Whose fault is this?

Last time I wrote about reducing frustration in tabletop RPGs, including a lengthy suggestion for resetting the baseline of combat so that the minimum result of a combat action is still a little bit of progress. Near the end of that discussion I hinted at why it is better to be dodged than to miss, and said that more would be forthcoming.

My main question today is “Does the player feel responsible for their character’s success or failure?” I won’t be talking a lot about narrative agency (although it’s important), more about how you can use mechanics to influence those feelings.

A Hierarchy

From most to least, I propose a rough ordering of how well players might feel that they own events in the game:

• The players’ actions
  • Varying with how much they feel responsible for the success or failure
• NPC reactions to player actions
• Events out of the players’ control but consistent with their understanding of the game
• Other events

Sometimes you’ll be able to move an event up a category by reframing it or making mechanical tweaks, although clearly the players can’t be the only actors in a drama. You might be able to clear the bottom category though.

Players Owning Their Actions

Obviously this is going to be best, but it’s a pretty broad category where you can do more work to move things to the top. For the best feeling of ownership you want players to feel responsible not only for what their characters attempt, but what then happens, success or not.

Your biggest challenge is getting players to own their characters’ failures. This is difficult because a lot of people have trouble owning their own failures; don’t beat yourself up if you can’t get your players out of a ‘we would have won if it weren’t for [some thing I take no responsibility for]’.

As an aside, a note about who is actually responsible for the outcomes of combat, because you’ve got two choices: ‘you’, or ‘the dice’. If you want to run a competitive game where the players could lose, that’s fine, but I suggest you agree that with the players in case they’re expecting a well-formed drama that can’t accidentally end at an anticlimactic juncture. For best drama, be prepared to cheat (but hide that from your players better than I am by writing it here).

So assuming you’re already avoiding players feeling that ‘the dice weren’t with us’, I suggest you’re aiming for ‘we did our best [with what we knew/had] and it wasn’t our day’, rather than ‘that was an unfair fight’.

That bit in brackets is narratively your responsibility. Consider the hero’s journey: the hero comes to a point where he must attempt to confront the antagonist, because he doesn’t know how to prepare any better (or because he doesn’t understand that he must, although he’s less the owner of that unless someone can convince him of his hubris). He’s defeated, but he can understand why and he gains something that will help next time.

Sure Things

Which brings us to success chance and player perception thereof.

If a player attempts a ‘sure thing’ and it fails, they’ll feel robbed. Never mind ‘well you knew there was a 5% chance of failing’: that’s a statistic that means something to people who attempt something many times and notice they sometimes fail, it’s not a comfort to one-in-twenty of the people who only have one shot at it and need to succeed. You may want to simply avoid this situation (many systems have an explicit ‘automatic success’ rule), or reduce the severity of the failure (by allowing a way of recovering from it, for example).

Sometimes in movies (for example) drama is generated by ridiculously unlucky things happen to protagonists. In the good ones, the unlucky event is foreshadowed, with the audience drawn to the possibility of the event (and in many cases to a sequence leading to heightened probability of the failure). The Final Destination movies are excellent at this, with the added satisfaction that they’re normally running multiple threads of potentially catastrophic happenstance and leaving you guessing what will actually happen.

This mostly works in tabletop, but unless your players have an excellent in-character/out-of-character distinction be prepared for them to change their plans because the dramatic irony seeps into character knowledge. You’ll just have to make sure they have few enough choices that they’ll have to attempt something risky from time to time.

I’m not saying that players shouldn’t fail, in fact they’re going to have to some of the time, for drama. With the same proviso as before, I suggest that they fail when you mean them to. Most parties will do risky things all over the place; for more organised players you’ll just have to keep narrowing their options until they need to take risks.

Note that a choice between two dangerous and unlikely options still allows much more player agency than not having that choice. (It sounds trivial when I put it like that, but agency isn’t linear with how many choices you have: ‘only one choice is no choice’).

I don’t have much to say about players doing truly stupid things, except that if they’ve run out of other options you’ve probably gone wrong. Well, maybe they did, if they were responsible for frittering away all their opportunities, but it’s your job to fix that, either by guiding them away from the most destructive options while you can, or just inventing more opportunities.

I suppose the point of this section is that players should be able to gauge how risky an action is. I’ve always been an advocate of obfuscating target numbers to try and stop the maths distracting from the narrative, but it’s important that players have an honest sense for whether they ought to risk something. They don’t need the exact percentage, but that might be a convenient and fair way to allow them that risk assessment. I don’t have visibility of the game mechanics in real life, but for many tasks I do have experience enough to estimate it; experience that my player (presumably) wouldn’t have.

