Do you fall into this trap?

[Higher Grounds]

Players will always do the thing you didn't expect. I always find that it's best to give your story broad strokes and problems, and surf the wave as your players create the biggest portion of content.

 

[Sunseeker]

Early on, I create conflicts. The real stories in my gameworld don't really kick in until the players get a little rep for themselves. (like level 5). This gives me a chance to gauge how my players are going to react in the future to similar stimuli. Some players are predictable, some players aren't.

When creating a game, I always start with the "what", in this case that's the conflict.
"An ancient artifact sits in front of you, slowly growing warmer and brighter."
Okay, now I look at the situation and address the "why".
"A wizard created a device of unknown power and his fried corpse is in the corner."
Now that I know what the problem is, why it's a problem, I look for answers. Obviously an arcana check exists here, but what if noone is good with Arcana? Okay, how about a "use device" type check? What about just smashing the dang thing?

I can't respond to the players actions if I don't know what the possible outcomes of these questions are. D&D doesn't run itsself. Maybe the kobolds run away when the Paladin crit one-shots their leader, maybe they go kamakazi. These are decisions I make beforehand.

Otherwise the players are fighting meaningless battles for no reason.

 

[Vladar]

I usually try to implement a couple of different ways to solve a dungeon, but always leave a place to player's personal plans and ideas.

 

[Gamgee]

I improvise almost everything so my games are truly open world if players are creative enough. Usually though they do "typical" things that aren't all that crazy. It's like they learned to railroad themselves at this point.