I had intended, at some hazy point in the past, to write a series of “Skill Failure” articles griping in blistering detail about each individual skill in 5E. I set that idea aside as unnecessarily pedantic.

5E has been cropping up in my YouTube feed recently and the game is well liked among my players so I decided to delve into the subject of skills once more. I’ll endeavor to draw new conclusions rather than repeating my earlier thoughts on skills; AD&D for 5E Players: What Drove Me Away: Skills.

I very broadly dislike skills because I don’t think they make D&D better. I’m fully aware of course that skill systems are necessary to other games, “Call of Cthulhu” for example. But it seems to me the action of CoC pivots around skills. The action of D&D, on the other hand, pivots around combat.

The pro skills crowd has an interesting argument, “My character can do things that I can’t.” Skills, they argue, capture that difference.

This argument is one of those unassailable truism that crumbles under further scrutiny. D&D models a lot of things we can’t do in real life, fighting trolls for example. So the question isn’t whether we roll dice in a game to model imaginary interactions. The questions is, “What imaginary interactions do we want to model with dice?”¹

D&D uses dice in combat to simulate the chaos and uncertainty of armed conflict. Lucky rolls represent just that, luck. Just as highly skilled athletes may have bad moments in real life, imaginary warriors sometimes have bad moments in imaginary battles.

So the question isn’t whether we need dice to represent talents our characters have that we don’t, the question is whether the random uncertainty of dice models character abilities satisfactorily. I should add here that we should also ask whether play is enhanced in any given moment by the introduction of random outcomes.

The answer isn’t simple. It may be that the introduction of a random uncertainty makes sense for physical skills. Picking a lock in a stressful situation may be a good case for dice rolling. That possibility of random failure is far less satisfying when, due to random luck, the Barbarian’s “Arcana” check exceeds that of the Wizard’s. At yet other times a player may have such a wonderful idea that the whole table is rooting for its success. Is that a moment we want to inject random uncertainty into?

A second sentiment I often hear from pro skills players is that rolling a check is (or seems) more fair than the alternative. That being arbitrary DM decision making (DM fiat).

It certainly may be fair in 2E where the target and modifiers are known to the player in advance (perhaps with minor situational modifiers from the DM) of rolling. I can’t agree that roll over skills systems are more fair. In these games the DM may set arbitrarily high difficulty levels. I hope we can agree that arbitrarily high difficulty is no more fair than DM fiat. If we can’t agree then read no further.

Interaction Framework

There’s a positive correlation between D&D with roll above skills systems and skills being printed on the character sheet. Don’t take my word for it. Look at the character sheets for D&D 3E+. The fact of the matter is that when all (or most, 3E) of the skills are printed on the sheet they become the player’s tools for interacting with the game world.

Sometimes the framework seems to makes sense.

If a player wants their character to charge into battle on an untrained horse an “Animal Handling” check may be in order. But do we want arbitrary dice to cause the character to fail at a pivotal moment? Are we modelling how well horses are trained for battle?

If the players want their characters to sneak past a guard then a “Stealth” check should be called for. But what of preparation? Have the players observed how frequently the guards pass? Have they blackened their armor to stifle the glint of steel and stuffed it with cloth to muffle its sound? Should we introduce the random results or reward clever play?

And what of the social skills? Are “Deception”, “Insight”, “Intimidation” and “Persuasion” the only ways to interact with the world? Can a successful “Deception” check make an NPC believe a lie when they posses evidence to the contrary? Is “Insight” a lie detector? Can your first level character really “Intimidate” the polymorphed ancient dragon? Can you “Persuade” the chaotic evil fiend to be friends?

Worse is the player with a unique idea that doesn’t neatly fit the skills, struggling over their character sheet trying to fit it into an incompatible framework. The player justifiably seeks to validate their understanding of the rules and the choices they made for the character within the skill framework. If the player doesn’t attempt their unique idea because they can’t see a way to fit the idea into the skills then we must accept that as evidence of a boundary the skill system places on the player.

I have yet to find (as a DM) the correct amount of encouragement that permits players to ignore skills and try their ideas. The DM says the skills aren’t that important. The players understandably dissent. If the skills aren’t important why are they in the rules?

The sad end is an invisible boundary that limits the ways in which players interact with the game world.

Perverse Incentives

The presence of skills creates 2 main incentives for the player. Firstly the player is incentivized to optimize their character build around the realization of some concept which may (but doesn’t have to) revolve around maximizing some skill or set of skills. Secondly the player has the implicit expectation that their choices are meaningful and should be rewarded by consistent successful checks throughout the game.

In the first case a player may reason that in prior games their character has missed certain clues. The player therefore, builds their next character around “Investigation” in order that they don’t miss future clues. The concept is sound but not without limits. Sometimes the player will roll a 1. Sometimes there won’t be any clues to find. Sometimes what can be found through “Investigation” is related to another character’s personal story and it would be far more satisfying for that player to make the discovery.

