attack is 5 damage, and the steady one is 0.
You do not need to calculate this every time. The practical point is that a higher standard deviation means more spread: the best and worst cases are further apart. Game designers deliberately adjust this, which is part of what is meant by game volatility.
When the situation changes the right choice
The “best” option often depends on where you stand, not only on the averages.
When you are ahead. If you are leading, you generally want fewer ways for the game to go wrong. A steadier option reduces the chance of a reversal. The averages are still the same, but you no longer need a large swing.
When you are behind. If steady play is unlikely to close the gap, a higher-variance option can be sensible, because the steady path may have a very low chance of winning at all. Variance creates the possibility of a comeback that a certain small gain does not.
When the target is a threshold. Imagine you need exactly 10 damage this turn to defeat an enemy. The steady attack never reaches it. The swingy attack reaches it half the time. Here the average (5) is irrelevant; what matters is the chance of reaching the threshold.
This is why an average should not be the only number you look at. Compare options by the probability of the outcome you actually need.
Risk tolerance is about the situation, not personality
People often describe themselves as cautious or bold players. That can be a useful starting point, but the better habit is to ask what the position demands. A cautious player who always plays safe will sometimes fail to take a sensible gamble when behind. A bold player who always gambles will sometimes throw away a lead.
Some other factors worth weighing:
- How many chances remain. With many turns left, a bad result can be recovered from, and the average matters more. With one turn left, variance can matter more.
- How costly failure is. Losing a minor resource is different from losing a character, a piece or a round.
- What the opponent is likely to do. In games where opponents react, predictable play is easier to counter, which is where ideas like mixed strategies come in.
- How reliable your estimate is. If you are unsure of the real chance, a guess about a swingy option can be badly off, so a steadier option has a hidden advantage.
Stacking risks in video games
Many video games combine several random elements, such as a hit chance and a critical hit chance. Each layer increases the spread of results. A character with a modest chance to deal a large critical hit is a swingy choice; a character with reliable moderate damage is a steady one. When you choose a build or a move, it can help to ask which kind of result you need, particularly when facing risk under uncertainty about what the opponent will do.
Also keep in mind that streaks happen in random sequences. A string of misses does not mean the next attack is “due”. If the game uses fair independent chances, the odds stay the same each time.
Where this does not apply
Everything above is about games where the stakes are in-game. In gambling with real money, the situation is different: every wager has a house edge, so the average result for the player is a loss. A “high-variance” bet is not a smart tactic when you are behind; it is just a larger bet with the same negative average, and it can speed up losses. Casino games are not a place to apply comeback logic, and playing more to recover a loss is a well-known way to make things worse.
If you gamble, decide in advance how much time and money you will spend, and stop when you reach the limit. If gambling is causing harm or you feel unable to stop, contact a local support service.
Summary of the idea
Before choosing, compare the average, then ask how much spread you can accept and what the position needs. If you are ahead, prefer fewer ways to lose. If you are behind, consider whether a steady path can really win. In both cases, the result of a single play is uncertain, so judge the choice by the reasoning behind it, not by one outcome.