Matt Weber's Blog

“We wanted causality from observational data. That was the dream.”

“The Desperation of Causal Inference in Ecology,” by Lizzie Wolkovich.

Some gems in here:

“And that’s the other problem — the bigger one: how much people want this causality. They want to believe that we have the data and methods to show that a disease that wipes out bats leads to an 8% increase in infant mortality.”

“I fear the next 10 years I will live in a sea of piranhas where lots of ecological problems explain 5-10% of infant mortality and plant productivity.”

“… be less gaga over any statistical method (and I do love my own statistical methods so I could practice a little more of what I preach) and teach everyone basic mathematical notation and basic biological models.”

I am not now, nor have I ever been, an ecologist, but this all sounds very familiar from cognitive neuroscience.

Currently reading: THE IMPENDING BLINDNESS OF BILLIE SCOTT, by Zoe Thorogood.

No silk, no song

Just wanted to say I was there for the great Steam Store Crash of 2025, 5 minutes after SILKSONG launched. Can’t login, can’t do anything with my cart, &c. Guess I’ll give it a couple hours and try again.

Currently listening: CAUGHT STEALING, by Charlie Huston, read by Christian Conn.

what is thu he yells

talcan all my lif of thy grandfather and anglisc gods and thy great sweord what thu has nefer efen swung and talcan of our folc and ingengas and feohtan agan all from outside and now they is cuman thu is sittan here on thy arse talcan scit about barns

—The Wake, by Paul Kingsnorth

It’s funny—The Wake was written in 2013, and I started reading it in 2025, and what with one DOGE and another making me try to figure out how it could have been a Trump novel, it’s taken me this long to realize it’s a climate novel. Buccmaster of Holland, poastre and slacctyvyste.

I mean this is probably reductive, there are lots of awful collective action problems in the world right now and many of them are more similar to the Norman conquest than climate change is. But Kingsnorth’s background would put a thumb on the scale of climate as a theme.

Currently reading: The Wake, by Paul Kingsnorth.

OSR=KOB

I recently read Matt Finch’s “A Quick Primer for Old School Gaming,” which is maybe the first thing that’s ever made me think I might understand what the “Old School Renaissance” is, does, and/or stands for. Finch describes four “Zen moments” of old-school gaming, which I’d describe more as principal components, but I think it’s the first two that are most interesting—“Rulings, Not Rules”:

I believe that the defining characteristic of an old school system is that it’s improvisational, played mostly with Rulings, not Rules.

… and “Player Skill, Not Character Abilities”:

“So, how do you know if your character can do something? You tell the Referee what you want to try. Most Referees will give you an idea of the probabilities, and most will also entertain a reasonable discussion about whether their first guess at a probability is fair.”

And these two ideas are obviously closely related, so the DM’s ruling on the difficulty of noticing something depends more on how you (the player) narrate your exploration of the scene (“player skill”) than it does on your character having a high Wisdom and being proficient in Perception (“character abilities”). (It’s worth noting that the assignment of a DC in 3/4/5E is usually a ruling, rarely a rule, and can be fully responsive to player actions as well as ambient circumstances, although the convention in 5E is to grant advantage for ingenuity rather than modify the DC. But this doesn’t especially nullify Finch’s point that rolls can short-circuit narration, which I think is broadly correct.)

So far, so good. But it’s interesting to me in that the OSR seems to be strongly associated with “old school” dungeon crawls—and Finch has some very perceptive defenses of things like treasure-based experience and random encounters in this context—whereas the “rules-light” system I’m most familiar with is Kids On Bikes, most famous for its plasticity in completely non-dungeon-like contexts such as noir (Mentopolis), wizard school (Misfits & Magic), and ‘80s action movie mashups (Never Stop Blowing Up).

Now this is very much a long way of saying “I’m an AD&D 2E baby who got back into TTRPGs thirty years later because of my parasocial one-way bromance with Brennan Lee Mulligan,” and I’m ready to own that. But I do think there’s something unexamined in Finch’s statement “there’s really only one underlying characteristic that defines an old-school system (a fairly open-ended, simple set of rules)”—because Kids On Bikes isn’t an OSR system, as far as I can tell, and it’s not because there are too many rules. If OSR is what Finch says it is, it’s not just because there are few rules, it’s also because of what those rules are about.

I mentioned random encounters and treasure-based experience—Kids On Bikes doesn’t provide any mechanical advantages for accumulating wealth, or indeed any means of “leveling up” at all. Relative to pre-3E D&D, it subtracted a whole system right there! But treasure-based experience supports the OSR vibe because it rewards treasure acquisition, which encourages players to go into dungeons, because that’s where the treasure is. (Why doesn’t it encourage bank robberies? Because getting tougher and more lethal—the effect of accumulating experience—supports crawling more difficult dungeons in a way it doesn’t particularly support robbing more difficult banks… at least in a system that, by stipulation, doesn’t have articulated rule systems relating your character level to a set of skills that support bank-robbing abilities.)

