Creating Encounters means to craft a situation that presents to the players, choices with significant, often uncertain, consequences.

Seems simple and direct, but note that you aren't just creating a building and filling rooms with monsters. You aren't even doing the extra step of developing relationships between those rooms. What you are doing is making judgments about what is relevant and why. 

Yes, if you consistently drill down in play on random 'realistic' things then something interesting might occur eventually. But then how many hours of table time is wasted doing things that don't lead somewhere?

The assumption that this somehow limits character choice is incorrect also. I'm talking about what you design and choose to present as important. This in no way means that my players can't pull over random townsperson X and talk to them or grab some bedframe in a dungeon and break it up to burn it. 

"You see a bedroom, with several items of note. A pair of boots lie on the floor, an ancient oak armoire stands next to the window, and a silver sword hangs on the wall."

Don't bother describing them. Like an artist, what you choose to present is your work. It is ok to have an item be Nothing in order to increase suspense, but there is no reason why you must present every last thing in the room and make them go through each and every item. Decide as a designer what interesting choices their are and make the items that are designed to be Nothing seem suspicious so they do their job of increasing tension and acting as red herrings. Note specifically that uninteresting and unimportant items designed as Nothing are exactly that. The (non-described) bed in the example situation is a waste of everyone's time.

"There is nothing of particular interest in the west wing."

Why is this room empty? We are all familiar with good solutions for dealing with empty rooms. Instead of tediously forcing the players to explore each room, present the room or series of rooms as a single interesting choice. "It will take you X turns to toss the rooms for anything interesting. This will result in Y wandering monster checks."

But in reality, this is a way of avoiding responsibility for bad design. In modern modules this is "solved" by creating nothing but rooms that are encounters. This is also dull. By removing nothing (and often tricks, and traps as puzzles) the game loses variation and becomes tedious.

The same thing that happens with an abundance of empty rooms.

The idea is to not have the nothing option be boring, but have empty items increase tension every time they appear. If your players are bored with empty rooms it is because the nothing option is too common, poorly implemented, or threats they encounter aren't serious, frequent, or severe enough to worry about their choices.

A lot of people have asked "isn't this how I'm running my game now?" I ask them, is it?

Do you create two nothing items in a room deliberately, one that looks like treasure and another like a trick?

Do your players walk away from empty rooms convinced that they have missed something or thankful they didn't die?

Are you using these empty rooms with purpose? Because if you are, searching empty rooms should never be tedious.

"You see what looks like a golden shield hanging on the wall."

When designing these encounters things things seem pretty simple, because you know what is going to happen. The players have no idea what the encounter conceals. They parse the basic information you give them into a guess of which one of the five categories it falls into (see terminology at the end of the article).

But even though things seem more complex to them, consider taking it even further. Not only can you misdirect them by having one thing (trick, treasure, etc.) appear like another, consider multi-level encounters. A monster that is actually treasure (Owlbear->eggs) a trap that hides a monster (Swiveling wall->Otyugh) a trap that hides a trick (Pit-Secret Door) and so on. 

"Yes, there's nothing there. There are however the boots, the sword, and the armoire."
Remind them of what the actual interesting objects are. Use the Quantum Ogre advice to present this information to the players. If you design the encounter to avoid having boring uninteresting choices, then reframe the options the players have available if they focus on meaningless items.

This is another place people seem insistent on making sure there is plenty of opportunity to be bored. Just because you present actionable options to your players, does not in any way prevent them from coming up with their own plans. Also, nothing about this means you can't use a good idea you come up with in play (e.g. Maybe there is a secret door under the rug. . . )

Again, don't design tedium into the game and then make up mechanics to bypass boredom you designed in.

"Yes, you climb down the shaft no problem."

I have only rarely played in games that didn't require a ton of useless rolls. (And every single one was an OSR blogger! Thanks Zak, Jeff, and Noisms!) Think long and hard before requiring a roll.

Rolling in OSR games isn't making choices. They are simply finding out the results of a choice they have already made. 

"The salagtite shifts and a single red eye opens. You hear it say, 'Wizard? I haven't had wizard in aaaageeeessssss.'"

Antagonists fall into several categories. These can be split into two types. Simple and complex.

Types of simple monsters include: