The Dungeons & Dragons team released the second One D&D playtest document, which focuses on the Bard, Ranger, and Rogue classes. Like the first packet, the changes in this release convince me that the update remains in good hands. Nonetheless, many changes deserve attention. This post avoids repeating things lead designer Jeremy Crawford mentioned during his video commentaries on the release.

1. Hubris and power level. The D&D team runs public playtests to measure players’ enthusiasm for rules and game elements, rather than to measure power levels. So each packet begins with friendly reassurance that power levels may change. “Don’t worry about broken features,” the note seems to suggest. “Count on us to set the power levels ourselves.” But from Sharpshooter, to healing spirit, to twilight domain clerics, the team keeps releasing features with busted power levels, so I feel unconvinced. Still, during his video, Jeremy Crawford says that future playtest packets will revisit the successful elements, enabling fine tuning.

2. Rules that give the Sage some rest. Even though the D&D team hasn’t shown an unerring sense of power levels, I’m certain the team boasts a hard-won understanding of the rules that raise questions and cause confusion. Jeremy logs D&D’s common misunderstandings and pain points. Exhibit A: The playtest changes Armor Proficiency to Armor Training, so folks learning the game can stop wondering where to put the proficiency bonus when they wear armor. I’ve seen proficiency bonuses mistakenly added to AC, creating bulletproof characters.

I suspect Jeremy never wants to explain how jumping fits with movement either. How else can we explain the playtest separating jump into an action? At that price, no one will ever jump again.

Credit—or blame—the playtest’s careful rules for hiding, influence, and searching on a matching drive to add rigor to certain common tasks. During the fifth edition design, the team opted to favor a dungeon master’s judgement to handle such actions. Rodney Thompson described the goal. “We want a system that makes it easy to be the DM, and at the same time trusts the DM to make the right call for any particular situation, rather than create many highly specific chunks of rules text in an attempt to cover every possible situation.”

The 2014 rules for Stealth and Perception, for example, left room for a lengthy Sage Advice discussion. The playtest rules work to pave over the DM’s judgment and the monsters’ passive perception in favor of a roll against a close-enough DC of 15.

Apparently the D&D team also listens to complaints about a lack of social interaction rules. The playtest moves some social interaction guidelines from the Dungeon Master’s Guide to a place where the table’s rules lawyer can more easily cite them at the table as the reason Vecna must cooperate based on a strong persuasion check. I’m all for helping tables handle social interaction, but leave DMs room to work.

3. Class groups. The playtest recalls the Warrior, Mage, Priest, and Rogue class groups introduced in 2nd edition by putting classes into similar sets. “A Class Group has no rules in itself, but prerequisites and other rules can refer to these groups.”

I imagine a design meeting where the team matched classes to groups, and then faced a jumble of leftovers like Bard, Ranger, and Rogue that defied an obvious group name. What did these classes share in common? They all rate as the most knowledgeable and skilled in their province, whether a tavern, a back alley, or the wild. Designing each class around Expertise and calling the group Experts builds on that trait.

The Expertise feature doubles a character’s proficiency bonus, so at higher levels an expert can succeed at nearly impossible tasks and routinely accomplish merely difficult ones. D&D tests use a d20 roll, and the 1-20 random swing can overwhelm the relatively small bonus delivered by proficiency and ability scores. Even the most talented and skilled characters often fail, creating a system that often fails to reward competence. Expertise delivers enough of a bonus to reward masters of a skill with a reliable chance of success.

Meanwhile, the playtest’s jump rule seems designed to enable a gross range of possible outcomes. An average, untrained person making a running jump for maximum distance can leap between 5 and 20 feet. If this rule had reached print in the 80s, Space Gamer magazine’s Murphy’s Rules cartoon would skewer it for laughs. Basketball games in D&D worlds must be something to see.

4. Inspiration works the way most players think it works. In the last playtest, Jeremy Crawford championed some changes that matched the game rules to the way players incorrectly assumed the game worked. That goal makes this playtest’s change to inspiration inevitable. Players can use inspiration to re-roll after rolling a d20 test. This makes inspiration more valuable, but under the old inspiration rules, few DMs awarded much inspiration, so the house rule’s bigger benefit hardly mattered. My earlier post discussed the merits of giving inspiration for 5% of d20 tests, and how that generosity tilts a game already stacked in the players’ favor. Won’t someone think of the monsters?

5. Bards stay busy every moment. The playtest class descriptions feature numerous small changes that improve play. For example, the Bardic Inspiration and the Cutting Words features include changes that improve the Bard’s agency and remove a source of friction. Now instead of giving another player a Bardic Inspiration die to control and often forget, bards can use their Reaction to add an inspiration roll to a failed d20 Test.

The new design eliminates the requirement that players choose to use a Bardic Inspiration die after they make their rolls, but before the DM determines whether the attack roll or ability check succeeds or fails. That requirement interrupted than natural flow of the game. For Bardic Inspiration, the requirement also blocked the DM’s option to reveal DCs and ACs despite the advantages of transparency. Now in the game fiction, the bard sees a companion falter and gives a magic boost that might win success.

6. Hunters mark gets a fix. The 2104 ranger class suffered from an need to concentrate on the hunter’s mark spell, which underpins the ranger’s flavor as someone who targets prey and pursues it to the finish. With a duration marked in hours, hunter’s mark seems meant to last through a ranger’s daily adventures. But the spell requires concentration, so 2014 rangers who cast another concentration spell lose their mark and what feels like a key feature. Also, 2014 rangers who aimed to enter melee suffered an outsized risk of losing their mark. The playtest version of the ranger no longer needs to concentrate on hunter’s mark. In the last 8 years, would that errata have proved too much?