None of this is bad! It’s good for folks to be able to play the kind of game they like, and I think Finch’s rules vs. rulings framework is a useful way to think about the evolution of D&D and the satisfactions offered by different editions. But whatever the OSR is, a sparse ruleset doesn’t seem sufficient to delimit it; and I can’t help but think it’s telling that there’s this impulse to delimit it mechanically, when what’s really being sought is, as far as I can tell, a vibe.

What’s being told, I’m still working out, but it feels like a reluctance to cop to wanting more of what you used to have just because you liked it. Finch allows that people should be let to like things, but his pro-old-school arguments are phrased in terms that gesture at a broader superiority of experience: combats that are quicker, more exciting, and less monotonous, play less encumbered by needless complexity, narrations that are richer and more demanding. Which are all features you might reasonably expect from a rules-light system, but which don’t lead with any particular directness to the old-school dungeon crawl. And maybe from the other side, it’s striking how “old-school” gaming identifies itself so closely with the art styles prevalent in older versions of D&D—or at least that’s what you see in Finch’s primer, Dungeon Crawl Classics, &c. If the point is the mechanics not the vibe, why is the vibe always the same?

One more point about the buried politics of “player skill, not character abilities.” Finch resists the idea, but it seems pretty unexceptionable that this forces a closer identification of player and character: Players who are more persuasive will be more successful at persuasion even if their characters have low Charisma, &c. And of course this is somewhat true irrespective of the system you’re in; characters will always partake of their players’ personalities to some degree, even in ways the player doesn’t intend. But the more you lean on this particular construal of “player skill” (i.e., outcomes are determined by the actions you narrate, not the dice), the more the character’s accomplishments are constrained by the player’s strengths and weaknesses.

Molly Ostertag and Brennan Lee Mulligan have a discussion of TTRPGs that’s worth watching in its entirety, but in particular around the 20-minute mark they talk about the appeal of D&D to the queer community, and Ostertag points out its utility for folks who are questioning: They can make characters who are different from their current presentation of themselves and closer to a self they’ve been imagining, and see how it feels to spend a while in those shoes. The conversation then expands to other reasons you might want to have a character different from you: Maybe you want to try on confidence, or athleticism, or perceptiveness. To do this, you need the scaffolding of rules that support your character in realizing those attributes even when you don’t possess them yourself. Not everyone is seeking these kinds of counterfactuals about their identity from their TTRPG of choice… but they’re seeking some kind of counterfactual, generally having to do with power and excitement and self-efficacy and so on. Again, the difference isn’t really “no rule is better than a rule most of the time”; the difference is in the kind of experiences your rules support.

I’m poking holes in some of Finch’s arguments because they’re interesting to think about, but overall I think his primer is insightful and there’s a lot to agree with. I’ve seen some of the frustrations he articulates arise at my own table, where a young player playing a druid went from a thrill at being able to change into a “baby owlbear” (brown bear stat block, owlbear appearance) to disappointment and boredom because they didn’t have many fun options in combat. Some of that’s on me, I could have paid better attention and given her some ideas for mixing it up, like grappling or maybe an Intimidation check with advantage for ursinity. But if the group had been following a norm of improvisational, free-form combat rather than the more by-the-book tactical combat that falls naturally out of 5E, I wouldn’t have had to.

I’m thinking hard about getting serious about running Kids On Bikes or a variant sometime soon, just to see how it feels. But don’t hold your breath on it being a dungeon crawl. (A Court of Kids and Bikes, on the other hand…)

Currently reading: Animal Pound, by Tom King, Peter Gross, and Tamra Bonvillain.

A self-published writer’s workflow, 2025

Warren Ellis laid out his current workflow in his newsletter; no one asked me, but here’s mine.

A It might be of interest because it’s very simple: Raw words into a reMarkable 2, paste into Atticus, do the rest there.

The reMarkable is more or less an e-ink notebook that will OCR your handwriting. You can pay $3/month to host your documents on their cloud and access it from most devices, or you can just email stuff to yourself for free. The OCR needs a lot of editing, but for me the device gives the low-distraction writing I want from a notebook while eliminating most of the cost of transcription.

So that covers getting the words In The Computer. “The rest,” done in Atticus, covers:

Edit from developmental through line stages (including killing all the typos inserted by the reMarkable)Produce formatted ebooks and printer-ready PDFs, including the cover if necessary.

I’ve used Scrivener for ebook and manuscript production in the past, and as Warren says in his newsletter, it’s a little too powerful; you can do everything you need to do and a whole lot of things you don’t, and the interface doesn’t distinguish between them very well. In particular, you’d better memorize all the parameters you used for Book 1 if you want a consistent look for Book 2… and there are a lot of parameters. Atticus has customizable styles that you can save. There are also things I don’t love about Atticus, in particular their inexplicable handling of smart quotes, and I’ve experienced performance problems when trying to do basic things with a large amount of text; but it’s what all the Streets of Flame books have been done in, and I’ve liked it enough not to shell out for Vellum.

Currently reading: THE WAKE, by Paul Kingsnorth.

