This will start off with a bit of a quick review of the advantage/disadvantage system, so feel free to skip down a bit if you’re already thoroughly familiar with that.

In 5th Edition, various circumstances and abilities grant either advantage or disadvantage to a character attempting an action: If you have advantage, you roll 2d20 and keep the higher result. If you have disadvantage, you roll 2d20 and keep the lower result. If you have both advantage and disadvantage, they cancel out (and you just roll 1d20). And they do not stack (so no matter how many things are giving you advantage, for example, you still only roll 2d20 and keep the highest, not 3d20 or 5d20 or 10d20), which also means that even one factor granting advantage will cancel out any amount of disadvantage (and vice versa).

There are several benefits of the advantage/disadvantage system compared to giving circumstantial modifiers to the die rolls:

• The modifiers you’re rolling against are not in constant flux, reducing the amount of in-game calculation required.
• It is viscerally pleasing and immediately rewarding to roll 2d20 and take the higher/lower result. It’s fun to do and you can literally see how your advantage benefited you or your disadvantage cost you by looking at the result on your gained/discarded die.
• It helps maintain the “bounded accuracy” of the system; advantage helps you, but you can still only get die results of 1 to 20.
• “Advantage” and “disadvantage” are incredibly useful terms of art, which designers, scenario writers, and DMs can quickly and efficiently use for any number of purposes. For DMs, in particular, they provide a very simple way to make a fast ruling.

The reasons for not allowing advantage or disadvantage to stack are:

• To maintain the simplicity of that fast, efficient DM’s ruling. Once you’ve determined that something in the situation grants advantage, for example, you don’t have to keep thinking about all the other things that might grant advantage: You have advantage. Move forward. Roll the dice.
• You don’t need complicated stacking rules, nor do you need to allow abilities to stack in potentially absurd ways. This removes one vector by which an RPG system filled with myriad options can suddenly break from the unexpected combination of those options.
• There are some mathematical effects of allowing advantage to stack multiple d20’s into a single roll, the most notable of which, in my opinion, is that your percentage chance of scoring a critical hit radically expands. (This last point is debatable, however, as many would argue that this is perfectly reasonable if you’re enjoying a massively advantageous situation. It also only applies to actually stacking additional dice, but not to the scenario in which you stack all advantage and all disadvantage and then compare the totals to see whether advantage, disadvantage, or neither applies.)

This system has one additional advantage (pun intended) that I want to call specific attention to: The simple, clear-cut mechanical concept of “advantage” also encourages players to engage creatively with the game world in order to create fictional positioning that grants them advantage.

Another example of this that I’ve seen in actual play is Numenera‘s concept of an “asset” — on any given task, PCs can have up to two assets, each of which shifts the difficulty of the task by one step. The first asset “slot,” so to speak, is often occupied by having the right tool for the job. The second asset slot is usually dependent on having some sort of advantageous situation in the game world, and this naturally results in players seeking to create those in-world circumstances that will give them an asset on a task.

In both cases, the clear-cut term of art coupled to the specific fictional situation in the game world reinforces the fiction-mechanics cycle. The mechanic thus, almost paradoxically, encourages players to engage in the game in non-mechanical ways: It’s not enough to just “play your character sheet” by saying “I hit the orc with my +6 attack bonus,” because the mechanics are no longer confined to the bonuses on your character sheet.

Arguably, of course, you can get the same benefit from any system that allows GMs to assign situational bonuses and penalties. But in actual practice, the clear-cut mechanical concept with a term of art attached to it provides a common framework. People just talk about and think about “advantage” and “assets” in ways that they don’t talk about and think about a miscellanea of +1, +2, or +5 bonuses.

THE PROBLEM

Speaking of actual practice, however, this final — and arguably most important — aspect of advantage tends to frequently disappear at the table.

The problem, ironically, is the very versatility of the system. Because advantage is such an easy mechanical hook to use, the designers of the game have used it to model all sorts of things. It’s hard-coded into everything from class abilities to spells to magic items. For example:

Dwarven Resilience. You have advantage on saving throws against poison, and you have resistance against poison damage (explained in Chapter 9).

Or:

Beacon of Hope. This spell bestows hope and vitality. Choose any number of creatures within range. For the duration, each target has advantage on Wisdom saving throws and death saving throws, and regains the maximum number of hit points possible from any healing.

Or:

Boots of Elvenkind. While you wear these boots, your steps make no sound, regardless of the surface you are moving across. You also have advantage on Dexterity (Stealth) checks that rely on moving silently.

Because advantage doesn’t stack, however, tension is created between the designer’s utility and the DM’s utility: If a character has hard-coded advantage from equipment or racial abilities or whatever, the DM immediately loses the ability to meaningfully model the game world through advantage and the player is simultaneously discouraged from engaging with the game world in order to create favorable circumstances.

