# Theory Thursday–Snatching Success from the Jaws of Failure

TL;DR: Narrative fallout is often offered as a way to escape failure in narrative TTRPGs. This post analyzes three systems implementing “costly success” and concludes with an observation, a hypothesis, and a gameplay option.

One famous result in the logic of questions is that yes-no questions can do the job of any question whatsoever. You may need multiple yes-nos to do that job—to answer, “What is the name of the tallest mountain?” you’d need to ask a sequence: “Is [Mountain name] the tallest?” But as long as your list is complete, you’ll have your answer eventually.

By their relation to the range of attention (it’s a thing), questions can draw attention to particular features. That’s basically how the Socratic Method works. So, the list-of-yes-nos trick is a time-honored analysis method. For instance, here are three question-answer pairs characterizing a TTRPG situation.

1. Did the PC fail a resolution test? Yes.
2. Did the PC get what they wanted? Yes.
3. Did the PC’s circumstances worsen? Yes

While (1)+(2) is just plain disregarding resolution rules, adding (3) with a complication correlated to (1) yields a garden variety of “failing forward,” or “success at a cost.”

The two are not equivalent, and the first part covers differences. Narrative TTRPGs implement the situation in various ways, and I’ll look at two that do, vanilla Apocalypse World and Fate RPG, and one, Cortex Prime, that does not (but probably could).

## Preliminaries: The Cost of Success

“Failing forward” is not identical to “succeeding at a cost,” which is sometimes success, to begin with—that would be a No answer to the question in (1)—rather than “bought-off success.” And the Table-Conversation distinction, which I borrow from D. Vincent Baker, differentiates two varieties of “bought-off” success.

• Table buy-in, aka mechanical. Ex: A test (at the table) is initially failed; metacurrency spending (table) turns it into a success (table); it is narrated as such (in the conversation).
• Conversation buy-in, aka narrative. Ex: A test (table) is failed but narrated as a success accompanied by a complication (both in the conversation).

I’ll adopt a conversation-centric perspective and won’t pay much attention to table-only buy-ins, until the end. Generally, conversation buy-ins need not be conversation-onlyThe table-conversation dichotomy is not a sharp boundary (for nerds: it’s a cover, not a partition). Extra “table” costs (stress track tick, ongoing malus, etc.) still count as “narrative” if they are conversation-first.

Now, for “failing forward.” It can happen with “double” failure (table and conversation)—a Yes-No-Yes situation rather than the Yes-Yes-Yes of (1)-(3). This, however, is a storytelling challenge for the GM: narratively moving the story forward while registering PC’s failure. And this is a topic for another day (or maybe not, I’ll tell you why). And so, I’ll proceed to look at cases of:

• Failure, according to mechanics: the “test” of (1) is a “table” event.
• Success, according to the narration: “get[ing] what they wanted” in (2) is a “conversation” event.
• Complication, according to either (or both): the “complication” of (3) is generally a “conversation” event first but might have “table” consequences.

## Forward and Upward (or not)

If you skipped the foreplay, this section is about game mechanics allowing for failed tests that grant PCs what they want anyway, but with strings attached. The terminology I used to frame questions (1)-(3) is not a common ground for the TTRPGs I’ll consider, so there’s a fair bit of interpretation on my part. But I’ll define my terms, so the fair bit should be fair.

### Apocalypse World: Prepare for the Worst

Let’s start with Apocalypse World (hereafter AW), whose resolution system is based on 2d6+Stat; a single action super-type, the move; and three outcomes: strong hit on 10+, weak hit on 7-9, and miss on 6- (see AW 2E, p.10). A miss is close enough to fail[ing] a resolution test” in (1)A PC gets all they’d want with a strong hit; with a weak hit, some—so the latter is valid for (2). On a miss, the PC “prepare for the worst” and the MC (Mistress of Ceremony, the AW lingo for Game Moderator) has “a golden opportunity [to] make as hard and direct a move” as they like (AW, 2E, p. 114)—close enough for (3). Thus, an AW-friendly interpretation for (1)-(3) is as follows.

1. Did the PC miss? Yes.
2. Did the PC get a weak hit result? Yes.
3. Does the GM have “a golden opportunity [to] make as hard and direct a move as [they] like”? Yes.

