9 min read

Method men

Writeups.org employs specifics heuristicsMethods to solve problems or explore data to stat and present characters.

I suppose it could be called evidence-based quantification.

The main principles are :

1/ Seen trumps heard

What we actually see the character do always has priority over descriptions of their abilities.

A character being described as the strongest person in the land or the most gifted scientist of their generation isn’t that useful. Unless it’s the only information we have.

What is of import is the feats of strength they actually perform, or what they actually achieve when doing science. Anything else is hearsayInformation received from other people which cannot be substantiated.

This avoids bias from pufferyExaggerated or false praise. It also helps navigate between universes with different power scales. “Most gifted scientist of their generation” in a realistic setting means a specific form of competence in a narrow domain. But in a 1950s comic book, it would mean a superhuman breadth of knowledge and deductive power. Also, smoking a pipe.

2/ Primary sources trump secondary sources

Primary sources  are the material itself.

Secondary sources  are material about it, written by other persons. Such as handbooks, guidebooks, “secret files”, wikis, etc.. And, yes, writeups.org.

What is of import is what happens in the primary source. No matter how “authoritative” or “official” a secondary source is purported to be, it always gets overruled by the primary.

If the sourcebook says that Drax can lift 50 tons and we see him punching huge asteroids into gravels in the comics, then the sourcebook is wrong.

If a character is stated to be 5’10” but we can plainly see in the art that she’s a good 5″ shorter than random male passerbies, then the sourcebook is wrong.

The goal is to avoid piling up layers of interpretation in a game of telephone. Or be “closer to the metal”, if you want it to sound macho.

Secondary sources definitely inform a profile, but they do not shape it.

3/ The rose trumps its name

How things are called is unimportant. What counts is what they actually *do* in the story.

Many, varied things can be described as “telekinesis” or “reality alteration”. These terms tell us very little. What is of import are the actions actually being performed.

That Daredevil (Matthew Murdock) has a radar sense doesn’t means that he has APs in what the DC Heroes RPG calls Radar Sense. Much for the same reason that people called James Smith aren’t all the same man.

Likewise, that the DC Heroes RPG has an Advantage called “Attractive” doesn’t mean it’s applied to all attractive people. Since what it does is apply a specific bonus to specific rolls when interacting with specific people. What is of import is whether this bonus to these rolls is observably in play. Whether we call this bonus “Attractive” or “Ploofurshuh Floombarrah”.

Deciding that the name is the thing and vice versa is a common human cognitive shortcut. But it’s… y’know… bad.

4/ The track record trumps assumptions

There are abilities that a given character “logically” should have. Scare quotes around “logically”.

Ideally, what counts is what they’re actually shown doing. Not what we think they should be able to do (or be unable to do).

Yet this has limits. Particularly for low-screentime characters. Assumptions are bad, yet one cannot possibly function without them.

So from time to time you’ll see notes stating “such-and-such wasn’t clearly demonstrated, but seems so likely that it was included in the stats”.

For characters who have very few appearances we might even skip the note, since it’s obvious. We couldn’t possibly issue an authoritative, accurate number for a character’s people skills if there are no complex character interactions in the material. Pointing that out every time is a “caution, water is wet” disclaimer.

But for characters who have lots of appearances, the notes will be explicit. Say, “we assume that they speak at least basic German, based on such and such elements of their biography”. Or “it would be logical if they spoke German, but that is not clearly demonstrated and seems too tentative to include in the stats.”

5/ The story trumps the tools describing it

The DC Heroes and Mutants and Masterminds RPGs are tools. And we consider them a means, not an end.

There’s nothing wrong with considering the tools as an end in themselves, mind ! But it’s not what we’re doing here.

So the goal isn’t to produce efficient, elegant, clever “character builds”. It’s to model what’s on the page. The stats are in service of describing the story.

Frex, a Power that costs a jillion points but represents what the character does will be preferred over a Power that costs five points but is a crude approximation.

Crucially, it’s *much* easier to streamline an exactly-as-seen-in-the-material character writeup in a way you prefer, than add missing detail. If it’s too fiddly for your tastes, dial it down to what you prefer.

We’d much rather be too fiddly than oversimplify, meaning you’d have to redo all the research to get the omitted information.

6/ Statistics trump everything

The right numbers and attributes are the ones with the highest odds of producing the outcomes observed in the stories.

All other numbers are wrong and shameful, and should be booed.

Now, we aren’t building probabilities matrices for every character’s every score. But numbers that feel iffy can be probabilistically tested. This is usually done in a simple manner –

You can see examples of that in, say, the XCOM weapons locker article (in the appendix) or the deathclaws profile.

7/ Context is good

Early on our #content was internal to a specific, highly geek-literate, small community.

But our readership keeps expanding. So we now provide more explicit support when it comes to context, concepts and references.