This is not a desirable outcome: Basically every time the game hard-codes advantage in this way, it makes the game less interesting in actual play.

Numenera recognizes the same basic problem, which is why it provides two asset “slots.” The solution is not quite so straightforward for 5th Edition because there are so many pieces of equipment, for example, that provide identical forms of advantage, that simply providing two slots will simply create min-max builds that stack multiple advantage and still shut down situational creativity (because both slots will already be filled).

But we can find a solution, I think, through parallel thinking. And by simply cutting straight to the heart of the matter.

SITUATIONAL ADVANTAGE

Situational advantage is any advantage which is derived from a character’s immediate circumstances; particularly and specifically those cases of advantage resulting from characters taking actions or positioning themselves in order to create specific situations which grant them advantage.

Situational advantage, and only situational advantage, stacks with other advantage:

• If you have hard-coded advantage (e.g., the advantage against poison damage from dwarven resilience) and situational advantage (e.g., the character manages to dilute the poison before being forced to swallow it), then you roll 3d20 and take the best result.
• One source of disadvantage can cancel EITHER all hard-coded advantage or all situational advantage, but not both. (So if you have all three, you’d still have advantage and roll 2d20 while keeping the best result.)
• If you have two or more sources of disadvantage, then they will cancel both hard-coded and situational advantage. (You would simply roll 1d20 and resolve the action check normally.)

Note that not all forms of advantage appearing in the rulebook are necessarily hard-coded. Some describe situational advantage. (The optional rules for flanking, for example.)

What about advantage from spells? Is the advantage provided by a spell situational? There is a potentially simulationist argument that they are (or perhaps that some subset of them are, although that way probably lies madness). But the primary meta-game point of all this is to encourage players to think creatively as they engage with the game world instead of just throwing a prepackaged block of mechanics at a problem.

And a spell, after all, boils down to a prepackaged block of mechanics.

So if I had to make an ironclad rule, I would say that advantage from spells is always considered hard-coded advantage.

Fortunately, the entire point of situational advantage is to prevent hard-coded rules from disempowering the GM. So I will by happy to override this ironclad rule whenever players think creatively in order to create situational advantages from their spells. For example, by using a create food and water spell to water down the poison before it’s fed to the dwarf.

Right off the bat, I want to note that these are literally my first impressions of the D&D Next playtest rules. I haven’t actually played a session with them and it may be awhile before I get the opportunity (if I ever do).

In order to help you understand my perspective on these rules, I want you to understand a couple of things about where I’m coming from.

First, I have come to realize over the past few months that 5^th Edition will have a tough time selling itself to me. I have an immense amount of time, expertise, and money invested in 3^rd Edition. In order to overcome the inertia of that investment, 5^th Edition would need to radically improve on 3^rd Edition. But the reality is that I am overwhelmingly satisfied with 3^rd Edition as a ruleset. Yes, there are a few problem areas, but I’ve been able to fix most of these with less than 8 pages of house rules. 5^th Edition needs to show me a radical improvement; but there just isn’t that much room to improve.

Second, I want the game bearing the “Dungeons & Dragons” trademark to have the fundamental gameplay that Gygax and Arneson created in 1974. I’m very comfortable with the game gaining an accretion of new mechanics – something which can be seen in every edition of the game from 1974 to 2008. But once you start fundamentally altering the core elements of D&D’s gameplay, you’re going to have a very tough time selling me a product with “Dungeons & Dragons” on the cover.

So, bearing those things in mind, here are my first impressions of the D&D Next playtest.

FIRST IMPRESSIONS

(1) As I mentioned in “The Design History of Saving Throws” a few months back, 4E inverted the facing of the mechanic. I’m glad to see the playtest document revert this decision.

(2) I’m tentatively supportive of the decision to replace the Reflex/Fortitude/Will triumvirate with saving throws based directly on the ability scores (so that you have a Strength save, a Dexterity save, and so forth). Like 3^rd Edition, this offers a universal system.

However, it does potentially reintroduce the hierarchy problem that AD&D eliminated way back in ’78. And you can actually see this in the playtest document. For example, Wisdom saves are used to “resist being charmed” while a Charisma save is used to “resist certain magical compulsions, especially those that overcome your sense of self”. (You might think that Wisdom applies to non-magical charming, but you’d be wrong: Both charm person and command are specifically resisted with Wisdom saves.)

