In fifth edition Dungeons & Dragons, characters and monsters calculate the difficulty class (DC) number needed to save against their spells and powers using the same formula: 8 + proficiency bonus + ability score modifier. That means a 7th-level wizard and a CR-7 mind flayer, both with Intelligence 19 and a +3 proficiency bonus, both force their foes to make a DC 15 save against spells like fireball and powers like mind blast. But according to the designer who led the creation of fifth edition, the monsters were never intended to use the same formula as characters, making the uniformity a “secret error in fifth edition.”

The design strategy of using the same math for both monsters and PCs traces to D&D’s third edition, where characters and creatures shared the same rules foundation. This brought some benefits: In theory, designers could create consistent monsters by plugging attributes like size, type, target challenge rating, and so on into formulas to get the creature’s statistics. Monsters and characters could share customization options like feats and class levels. Third edition sets an effective character level (ECL) for some monsters. The ECL matched the creature’s power to a particular level of PC. A player could play a monstrous PC by just treating it as a typical character of the effective level, and then leveling up from there.

This symmetry suffered drawbacks too. In a discussion, third-edition designer Monte Cook says, “We had to make NPCs and particularly monsters work exactly like player characters. I get it. I understand why that’s cool and why that’s important, but boy that became cumbersome. You could look at some high-level monster and go through his skills and just say, ‘Oh, mistake.’ I’m not thrilled with the fact that I created a game where it was so easy to pick nits.”

Most monsters only survive about three rounds, so unless players love monstrous PCs and DMs enjoy customizing creatures with class levels, the cumbersome math rarely pays off. Worse, the math may result in monsters that prove less fun to battle during its typical three rounds in play. This sort of shared formula became the fifth edition’s secret error.

In a post on Bluesky, fifth-edition lead designer Mike Mearls explained that monsters were not supposed to include their proficiency bonus when calculating the DCs needed to save against their powers. “Monsters were supposed to have DCs…based on their flavor and design.” Monster designers would simply choose saving throw DCs that matched the creature’s lore, while making the creature an exciting foe for the level of character it typically faced. By locking save DCs to a strict formula, designers lost the flexibility to choose the best numbers for fun and flavor.

At lower levels and lower challenge ratings, using the save DC calculation for characters and monsters works well enough, but at higher levels, the steady rise of save DCs that add proficiency bonuses as high as +9 creates trouble. DCs become so high that only characters both proficient at the save and with a high ability score bonus enjoy a realistic chance of success. In a high-level party, that math dooms about two-thirds of the characters to failure. Against a mind flayer’s DC 15 save, a typical party of Intelligence 8 and 10 PCs must hope some of the heroes stand out of the mind blast’s 60-foot cone.

The climactic encounter of Phandelver and Below shows how this dynamic can lead many parties to a total party kill. (This example includes minor spoilers.) The showdown pits a single CR 15 monster against a party of 12th-level characters. According to the 2024 Dungeon Master’s Guide, this encounter rates as somewhere between low and medium difficulty. The saving throw math says something else. The creature’s CR leads to a +5 proficiency bonus backed with a +6 ability score bonus, setting a DC 19 save. The monster’s mind blast power targets creatures of its choice within 60 feet of it. Targets who fail an Intelligence save take damage and become stunned for 1 minute. Typical 12th-level wizards and artificers make that save about 55% of the time; druids and rogues make it about 30% of the time, and everyone else gets a slim 1-in-10 or 1-in-20 chance of success. Some groups need luck for just one character to escape the blast without cartoon stars circling their head. Stunned characters can redo the save every turn, but the barbarian and everyone else who dumped Intelligence still needs to hope for a natural 20—and to hope the monster doesn’t recharge mind blast on a 5 or 6.

The scarcity of characters with good Intelligence saves and the brutality of the stunned condition makes this example particularly harsh, but the dynamics hold true even for Dexterity and Constitution saves. As levels increase, saving throws for characters who lack proficiency swing toward impossible.

How does a mistake like that happen? “Earlier this year I was at a game design talk, and while mingling with other designers, someone asked about the hardest part of design. Every single person gave the same answer—communication and coordination.”

Some members of the fifth-edition design team likely preferred to make characters and monsters share the same formula for calculating saving throw DCs. Perhaps some on the team still favor this consistency. But even if they all agree on the secret error, after 10 years of fifth edition, it’s too late to change.

As a house-rule remedy for fifth edition’s secret error, Mearls suggests adding the proficiency bonus to all a character’s saves rather than just the few with class-assigned proficiency. Obviously, this rates as a major change at the game’s foundation, but it would make fifth edition play more like D&D’s early editions where every character improved at every save as they added levels.