I play a decent amount of Magic: the Gathering. One of the most useful concepts I've learned from playing – that's useful in both games and in life – is the idea that you should avoid being overly results-oriented – that you should avoid letting the results of a game affect the evaluation of the decisions that you made during that game.

This is an important principle in Magic because Magic is a game with a lot of variance and hidden information. It's possible for a stronger player to lose to a weaker one, a better deck to lose to a worse one, for a good decision to go poorly and a bad decision to turn out well. Large sets of results can be useful for evaluating long term performance, but the outcome of any particular game says basically nothing about your skill as a player, or the quality of the decisions you made during the game. If you constantly adjust your decisions based on results, you'll make decisions effectively at random, and your performance suffers.

For example: You're playing a game in a format that has just one very rare card that "wipes the board" – removes all the played cards from the battlefield. There's a less-than-one percent chance that your opponent even had the opportunity to put that card in their deck. You draw a good hand with lots of cards to play and the resources needed to play them, and you play your cards out "on curve" – playing each card as soon as possible, for the most efficient resource usage. Usually this will allow you to win the game easily – Magic games often come down to "who did the most stuff."

Instead, your opponent wipes the board, removing three of your cards from play in exchange for one. At this point you've already played out your whole hand, and you don't have much else left to play.

Did you play incorrectly? Should you have held a card back? No! The vast majority of your opponents won't have that particular card. If you play again against that particular opponent with that particular card, you might hold something back in reserve, but that will reduce your chances of winning any game where your opponent doesn't have that card. Your decision was correct based on the information that you had at the time.

The alternative to results-oriented thinking is process-oriented thinking. With process-oriented thinking, your focus is on the quality of the decisions that produce the result – the part of the process that you can control, instead of the random variables that you don't control.

This is a little bit counter intuitive, and it's counter to a lot of business advice, which praises bosses who are "results-oriented" and don't fuss too much about the details. Likewise, it runs counter to standard performance management doctrine, which says that we should reward employees based on their impact. "Outputs over inputs."

Personally, I think this explains a lot that goes wrong in businesses – and a lot of bad decision-making generally. While it's definitely possible to take being process-oriented too far, and you need to pay at least some attention to results in order to evaluate your process, excessive results-orientation makes it easy to let random or inessential factors control your decisions. It causes you to learn inappropriately from failures where random chance overcame good decision-making – and fail to learn from "success" where bad decision-making was saved by random chance – or even by a counteracting failure.

Imagine a general who receives a report on a friendly fire incident. The report notes that the casualties were limited, due to a malfunction in the weapon system. Should the general be happy about this? If you were doing a retro, would "weapon system failed to work" go in the "happy" column?