A new leaf for Shark Week

My main pop-cultural reference point for Shark Week—other than the thing itself—is 30 ROCK, wherein Tracy Morgan’s Tracy Jordan exhorts his audience to live every week like it’s, you know. So when I read a greeting card declaring its recipient “the human version of Shark Week,” this reads straightforwardly as a compliment: You are rare, vicious, and to be cherished, or words to that effect.

So I was surprised, showing this to my teenage daughter, to get the reaction “ew, how is that a good thing?”

I tried to explain the basic facts and cultural connotations of Shark Week, but she’s 13 and not that interested in history. Eventually she explained, “I thought it was the week of your period.“

Reader, maybe you knew where this was going all along, but I about sprained my diaphragm from laughing.

I’ll never use it, of course. But it makes me happy to know it’s out there.

Currently listening: Last Call, by Tim Powers, read by Bronson Pinchot.

That crucial sense

To be writing on a blog in this day and age is, let’s be honest, like being a scrivener of three-decker novels in the days of the paperback original. Or, perhaps more correctly, to be a paperback writer in the age of TikTok. And yet, all I ever wanted to be was a paperback writer. Affordable, portable books that would eventually get remaindered or traded or dumped into a charity shop where someone like the fourteen year old me could find them and discover new worlds and that crucial sense of not being alone.

“NINE BELLS paperback writer,” Warren Ellis

Currently reading: ONE FOR SORROW, by Jamie McKelvie.

The Ramsian Doctrine

Seven episodes into A CROWN OF CANDY and it still wasn’t on my bingo card that Brennan would use his food-themed world to do a full frontal assault on diet culture and organized Christianity at the same time. Which is a skill issue on my part, to be sure.

I’ve said elsewhere, in full seriousness, that if the estate of George R. R. Martin doesn’t hire Brennan to finish A Song of Ice & Fire, it’ll be an artistic tragedy. His ability to create GoT feels in real time, at a table full of comedians, in a world where everyone is walking food, just gobsmacks me over and over again. What a great show.

Currently listening: A CROWN OF CANDY e7, “Safe Harbor.”

“Vertical microdramas”

Sometimes known as vertical dramas or mini-dramas (a consensus genre label is sure to emerge soon), these series are filmed vertically and typically have dozens of episodes — or chapters — lasting no longer than 60-90 seconds each.

Elaine Low and Natalie Jarvey sketch a trend. (I’ve only read the non-subscribers part; h/t to my gym-mate Brian Eugenio Herrera.)

I have never considered myself a screenwriter and I’m not interested in the angle of “you only have 6-12 months to get in on this”; as a chronic non-professional I have the privilege of ignoring the market. But the format is interesting.

A 60-90 second vertical video is, in 2025, the culmination of “content marketing”: The content is, one for one, the marketing, each episode something you can drop in as a YouTube short, Instagram reel, or TikTok. Low and Jarvey observe that the dominant microdramas “over-index on billionaires, vampires and werewolves,” “share DNA with Lifetime or Hallmark films and the… romance publishing industry,” and are “soapy, telenovela-style shows,” but the potential that struck me here was in science fiction or horror—the kind of thing where you could tell a bunch of unsettling or mysterious micro-stories around an event or premise, rather than slicing a linear narrative into deli-thin chapters. This feels like it could work for a lot of genres, but might have exceptional potential for horror because you could do it on a budget; you don’t have to show the monster. Video creepypasta, but told as a mess of shards reflecting something, rather than straight linear narrative—the usual weakness of creepypasta, where the portent of the premise collapses under the technical demands of storytelling. Comparables: Angel Hare; Observable Radio; Gone; The Department of Midnight.

Anyway, this sounds like fun. I’d love to see what some low-budget, high-insanity creators can do with it.

Currently reading: ROSEWATER, by Tade Thompson.

“She almost leaves and gets pancakes instead of dealing with the climax of the narrative.”

“Our Default Story Structure is Literally Killing Us,” by Charlie Jane Anders:

The rising action of the story means that a problem slowly gets more and more dire, perhaps also spawning a host of associated smaller problems. At the start of the “third act,” things reach a crisis in which everything appears to be headed for irreparable harm. Then, by some combination of luck and courage, the good guys win and the day is saved…

I’m just not sure if this structure is good for us — because we really seem to believe that making things worse and worse will somehow result in a dramatic improvement, via some fantastic reversal. And meanwhile, we have a nasty habit of responding to trauma by inflicting more harm, and we somehow seem to believe that this will eventually make people happy.

I should probably spend some more time with this one. As you can imagine, Anders gives some examples of what she does to complicate this structure in her own fiction: “My own favorite way to mess with story structure is to build in a lot of anti-climaxes and false climaxes.”

It’s a helpful train of thought as I work through the penultimate book of Streets of Flame, with a view toward the finale. The partial skeleton of the plot I have in mind for Wildfire Riptide is very much in the problems-escalate/action-rises/plot-culminates-in-a-huge-moment-of-tension-and-catharsis kind of mold. It’s a helpful reminder that other things can, and maybe should, be done.

Currently listening: Doppelgänger, written and read by Naomi Klein.