Who Should Roll?

The previous section was almost entirely narrative, even though I said it wouldn’t be. Let’s look at a key system point: who is rolling? There are competing principles/benefits:

• The owner of the acting character (e.g. attack rolls, skill rolls)
• The owner of the character at stake (e.g. saving throws)
• Both (opposed rolls e.g. damage vs. soak)
• Asymmetric (normally to prefer players)

The last one seems to make a lot of sense for involving players in the game, and many players do simply enjoy rolling dice, but be careful of giving them involvement without agency. The cost of letting them roll dice for NPCs (or similar) is that it detaches those rolls from their context, reduces the mystery around NPC abilities and statistics, and emphasises the ‘game’ aspect; a lot of players will feel like this doesn’t make sense.

The third option I suspect is an attempt to make sure you’re always giving the player something to roll, while also keeping the system symmetric. The cost is much more rolling (and in the case of WoD soak rolls the extra point of failure also goes against what I was saying last time about keeping combat progressing). I’ve been a big fan of the World of Darkness for ages and continued to run Masquerade/Ascension/Apocalypse/etc. well into the Requiem/Awakening/Forsaken era mostly because I preferred the settings; the new mechanics are much smoother and largely superior. That said, rolling attack then soak allows you to make a big deal of accuracy versus damage potential, which are equivalent benefits in the combined attack+damage system; I don’t like to lose that distinctiveness.

Saving Throws

Here’s a mechanical change you could make for your D&D games. Currently many spells work on the model of saving throws, in which the spell automatically works but targets own have a chance to resist (I’m picking words carefully here, because it’s a defender roll that takes no action). You may want it that way, but to me a successful save feels like a failure on the part of the caster, often a player, even though it was out of their hands.

I want to recast it as ‘the caster of a spell must overcome the target’s innate defences’. This is pretty easy. In 5th Edition D&D a saving throw is

d20 + proficiency + attribute mod, versus a difficulty of 8 + proficiency + attribute mod

If both parties have neutral attribute and are unproficient then the save passes 65% of the time (in practice I don’t think the caster can be unproficient, which is why that base save chance is actually 55% at first level), so we can replace it with an ‘active casting’ roll with a 35% (plus proficiency) chance of passing, i.e.:

d20 + proficiency + attribute mod vs. 14 + proficiency + attribute mod

You can probably see the other reason that saving throws often feel bad: they’re balanced differently to attack rolls. Attack rolls have more variables, but assuming proficient attackers and neutral attributes we see saving and attack rolls equally likely when an opponent has 4 points of armour. Most regular opponents aren’t proficient in many saving throws, but that’s not even a factor for AC so those that are have a significant additional advantage. Ability modifiers depend mostly on whether the enemy favours Dex or the ability appropriate to the save, except that more heavily armoured characters may forgo their Dex mod, meaning that on average the savable effect my suffer slightly more from defender abilities.

I suppose in a way it helps differentiate enemy types, since a fighter-type enemy will be relatively weaker to spells (at least Int/Wis/Cha ones), but even for that I’m not fond. I wouldn’t want to try and rebalance it without a good survey of which saves monsters tend to have, but if you pushed me for an answer I suggest you playtest a base target of 12 rather than 14.

Anyway, even without rebalancing it, we’ve reversed the agency of the roll: it’s not now the chance that your opponent fails to resist your spell, but that your casting ability overcomes their defences. I prefer this fiction, and I like the idea of players being responsible for that roll.

For Enemy Spells

This change could be a player spells only, or for all spells, up to you. Leaving players responsible for their own saving throws preserves that piece of involvement, and might be preferable, but personally I prefer the symmetry and would switch all spells.

Until Next Time

I’m going to stop there, partly because this got really long but also because I’ve covered the mechanical things I wanted to.

The idea of linking NPC actions back to player responsibility is one that deserves some thought, but I’m less well qualified for it than game mechanics (think about not just NPCs reacting directly to players, but also ‘player-made villains’ and other NPCs whose decision-making is deeply influenced by their previous experiences of the players).

Players should understand the world well enough to appreciate why it behaves as it does, and this is similar in RPGs to other media. You may expect it to be trickier when revealing the workings of the world may influence your players, but it’s not so different to a novel, whose reader should be well enough informed to understand events but not so informed as to predict them all.

I could talk about these things more another time, if anyone is interested.