(3) The advantage/disadvantage concept seems like a really valuable tool. Basically, if you have advantage you roll 2d20 and keep the highest. If you have disadvantage on a roll, you roll 2d20 and keep the lowest. Not only does it provide a really useful mechanical hook that you can hang things on, it gives both the players and the DM a firm concept to aim for: “I’m going to try to get an advantage on my attack roll by swinging on the chandelier and dropping on him from above.” or “I take extra time to cover my tracks, hopefully disadvantaging anyone trying to follow me.”

(4) Similarly, the “hazard” concept seems like a great tool. Essentially, if you fail a roll by 10 or more you suffer the hazard. This immediately gives you a consistent mechanical framework for all kinds of stuff: Fail a climbing check and you don’t make any progress; but if suffer a climbing hazard you fall. Fail a check to disarm a trap and you didn’t disarm it; but if you suffer a hazard on the check you’ve actually triggered the trap. And so forth.

(5) I suspect that loosening up a character’s turn during combat will be very advantageous. You can start your move before taking an action and then complete your move afterwards. In addition, all sorts of incidental actions (like drawing a sword or opening an unlocked door) are just assumed to happen “during the round” without need to take an action. I suspect the combination will make it a lot easier for people to improvise, take chances, and generally keep combat more dynamic.

(6) There are dissociated mechanics all over the place. The rogue’s Knack ability (you’re really good at doing something, so twice per day you can choose to be good at it) is a good example of this.

If there’s one thing that I would absolutely, 100% qualify as a complete dealbreaker for 5^th Edition it would be ubiquitous dissociated mechanics. So this does not bode well.

(7) On a similar note, the healing mechanics are essentially identical to 4^th Edition’s approach, except they’ve replaced the term “healing surge” with “hit dice”. I’ve never liked the dissociation of this system. I also don’t like the fact that it allows you to fully recover your hit points after a rest (suggesting that the designers are still fixated on a tactics-only version of D&D play instead of embracing a balanced mixture of tactical and strategic play). And I’m also not a fan of appropriating terms from previous editions and applying them to completely different mechanics in an effort to appeal to nostalgia.

(8) There’s quite a bit of math in the playtest document that looks really questionable to me. I understand it’s a playtest and the whole point is to find stuff to fix, but some of this stuff seems really self-evidently broken.

For example, both splint armor and banded armor cost 500 gp. Splint gives you AC 15 + half Dex modifier, whereas banded gives you AC 16 and a -5 feet speed penalty. If you’ve got a +2 Dex modifier or better, splint is obviously better. And, at best, banded is giving you a +1 bonus to AC. Is a +1 bonus AC really worth a -5 feet speed penalty? Probably not.

Consider, also, studded leather vs. ringmail. Studded leather is cheaper and gives you AC 13 + Dex modifier. Ringmail is more expensive and gives you AC 13 + half Dex modifier. If you have a -1 penalty to Dex, ringmail is superior. But in all other circumstances, the studded leather is strictly better.

(9) As a note of incredibly minor interest and consequence, flasks of acid are inexplicably nerfed even more. This is part of a long trend line of nerfing acid, but we’ve reached the point where it no longer makes any sense at all: Acid in the playtest document is a ranged weapon with one use that deals 1d4 damage. It costs 10 gp. By contrast you can buy a sling for 5 sp and deal 1d6 damage.

(Alchemist’s fire also gets slightly nerfed compared to 3E, but not as severely.)

(10) Spellcasters can now cast cantrips and orisons as often as they like. It’ll be interesting to see how this feels in playtest, but based on the pregenerated characters it feels to me like straight fighters really are actually getting screwed from Day 1 for the first time in the history of D&D.

(11) It appears that absolutely nothing scales with level: Not attack bonuses, not skills. Nothing. It will be interesting to see what accumulating abilities without a commensurate increase in basic capability feels like in play. But I wasn’t a fan of 4E picking a “sweet spot” and locking it in for everybody, and this seems to only be making that even more explicit. At the very least, it’s tickling my “this doesn’t play like D&D” reflex pretty heavily.

MY MOMENTARY CONCLUSION

There’s some innovative and interesting stuff to see here. But I’m not seeing the knockout punch that convinces me that 5E is offering something worth abandoning the time, money, and expertise I have invested in 3E.

In addition, the infestation of dissociated mechanics I’m seeing are a complete poison pill. There’s no way I’m playing 5E if they stick around: They are, as I’ve said before, completely antithetical to everything I want from a roleplaying. They are antithetical to the act of roleplaying itself.

Finally, the system currently feels a lot more like D&D than 4E did. On the other hand, all we’re seeing is a very minimalist, very stripped-down version of the rules. If you similarly stripped 4E down, you’d also end up with something that feels a lot more like D&D than 4E did. And even what we’re seeing is distinctly “not D&D” in a lot of key ways.

So my first impression is one of skepticism leaning towards disappointment. Take that for what it’s worth.