Fifth edition Dungeons & Dragons started as a game with a strong foundation, strong enough that when I imagined changes that would best improve the game, I just wished for replacements for the annoying spells, overpowered feats, and toothless monsters—the game’s features atop the foundation.

The D&D team agrees. “We did a smart thing with fifth edition by listening to the fans and what came out of that process was a system that is stable, that is well loved, that incorporates the best elements of earlier editions.” Designer Chris Perkins says. “Now that we have that, we are no longer in a position where we think of D&D as an edition. It’s just D&D.”

The D&D team started fixing trouble spots years ago. For example, newer books like Xanathar’s Guide To Everything revisits the rules for downtime with a more evolved take. Tasha’s Cauldon of Everything includes the most updates, with a new way to assign ability score bonuses, alternatives to game-stopping summoning spells, and new beast master companions that strengthen the ranger archetype. The changes improve the game without invalidating anything in the 2014 Player’s Handbook. (See D&D‘s Ongoing Updates and How a Priority Could Lead to New Core Books.)

In 2024, the D&D team will release new core books, making that 2014 Player’s Handbook obsolete. In a way, this 2024 update resembles the jump between first edition Advanced Dungeons & Dragons and second edition. In the 80s when designers started work on second edition, copies of first edition adventures and books like Oriental Adventures were staying in print and selling well for years. TSR management wanted to keep those evergreen products earning, so they required that second edition remain broadly compatible with first. Second edition’s most important goal was “to make sure the game was still the one you knew and enjoyed.” Of course, first edition had already seen changes and new options would continue to evolve second edition. (See The Dungeons & Dragons Books that Secretly Previewed Each New Edition.)

For the next 12 to 18 months, the D&D design team plans to release monthly playtest packets, enabling gamers to sample and provide feedback on the game’s 2024 release. “You’re going to be able to use all of these playtest docs with your existing core books,” says designer Jeremy Crawford. “We’ve designed these docs so you can take each one, and other than the places where we tell you here’s an update, all of this material works with the core books you already have.”

The D&D team emphasizes how the new release will just build on the game we play today. Their claim and my feeling that the game’s foundation is good leads to the playtest package’s biggest surprise: The document makes changes to rules such as critical hits and conditions—changes at D&Ds foundation. Make no mistake: I’m fine with these changes and the package convinces me that the designers will improve the game.

The changes to D&D’s foundation hide in the packet’s unremarkable sounding “Rules Glossary.” Roleplaying game design often means making choices between the benefits and drawbacks of a particular choice. To weigh the choices revealed by the playtest, I like looking at both sides of this equation. My listing of the drawbacks of a choice doesn’t mean I wouldn’t choose the same.

Critical Hits rate as the candy of D&D. No one ever accused D&D co-creator Gary Gygax of giving players too much candy, and he hated crits. (See page 61 of the original Dungeon Masters Guide.) Like candy, crits give joy, but they’re also bad for us, and especially bad for our new characters. Forget bugbears and goblins; blame most new character deaths on a natural 20. First-level characters lack enough hit points to survive the extra damage. D&D’s designers aim for a game that makes players feel like characters can die while rarely actually killing them. (Some gamers enjoy a more dangerous game, but fifth edition needs optional rules to cater to that taste.) Removing crits helps D&D avoid wasting new characters, but we love our candy, so the test rules allow only player characters to score crits—a change that would have appalled Gary. I like it.

As a DM who speeds play by using average monster damage, monster crits add extra friction. That 20 interrupts my flow and forces me to hunt for damage dice to roll and total. (Yeah, I know I could find a short cut.) A crit and a miss deal less damage than two hits, so the slowdown adds little to play.

Some folks complain that not letting monsters crit makes them too weak, and I’m sympathetic because D&D’s mid- and high-level monsters are too weak, and I’ve complained as much as anyone. But the fix comes from much more damage than the occasional critical hit delivers. Hopefully, the 2024 Monster Manual will deliver the power bump foes need.

The test critical hit rule also affects players. Spell attacks no longer deal crits. This just brings the rule in line with what new players expect: Only weapon attacks and unarmed strikes crit. We D&D enthusiasts can master this change.

The new critical rule also changes the damage formula: Only weapon damage dice get doubled. The designers probably aimed to weaken characters designed to farm criticals with feats like Elven Accuracy. The new formula hinders paladins and rogues by eliminating doubled smite and sneak attack damage. Paladins rate as one of the game’s strongest classes, so this change helps bring them down to Oerth. Rogues suffer more from losing a double sneak attack damage.

Still, in D&D specific rules beat a general rule. The critical rule works like this in general, but a class like rogue might gain a feature that adds additional damage to crits. If that feature worked for melee attacks and not ranged attacks, then it would help make up for the inferiority of melee-focused rogues. A guy can dream.