Explaining the historical, genre, cultural, technological, societal, etc. references does impact writeups – and even statting. Explaining something to others instead of assuming it known is a great way to realise that you’ve overlooked information.

8/ Occam’s razor trumps like mad

The razor  and other basic tools of scientific investigation and textual analysis you presumably have been repeatedly taught about in school.

Plus empathy toward the characters.

9/ These heuristics are a choice

So, this can be summed up as an evidence-based approach, living on the same block as the scientific approach.

Now, these are certainly *not* the only possible approach (or épistémè  , if we absolutely must…).

In fact, it’s not the dominant one. An experience-based approach (“I’ve read tons of comics over many years, therefore I can speak about them with authority”) occupies that slot. In fact, it’s the approach we adopt whenever there’s no solid data we can use.

And there are of course others – say, some Star Wars online discussions that use something much like a faith-based approach. Which has its own coherence, its own truth, its own tools and so on.

We just happen to use different ones. And part of our approach is that it’s better, when practical, to explain these things rather than just assume them known.

Tigra (Marvel Comics) and Spider-Man running over an animal stampede

It’s getting long and there are no germane illustrations to break the wall of text, so I just put in a cat picture.

Abductive reading

While waiting for my order at the burger truck, I started reading Yoon Ha Lee’s Ninefox Gambit (yes, there’ll be a Kel Cheris character profile at some point).

Mr. Lee does a Le Guin, dropping the reader into the far future with very little in the way of non-diegetic  explanations. You have to infer what’s going on from Cpt. Cheris’ thoughts as she’s reviewing the situation and adapting her tactics.

Because *of course* the far future isn’t going to just make sense to a XXIst century person.

Therefore, the reader has to engage in what Professor Saint-Gelais  calls “abductive reading”.

Which, as it happens, is a core approach to writeups.org work.

(Also the burger was fine, thanks for asking.)

World’s greatest detective

There are three main types of reasoning to infer stuff.

Deduction infers from known-as-a-fact elements.
“Socrates is human”, “All humans are mortal”, therefore I can deduce from these two certainties that Socrates is mortal.

Induction infers a tentative truth from factual observation.
“I’ve seen hundreds of swans and all were white, therefore I can assume with *some* security that all swans are white. At least until a black swan — or, worse, a mauve swan — pops up”.

Abduction is kinda like deduction, but there’s no known truth. It’s about probabilities.
“Chelis talks about ‘vectors’. Her thoughts about this evoke teleportation/warps in other universes. Therefore the more likely explanation at this stage is that ‘vectors’ are a form of teleportation. This educated guess will be refined or discarded as more information becomes available, thus allowing for better guesses.”

You can compare abductive reasoning to a medical diagnosis. “Probabilistically, symptoms B correspond to medical condition A. Let’s treat for condition A, see what happens, and revise the approach if it doesn’t look like it’s condition A after all, making the next-most-common condition C a more likely culprit.”

Abduct harder

Abductive reading is extensively used when one begins reading comics. There’s too much backstory and genre conventions and imaginary entities and characters etc. for them all to be explained.

So the reader has to infer. Or decide it’s too much work and give up.

But it recedes as the reader’s genre mastery grows. Eventually, it only becomes necessary when tackling a new subgenre, or with stories that deliberately require the reader to work. The latter being commercially dangerous, unless you’re Grant Morrison or something.

However, if you’re doing profiles (especially with a strong RPG stats component), an abductive reading remains the default approach – forever. “This scene implies Scientist: 07 or so. Let’s jot that down for now and see if more information pops up”.

Hence the aforementioned Occam’s Razor. The simpler guess involving the fewest entities is more likely to be correct. When the bulk of the activity is making and revising informed guesses, this is definitely a useful heuristic !

(Though of course, in genre stories, “simple” can get weird. Also known as the “at this point, the simplest explanation is time travel” running gag about convoluted plotting.)

Tigra (Greer Nelson) (Marvel Comics) leaping over a white background


You may note that this is part of a running theme, which is to exclude the idea of absolute truths from the profiling and statting process.

  1. As role-players, we normally express everything in terms of probabilistic models. Some outcomes are more likely, but Schmidt happens. Say, rolling a double.
  2. As profilers, we use abductive reading. Don’t ask me whether Brickhouse is stronger than Funky Flashman. The answer will be “at this juncture, the preponderance of evidence points toward a STR 13 for Brickwall, subject to revision as more information becomes available. Do keep in mind that there have been discrepancies, and that part of the estimate relies on observed performance by *similar* characters, which may or may not…”.
  3. As writers, we know that what happens in the story is what the story needs to have happenin’.
    If it’s weird (e.g., Aunt May defeats Galactus) the writer will have to work A LOT to contain suspension of disbelief. But in the end the story always wins.
    And this is precisely why some of the RPG probabilistic tools (particularly doubles and Hero Points) exist.

If you prefer absolutes well, there’s always Darkseid.