A topic that comes up time and time again on my stream is the concept of playing to win versus playing not to lose, with the former being generally seen as better. Lines that play to win generally involve figuring out what would need to happen for you to win, and then optimizing for that. Playing not to lose is broadly seen as trying to predict ways your opponent could win and optimizing to avoid them. This dichotomy has always confused me, since those things have always seemed like two sides of the same coin. In a given situation, there’s just an optimal play, and using a heuristic like whether it plays to win or to not lose to determine what it is has seemed arbitrary and incorrect to me.
A better conception of this problem is instead looking at playing safely versus playing scared. This looks at the concept of “playing not to lose” and breaks down whether it’s justified or not in a given game. There are games where it is genuinely correct to play around the potential outs your opponent could have, and these lines can be considered playing safe. Playing scared involves taking the very same caution lines, just in situations where it’s unreasonable to do so. What it really comes down to is differentiating these two situations.
One of the times it’s most likely to be correct to play cautiously is when you’re significantly ahead. In this situation, if the status quo is maintained, you will win, so you need to try and preserve it rather than find a way to change it to your advantage. If the answers to “how does my opponent win this game” are fairly few and preventable, it may make sense to go out of your way to play around them just because you can. For instance, if you’re playing a game of Limited and are far ahead on board and life totals, you may opt to avoid playing into a top-decked board wipe just to play it safe, since it’s the only way you can realistically lose.
A common mistake I see people make in this situation is making plays that don’t actually meaningfully change how likely they are to lose. These plays are generally fairly nearsighted and focus on playing around something for a turn cycle or two while still not having the ability to beat that in the long term. For example, leaving up a counterspell rather than playing out a threat when dead to a single burn spell, and being unable to play a threat out for the next several turns while also holding up a counter. In this situation, you need to get a threat down to actually close out the game, and while leaving up a counterspell may seem safe, in the long term a burn spell in hand will be very hard to beat and you’re just giving your opponent more time to draw multiple. This leads into the next point: you often can’t play scared when you’re under pressure to end the game.