In each case the DM and player must now confront the player’s disappointment because the dice went against them or because the dice were with them but there was no reward or worse, that the dice were with them and they feel slighted that another player’s moment overshadowed their decisions². None of this touches on the fact that the only lever left to the DM to challenge this player is to raise an invisible number; difficulty class.

In the second case the DM is compelled by the nature of the game that provides such a skill system to reward players by designing encounters that allow them to use and succeed at using the skills. 4E explicitly enshrines this in its Dungeon Masters guide as “Skill Challenges”³. 5E makes no mention of skill challenges (probably to distinguish it from 4E) but 5E would benefit from 4E’s “Skill Challenge” mechanic.

The end result is that encounter design becomes (in part at least) a matter of making lists of skills, difficulty classes and results. Perversely changing the game from R-O-L-E-playing to R-O-L-L-playing.

Story Games

Role-playing isn’t storytelling. If the dungeon master is directing it, it’s not a game.

E. Gary Gygax⁴

After a lot of reading, and YouTube and consideration I’ve come to the conclusion that at some time in the 80s/90s the story game crew won the battle for D&D. What this group needs to play D&D their way is a framework to tell stories without player decisions (very random decisions) derailing them.

The initial foray into story driven games is now commonly known as the “rail-road”. In a rail-road game the plot marches toward a conclusion that the player’s decisions can’t alter. Nascent variants of this style often featured immortal NPCs, unavoidable combats and information that the players had to discover in order for the game to continue.

In these games you can’t befriend the “big bad” or ditch the “helpful” NPC or in fact do anything at all the DM hasn’t planned for⁵. Some DMs took this as a lesson to run other styles of games. Others have sought over the years to refine the game system to limit player choice to a set of mostly predefined dice rolling encounters⁶.

The result is a limited game in which the players only choice is when to roll their dice. This sums to the story DM being allowed to tell whatever the story is without player shenanigans disrupting their vision.

It’s a limited game with limited options and limited outcomes that I personally find woefully boring to run.

Skills vs Role-Play

Here again I risk repeating earlier thoughts, AD&D for 5E Players: What Drove Me Away: Role-Playing

While skills and role-play are not intrinsically at odds skills seemingly provide an irresistible escape hatch from role-play. A player could listen intently to a room description, describe a search in great detail and make some number of Investigation checks to determine their success at each stage. In practice they don’t.

The player shouldn’t be blamed. As outlined above the rules reward the optimization and use of the skills. Role-play would undermine the rules by removing the need to make rolls.

In the case of searching rooms dice rolling may not be so terrible since searching may become repetitive over the course of a campaign. But in this case the DM must still face the random failure problem and the challenge problem.

If bad dice rolls prevent the players from finding some necessary clue the DM must either lower the skill’s DC or move clues in front of the players. In either case, if the DM has to manipulate the frequency of successful die rolls, why roll at all?

Then there is the matter of challenge. Assuming that papers in a desk can always be found on a DC 10 Investigation check. Armed with the knowledge that the player’s skills gain bonuses over time. How does the DM make finding papers more challenging?

If the answer is that difficulty should remain the same then we must ask, “Why roll?” If the answer is that difficulty should increase to match the character’s increased ability then we must ask, “Is it fair to raise DC over time?” The DM must also ask is clever design worthwhile? If a scroll is hidden in a hollowed out bed post DC 20 aren’t we just rewarding lucky die rolls?

Skills may alleviate repetitive tasks but how do they affect unique situations? Can a player simply announce, “I give an inspiring speech, I rolled a 23 Persuasion, what happens?” And reasonably expect the DM to continue feeding them story? What’s the player trying to inspire the crowd to do? What events from the game is the character leveraging to stir the hearts of the crowd? Does the player have an obligation to provide the DM the fodder for realizing their objective?

Your answers may be different from mine.

Unlimited Possibilities

Skills create boundaries around play to limit the options available to players to a subset of actions a story DM can prepare for. Skills introduce random chance that may yield irrational successes and untimely failures. Mostly they’re just not necessary.

Role-play is limited only by the imaginations of those involved in a game. It’s perhaps not the easiest skill to build but it is simple enough. Pretend your a person in the situation that your character is experiencing, decide how you want your character to act and describe it for the entertainment of your fellow players.

Footnotes

1. Here I’m avoiding the discussing the topic of just how random of a result we want to model. Suffice it to say that a d20 produces a more random result than rolling several dice with fewer sides (i.e. 3d6).
2. The decisions the player made when they created the character and again when that character attained each level.
3. James Wyatt, Skill Challenges: Dungeon Master’s Guide (Renton: Wizards of the Coast, 2008), 72.
4. Dungeon Master: The Life and Legacy of Gary Gygax
5. These are the games that my group always tried to break.
6. It’s not possible to completely eliminate player autonomy but you can limit it to a high extent.