6. Playing a spellcaster becomes less daunting. Jeremy Crawford says the need to pick spells means that “sometimes playing a spellcaster can be a little daunting.” So the playtest classes add recommended spells to prepare. Good idea. I created a list of recommended spells for wizards, but 2014 spellcasters can prepare different numbers of spells based on an ability score, and that variable added complication to my lists. The playtest rules cut the formulas for number of spells prepared in favor of letting characters prepare spells equal to their spell slots. I’m happy to never again search the class descriptions for the formulas that I never remember.

7. Free hands and spellcasting. D&D’s rules for spellcasting components aim to reinforce the classic flavor of the game’s classes while adding the dash of balance that comes from, say, not letting Wizards equip shields. The simplest measure of these rules’ success comes from 4 tests.

• Do the rules encourage Wizards to carry an arcane focus in one hand while leaving the other hand free?
• Do the rules prevent exploits like letting you equip a shield between turns to maximize AC, and then stow a shield on your turn to cast spells? (DMs can say no, but we like the rules to back us up.)
• Do the rules enable Clerics to equip a shield, carry a weapon, and still cast spells?
• Do the rules enable Rangers to have two weapons or just a sword and shield, and to cast spells without any juggling?

Rules as written, fifth edition passes the first 2 tests, complicates the third test by requiring a cleric to free a hand to cast cure wounds (see the first question answered on page 16 of the Sage Advice Compendium), and botches the fourth test. Sure, a dual-wielding ranger can use their free, manipulate-an-object action on one turn to sheath a sword, and then on next turn use another free action to get out their components, and players can keep track from turn to turn, but few players see that as a fun enhancement to the heroic action. Ranger players could take War Caster, but the rules shouldn’t impose a feat tax just to allow the things we expect of rangers. Also, letting rangers do their thing hardly overpowers the class.

To be fair, the playtest makes a change that eases some of the friction. Now the attack action allows characters to “equip or unequip one Weapon before or after any attack you make as part of this Action.” I like how this enables characters to switch weapons in a single turn without dropping one, but the measure fails to let rangers be rangers without a juggling act.

8. Class capstone abilities come sooner so they get used. The 2014 classes rewarded players who reached level 20 with capstone features that often seemed almost too good. But level 20 represents the end of a character’s career, so players seldom flaunted those wahoo abilities for more than a session. The playtest classes move the capstone features to level 18, so players gain more time to savor them. Levels 19 and 20 now gain more ordinary-feeling rewards. For some players, this change makes the capstones feel less like an aspirational target to seek as the crowning achievement of a character. I say level 18 rates as enough of an achievement to reap these rewards.

9. Why would anyone take the Ability Score Improvement feat? The designers of the 2014 version of fifth edition made feats an optional system that groups could skip in favor of a simpler game. So the 2014 team tried to design feats that matched the power of a +2 ability score increase. Clearly, the One D&D team sees little point to keeping feats optional. Who can blame the team for this conclusion? I never saw a table choose not to use feats.

When Jeremy Crawford touts the playtest’s feats, he boasts that nearly all increase the power of the older versions. They achieved this using the highly technical design technique of packaging every 4th-level feat with an extra +1 ability score boost. Many feats nearly match the 2014 versions that the designers judged as powerful as a +2 ability score bonus, except now boosted by an extra +1, making them as good as a +3. Someone please check my math.

With One D&D awarding feats at level 1, and offering boosted feats at level 4, characters keep getting candy. I hope the monsters get some help keeping up. Won’t someone think of the monsters?

Does anyone else consider feats that bundle a +1 ability score bonus a nuisance? Odd numbered ability scores deliver no bonuses, so without planning, those +1 increases can feel wasted. For new players, the wasted +1 feels like a gotcha. For lazy players like me who rarely plan a character’s career, same.

The monsters and I applaud one change: The designers fixed the worst thing in D&D, the Sharpshooter feat, by removing the +10 damage option. I have just one note: Find a different benefit than bypassing cover.

Erasing the effect of cover means ranged characters can mostly ignore tactics, making combat less interesting for their players. Meanwhile, as a DM, I can counter sharpshooters by having monsters move out of total cover to attack before moving back to total cover. If sharpshooters cope by readying attacks, they lose their extra attacks and bonus action attacks. Unless you relish tactical crunch, none of this tit-for-tat brings much fun, so I would rather just play monsters benefiting from partial cover.

10. Can guidance be saved? The 2014 version of guidance rates as the game’s most useful cantrip and its biggest nuisance. Simply by interjecting “I cast guidance” before every single skill check, the cleric gets to improve d20 rolls by an average of 2.5. This proves both useful and tiresome. Frequently in play, someone blows a check, the cleric remembers forgetting the guidance mantra, and the game halts while players plead to add retroactive guidance. Forgetting guidance creates a feel-bad, gotcha moment. I’ve seen some tables bypass the I-cast-guidance spam by just adding a d4 to every skill check. I assume the DM secretly raises every DC by 3. Should we drop guidance from D&D and call it a win? In the cantrip’s favor, priests praying for divine guidance reinforces these classes’ flavor.

The playtest includes a new version of guidance that makes the spell less spammy. Plus the new version’s limit that characters can only benefit once per day might weaken the spell to extinction. I’m okay with that.