A decision is not its result
Imagine two players face the same situation. Each has a roughly 70 percent chance of winning the exchange if they commit. Both commit. One wins; the other loses. Afterwards, the winner is likely to feel that the play was smart, and the loser is likely to feel that it was a mistake. But they made the same decision, and it was the same quality.
That gap is the core of what is called results thinking, or outcome bias: judging the quality of a decision by how it turned out instead of by what was known and weighed when it was made. It is one of the most common reasoning errors in any field with uncertainty, and competitive games, with their mix of skill, hidden information and chance, are full of it.
Why outcome is a poor judge
A result combines at least three things:
- The quality of the decision, given the information available.
- Chance, including random effects and things the opponent did that you could not predict.
- Execution, meaning how well the plan was carried out.
Only the first is a decision-making skill. If you judge only by the result, you mix the three together. Over a large number of attempts the mix tends to even out, and good decisions usually do better than bad ones. In any single case it can point the wrong way.
This is why the same choice can look brilliant on Monday and foolish on Tuesday. The choice did not change. The dice, or the opponent’s guess, did. Understanding how much random variation sits inside outcomes, including something as simple as how often random clumping makes a run of identical results appear, makes it easier to treat any one result with suspicion.
The two mistakes it causes
Abandoning good plays after bad luck. A player tries a sound but risky play, it fails twice, and they conclude it does not work. They drop it, and lose a useful tool. Two attempts is a tiny sample size, and many good plays fail regularly. A play that works 60 percent of the time fails 40 percent of the time, which is a lot.
Repeating bad plays after good luck. The reverse is more dangerous because it feels good. An unsound play that happens to work gets reinforced. Players say “it worked” and the habit is kept, until someone competent punishes it. This is also one route to overconfidence after a win streak.
Results thinking can also affect how you judge other people. A teammate whose gamble fails gets blamed, and one whose gamble succeeds gets praised, even if the decisions were identical. This feeds the patterns discussed under blame bias.
What a decision process looks like
Judging by process means asking whether the decision was reasonable given what you knew. A simple version has four questions:
- What information did I have? Include what you could see and what you could reasonably infer, but not what you learned afterwards.
- What were the realistic options? Did I consider more than the obvious one?
- What were the likely outcomes of each, and how bad was the downside? The idea behind expected value is that a choice should be weighed by what you stand to gain or lose and how likely each is, not only by the most likely result.
- Was this consistent with a plan or principle I would endorse in general?
If the answers hold up, the decision was probably sound, even if it lost. If the answers show that you ignored something visible, panicked, or took a risk without a reason, the decision was weak, even if it won.
A worked illustration
Suppose a play has a 60 percent chance of netting your team a useful advantage and a 40 percent chance of losing a key resource, and the alternative is a safer play with a modest, near-certain gain. Whether the risky play is better depends on how large the gain and loss are and what the match situation demands. It would be wrong to say it is good because it worked or bad because it failed. Even a good play at 60 percent will fail on roughly two attempts out of five, so a run of two failures is not unusual: with independent attempts, two failures in a row happen about 16 percent of the time (0.4 multiplied by 0.4 is 0.16). That is not rare enough to conclude anything.
The numbers here are illustrative, and real situations rarely come with clean percentages. But the habit of estimating roughly, and not reasoning backwards from the result, is the point.
Making it practical
Review before looking at the result when you can. In a structured replay review, pause at key decisions and write down what you would do before seeing what happened. This protects against hindsight, where the outcome makes the right choice look obvious.
Keep a short decision log. For important choices, note the situation, the options and your reasoning. Later, you can judge the reasoning without the result changing your memory. Simple notes work better than recollection.
Look at repeated patterns, not single outcomes. If a type of decision keeps failing across many cases, that is evidence worth taking seriously. A single failure is not. The same principle applies to testing changes in your play, where sample size should limit how much you conclude.
Be willing to say “right call, wrong result.” This phrase is surprisingly uncommon, and surprisingly useful. It lets you keep a sound play, and it takes some of the sting out of losing.
Also say “wrong call, right result.” This is harder, since it takes away a win’s satisfaction. But it is where much of the learning is.
Limits of the idea
Process thinking can be misused. It can become an excuse: “my process was fine” is easy to say about any loss. A real process check has to be willing to find fault. Also, when outcomes repeat, they are information about whether your model of the game is accurate. If a play you consider sound keeps failing, one possibility is bad luck, and another is that your estimate of its odds is wrong. Good process includes updating on evidence, only more slowly and carefully than a single result would suggest.