The Zork II computer game from 1981 includes a locked door that you can open by solving a clever puzzle. The door has the old-fashioned sort of lock that lets you look through the keyhole and see the other side. Except here, the key is in the other side of the lock. You slide a mat under the door, and then poke the key out onto the mat. When you pull the mat back, you have the key.

Back when Dungeons & Dragons consisted of the original brown box, before skills, before rogues, before thieves, all the obstacles in the game invited that style of play. You overcame obstacles by immersing yourself in the game world, making decisions, and problem solving. See A Lack of Ability Checks Shaped How People Originally Played Dungeons & Dragons.

This style of play suffers from the same problem as the puzzle in Zork. When Zork II came out, I had only ever seen that sort of old-fashioned lock in my grandma’s house. And if you’ve never examined that kind of lock, the door puzzle simply leaves you stuck and frustrated.

In the old computer adventure games, when you became stuck and frustrated, you had to send money for a hint sheet, and then wait for it to arrive in the mail.

Fourth edition attempted to eliminate such frustration by emphasizing skill checks and skill challenges over concrete obstacles and over players’ problem solving skills. During this era, Dungeon magazine’s submission guidelines warned authors to create challenges for the characters, not the players. When every obstacle has a DC and multiple skills, then no one gets frustrated. If you find a locked door, you can pick the lock with Thievery, or break the door with Athletics.

No one gets frustrated, but no one feels engaged either. When the game only challenges character skill, the players never need to make meaningful decisions or engage the game world. They just look at their character sheet for the best applicable skill. This improves on playing guess-the-solution-I-thought-of with an inflexible DM, but the picking a skill and rolling is less fun than D&D can be.

The fourth-edition designers must have know this, but they emphasized selecting skills and rolling outcomes for a two reasons:

• To add weight to the choices players make when they build characters. See The Pros and Cons of D&D’s Ability Checks.
• To prevent inflexible DMs from hurting the game. Fourth-edition designer Stephen Radney-MacFarland wrote, “In the early days, DMs all too often felt compelled to demonstrate their cleverness and punish players for making ‘wrong’ choices—even a choice as simple and random as which passage to explore.”

Such inflexible, punitive DMs neared extinction decades ago. When Mike “Sly Flourish” Shea asked players to cite the traits of a good DM, flexible ranked first.

Dungeon masters can challenge players without risking player frustration, because DMs can allow creative solutions.

Have you ever noticed how the Tomb of Horrors makes the demi-lich only vulnerable to a short list of curiously-specific attacks?

The demi-lich Acererak’s skull can be harmed only as follows:

• • • a forget spell will force the skull to sink down without taking a soul
    • a shatter spell thrown at it inflicts 10 h.p. of damage
    • a power word, kill pronounced from the an astral or ethereal magic-user will destroy it.
    • only a fighter with a vorpal blade, a ranger with a sword of sharpness +5, or a vorpal weapon, or a paladin with the like or even a +4 weapon can inflict damage on the skull
    • an exorcise spell will cause it to sink as a forget does
    • a dispel evil spell inflicts 5 h.p. of damage
    • a holy word pronounced against it will inflict 20 h.p. of damage
    • a thief slinging one of the large gems in the crypt will inflict 1 h.p. of damage per 10,00 g.p. of value, i.e. 1, 5,, or 10 h.p. of damage, but the gem is thereby shattered

A power word, kill does nothing, unless you happen to be ethereal or astral! How would anyone think of that? Also, the demi-lich is vulnerable to the destruction of very expensive gems. That messes with the players in the best(?) old-school tradition. Only someone immersed in that tradition would even consider the gem attack.

Once, I thought that this list exposed Gary Gygax as an inflexible DM working to punish players. After all, he devised the tomb to challenge—and frustrate—those “fans who boasted of having mighty PCs able to best any challenge offered by the AD&D game.”

Now, I see the list differently. I suspect Gary created Acererak with no vulnerabilities in mind, but as he ran the adventure, players invented attacks. If Gary judged them reasonable, he allowed them to work. When Gary wrote the adventure for publication, he listed the attacks he had allowed so far.

Gary Gygax had the wisdom to allow a creative solution. In the Foreword to Return to the Tomb of Horrors, Gary wrote, “In one tournament use of the setting, a team managed to triumph by using the crown and scepter found earlier as the ultimate tool against the demilich. As Acererak’s skull levitated, one PC set said crown firmly upon the bony pate; another tapped the regal adornment with the ‘wrong’ end of the scepter. Poof! Scratch one demilich, and give the tournament’s first place to the innovative team of players who thought of this novel solution. Russ Stambaugh, the DM for the group, was stunned. ‘Could that work?’ he asked. I shrugged, admitted I certainly hadn’t thought of it and that it was a stroke of genius that deserved a reward.”

When I DM, I love to be surprised. One of the great joys of being a DM is crafting some trap or obstacle, leaving a couple ways to overcome it, and then watching as the players crack the problem with a third way. I’ve run campaigns for groups who proved so good at coming up with unexpected solutions, that I stopped worrying about planning any solutions. I just sat back and watched the players come of with something.

I have three bits of advice for refereeing game-world obstacles that demand player skill to overcome.

• Watch the players for signs of frustration. Be prepared to let he characters uncover a new clue, or to just have something on the other side of that locked door come and open it.
• It’s good to say yes, but avoid being too quick to accept implausible solutions. If a couple of players are deeply engaged in a predicament, and you allow any dumb idea to work, they just get annoyed. The last thing you want is a player arguing that something you allowed should fail.
• Watch out for clever, repeatable ideas that break the game. I remember a player who regaled me with a story that he remembered fondly. His party defeated a dragon by enclosing it in a wall of force shaped like a giant fishbowl, complete with an opening on top too small for escape. Next, they created water above the opening, filling the fishbowl and drowning the dragon. I suspect that no version of Wall of Force ever actually allowed such shenanigans, but as a one-time trick, the stunt created a moment the players’ loved. I wonder what the DM decided to do when the players kept trying to repeat it. If you can use this trick on a dragon, the dungeon becomes your aquarium.