Two basic moves fit (1)-(3): read a charged situation (p. 144) and read a charged interaction (p. 145). Both let PCs ask multiple questions on a strong hit, a single one on a weak hit, ditto on a miss, but with a “prepare for the worst” proviso. Some playbook-specific moves also verify (1)-(3) along the same lines or differently (see The Fine Prints). One of the latter, the Hardholder’s Leadership (Fig. 2.), encourages foreshadowing—e.g., announcing future badness—and the introduction of new Threats.

Now, and generally, AW’s Yes-Yes-Yes situations are supported exclusively by moves. The only players’ option to bring about a Yes-Yes-Yes configuration is steering the conversation toward the narrative trigger for a move allowing it. Our next stop is Fate, which takes a very different route.

### Fate RPG: Making Failure Great Again

In Fate RPG, N/PCs roll against a difficulty with four possible outcomes. Rolling under it is a failure, matching (1). Still, the GM can offer on failure the same outcome as rolling over the difficulty, matching (2), at a major cost (narrative), possibly with a mechanical consequence, matching (3). The Fate translation of (1)-(3) is as follows.

1. Did the PC roll under the difficulty? Yes.
2. Did the PC get the same outcome as rolling over the difficulty? Yes.
3. Did the PC pay a major cost (narrative), possibly with other consequences (mechanical)? Yes.

Fate Core includes a detailed discussion of narrative costs and mechanical consequences, and the short of it is: Fate has no rules for “major costs” in the traditional TTRPG sense. There is no input-output, failure-goes-in, cost-goes-out functional relation, only guidelines helping the GM narrate costs and, if appropriate, co-opt consequences mechanics.

Subsequently, there’s surprisingly little to say about “success at a major cost” beyond paraphrasing the Fate ruleset. The alternative option, that is, failure according to mechanics and narrative—the Yes-No-Yes situation—deserves perhaps more attention than in the ruleset. Then again, Rob Hanz has covered it already extensively, and I don’t think I could improve on that. So, maybe not a topic for another day.

### Cortex Prime: Logic and Player Skill

Cortex Prime’s default resolution is based on opposed dice pools. A roll outputs three values: a total, an effect, and (possibly) a complication. Resolution is pass-or-fail: pass when the acting PC’s total exceeds the opposition; fail otherwise. N/PCs declare intent narratively, so (2) is unchanged (handwaving the mechanical counterpart, the effect). The Cortex paraphrase of (1)-(3) is below.

1. Did the PC’s total fail to beat the opposition’s totalYes
2. Did the PC get what they wanted? Yes
3. Did the PC receive a complicationYes

Now, complications are possible on a pass, necessary on a fail. Explicitly (and pedantically): both No-X-Yes and No-Y-No are possible (with “X” and “Y” being variables with range {Yes, No}), but (1) entails (3), so as soon as you have Yes-X-Y, you must have Y=Yes. More importantly (but equally pedantically), Yes-X-Y entails Yes-No-Y. Equivalently: as soon as you have Yes-X-Y, you must have X=No.

Less pedantically: natively, Cortex Prime logically rules out Yes-Yes-Yes situations (the best kind of ruling, if you ask me). No currently available mod (optional rules) can change that. A mod for that is conceivable—and this post started with a half-baked idea for one. But along the way, I changed my mind. I’m not sure it’s such a good idea anymore.

To put it simply, table buy-ins for boosting one’s total are cheap and easy, and planning for them is part of the game. A fail is only inescapable if the player’s past decisions make it so (see The Fine Prints). Subsequently, the system does not need to compensate for randomization nullifying character skills because escaping a fail is player skill. And it’s the GM’s job to smooth the learning curve for that skill—with narrative “failing forward.”

Thus, a Cortex-equivalent to Fate’s “success at a major cost” would be redundant. Then again, there’s room for nuance. Not every Cortex player enjoys gaming the system. A “costly success” option would be an equalizer for players who don’t. It is perhaps not the developer’s intention, but it remains a legit way to play the game, much like the “story mode” of a computer action RPG. So, the idea that prompted this post could be worth a full bake, after all.