Rolling a 20. Another change deals monsters a more serious blow than losing critical hits. Based on the new rules for rolling a 20 and inspiration, characters will rarely fail saving throws. Now players gain inspiration whenever they roll 20 on an ability check, saving throw, or an attack roll. Players gain more fun candy for their high rolls. If you already have inspiration, you can pass the award to another character. “We wanted a way to feed people inspiration through the system itself. What the system is intentionally doing is encouraging you to use the inspiration.” Dream on. Inspiration proves so much more valuable for saving throws that I plan to continue hoarding it until I need to make a save. I suspect this will bring my characters closer to never failing a save. When I run games, players like me who hoard inspiration make monsters much less fun to run because characters rarely fail a save and so many monster abilities amount to “Action: Waste a turn while every single character laughs off your biggest threat.” At tables using the widespread house rule that lets players spend inspiration to re-roll, the heroes’ edge grows even stronger.

Instead of the players fighting ice cold dice who could use a lift, the inspiration-on-a-20 mechanic awards more success to the character already rolling 20s. Perhaps if a 20 let you inspire another character in the party, the rule would feel better.

Nonetheless, I have mixed feelings about the inspiration-on-a-20 rule. As a player, I love rolling natural 20s and hate failing saves. But even more, I love challenges that press my characters to the limit.

Ability score bonuses. The playtest’s update to ability score improvements rates at the playtest’s least surprising change. Now instead of pairing each race with set of ability score bonuses that reinforce a fantasy archetype, every player chooses where to put a +2 and a +1 bonus, or alternately three +1 bonuses.

Since first edition AD&D, each race has gained ability score modifiers that match the fantasy archetypes of robust dwarves, agile elves, and so on. This started back when everyone rolled characters at random and when good play meant making the most of whatever the dice gave you.

Now most players build characters to match their tastes, so ability score bonuses limit freedom to create capable characters who defy stereotypes. Also, for many, such adjustments raise troubling reminders of how real ethnic groups can suffer from racist stereotypes that paint people as lacking certain aptitudes. Sure, elves, dwarves, and half-orcs are imaginary species, but they become relatable reflections of us in the game world. After all, imaginary halflings, I mean hobbits, just started as Tolkien’s stand-ins for ordinary folks.

Setting ability scores should require just one step: Assign the scores you want to suit your character. Instead, the current design asks players to assign scores and add bonuses as separate steps, likely adding some back-and-forth friction as players find the right values. I would like to see a process that folds the two steps into one. That would work easiest if the game simply offered a few standard arrays of scores with the ability score bonuses included.

Feats at first level. Originally, the fifth-edition designers sought to make new characters as simple as possible. This returned to D&D’s 1974 roots. Then, characters just started with 6 ability scores and a class. Characters developed in play. Those simple characters proved especially easy for new players. You could immerse yourself in your role and play without knowing the rules. If you’re a hero with a sword and a monster charges, then you know your options: talk fast, hit it, or run. Now text like “a number of times equal to your proficiency bonus” weighs races, 1st-level feats, and classes. If you’re coaching a new player, prepare to explain “proficiency bonus.”

The playtest rules make a new character’s history feel more important by bolstering it with mechanics. “I’m super excited about this whole approach that we’re taking with backgrounds,” says Jeremy Crawford. “It’s all about building your character’s story and making certain meaningful game-mechanic choices that reflect the story you have in mind for your character.” Or instead, you can take the Lucky feat.

For new players, the added “game-mechanic choices” risk making the game feel overwhelming. Maybe that’s fine. New players confronted with a pregenerated character always find it overwhelming, but the end of the session, they typically feel comfortable with the basics.

The designers seem enchanted by the phrase “a number of times equal to your proficiency bonus,” but I wish fewer feats added things to track.

The playtest feats include a change that strike me as ingenious. Each feat includes a level. “One of the ways to make sure that feat selection is not overwhelming is to break feats up into smaller groups, and one of the ways that were doing that is with levels.” Credit Pathfinder second edition for adding this innovation first.

Grappling. The playtest changes the rules for grappling. Now, if your Unarmed Strike hits versus AC, then you can grapple the target. Likely this change aims to make grappling for characters work like all the monsters that grapple by hitting a target. Starting a grapple with an attack strikes me as odd because it defies a fifth edition design principle.

Fifth-edition designer Mike Mearls once explained that to determine whether to use an attack roll or a save, designers asked, “Would a suit of plate mail protect from this?” Armor protects against darts, scythes, and so on, so traps using such hazards make attacks. Poisonous fumes, lightning, and mind blasts all ignore armor, so targets make saves. Attacks to grapple fail this test. Surely though, rules for saves to avoid a grapple would add more complexity than the designers want. Besides, D&D hardly needs another reason to favor Dexterity over Strength.