How to Set Stakes, or When Binary Resolution is About More Than Two Choices

This is a good post by Behind the Helm. Maybe even good enough to go in the hall of fame.

It has the simple, elegant "rule of threes" vibe of this post, which outlines a principle so good that I have it attached to the computer I use for running games.

What both of these posts are getting at is that adjudicating uncertain or dangerous situations in an RPG requires the GM to think critically about possible outcomes, and what outcomes the dice should be suggesting.

Many RPGs fail to teach GMs that game action is viewed through the lens of the ruleset. Rules clarify  and focus the game action rather than define it. Calling for a roll of the dice is not (just) a chance for excitement and risk; it is a way to orient everyone at the table around what is happening. A roll of the dice syncs everyone at the table with the current fiction.

Yes, part of this is just knowing when not to roll the dice. It just succeeds. But if dice must be rolled, the GM must understand how to set stakes. 

I have just begun running Mothership, and its an interesting game to examine in the context of Behind the Helm's post. At a casual glance, it feels like a binary resolution game. Roll under, succeed; roll over, fail. Doubles are crits. My D&D-trained brain says, binary resolution plus crits. Easy, right?

But how closely did I read the Warden’s Manual? On page 20, regarding the marine class: “Interpret their failed Combat rolls more generously than others.” Hmmm. Page 26-27 includes a lovely adjudication flowchart; the "warden sets the stakes" phase of the chart is clearly an opportunity to choose binary resolution between success/success-at-cost/failure, although it doesn’t explicitly say that. Page 30, under “there’s more than success and failure at stake,” is essentially making the same point as Behind the Helm's post. 

But I missed it! I read the manual cover to cover before I started running the game, and I didn’t really internalize what those bits collectively mean until I reread it with this blog post in mind.

In one player’s very first session in our Mothership game, they rolled exactly the target number. We had to look it up and confirm that only results under the target succeed, so it was a failure. But I had already set the stakes for the roll at total success or success-at-cost, so I already knew how to adjudicate it. 

The character escaped from their pursuer, but wasn't able to fully close the airlock door. Their pursuer was still trying to pry it open. Binary resolution supported by thoughtful stakes-setting framed the action and made it easy to say to the players, “well, what do you want to do next?”

Playing the Game of Clue as a Deckbuilder

In TTRPGs like Dread and video games like Balatro, we can see how a simple and familiar game concept serves as a "chassis" for a more complex or customized game. When you start thinking about games like this, it can be hard to turn those thoughts off. Consequently an offhand comment about what the game of Clue (aka Cluedo) would be like as a deckbuilder turned into the following.

Assume “standard deckbuilder” rules, wherever a rule is not otherwise stated. Dominion is a good baseline, since the turn procedure is straightforward. This version of Clue would require making a new deck of cards, for various reasons that will be obvious. But it could use the existing Clue board.

First you would need multiple copies of each suspect, weapon, and room card (collectively, clue cards). The specific amount would be determined after playtesting. Shuffle the clue cards and then distribute them evenly in piles between the nine rooms, face down.

Each player starts with some mix of Shoe cards and Magnifying Glass cards (10 total, proportion would depend on playtesting). They draw a five-card hand per turn, ala Dominion. Playing a Shoe allows you to roll the dice to move. Playing a Magnifying Glass lets you search the room you’re in. You can’t search the same room two turns in a row (place a token at the end of your turn to mark your last-searched room). There is also an investigation deck that is available from all rooms.

Searching a room allows you to draft a card from either the clue deck in that room, or the universal investigation deck. Cards acquired this way go in your discard pile, and then you later draw them as you would in any other deckbuilder. The clue cards don’t typically do anything until the “solving the mystery” phase (see below), although some investigation cards might care about them and leverage them as resources before that time. 

The investigation deck includes cards that do typical deckbuilder-type things (see below). Three to five cards (again, determined via playtest) are always available face up in a “market” next to the investigation deck; searching allows you to take one of these and replace it with a new card from the deck.

Solving the mystery: On any turn in which you have each of a suspect, weapon, and room card in your hand; and you are in the room in question; and no other player has a card in their discard pile that matches any of those three cards: you may play those three cards and declare the combination to be the answer to the mystery. Do this in the typical Clue style (e.g., "Professor Plum, with the Wrench, in the Ballroom"). At this point, any player may reveal and discard one of these three cards from their hand to “disqualify” your solution. If they cannot, you win the game.

Most of the investigation cards would be familiar to deckbuilder afficionados (improved movement, more searching, card draw, trashing unwanted cards, etc.) A few ideas for special cards particular to Clue:

• Snoop: Privately look at one other player's hand. Then draw a card.
• Clever Reveal: When a player proposes a solution, you may reveal and trash this card to search any one room’s clue deck; if you find a clue that matches any part of their proposed solution, you can add it to your discard pile (thus invalidating their solution).
• Rule Out: Choose an opponent; they trash a clue card of their choice from their discard pile.
• You’ll Never Catch Me: When a player proposes a solution that names your character as the suspect, you may reveal and trash this card. If you do, they do not immediately win. Instead, you win if you can reach the “x” spot on the staircase on the Clue map. The accusing player can stop you by reaching that spot first. Play continues as normal for other players. During your turn, you may discard this card to draw a card.

Possible issues:

• It may be too easy to reach a fail state where no one can win? Possibly clue cards need to be retrievable in some way, even after they have been trashed. A straightforward solution would be for trashed clues to go to the bottom of the deck.
• Movement around the board may not be incentivized enough?
• Optimal play may obligate competitive players to try to track too much of what they see in their opponents discard piles, particularly clue cards.
• The “you’ll never catch me” minigame may be too cute by half.
• Random card acquisition versus more filtering (e.g., "look at the top three cards of the clue deck, choose one, then shuffle") may need to be evaluated.
• Getting lucky (e.g., finding a room card in the clue deck for the matching room) might allow a player to win too quickly.

Another Card-Based Method for Wrapping Up a Campaign

TTRPGs provide a lot of assistance for beginning a game and a lot less for ending one. I’ve written before about using cards to wrap up a game. Late last year, we wrapped another game with cards. The situation was as follows.

• We were wrapping a multi-year game, and the PCs had finished everything that we wanted to address through conventional, moment-to-moment play.
• We wanted to "zoom out" from a typical session's focus and resolve some big questions without rolling the dice. 
• The PCs had two goals that were in tension with each other.
• We wanted to break the action into separate stages to define gradients of success and failure.
• We wanted danger, risk, or conflict embedded both in the individual stages, but didn't want any particular resolution to be too time-consuming.

For brevity’s sake, I’ll skip the in-fiction details and focus just on the mechanics of the process.

The PCs needed to get through 14 “phases” to achieve their two goals. To ensure we could complete the finale in one session, I used a timer, with the PCs compelled to move forward to the next phase. Each phase began with two face-down playing cards, which the PCs revealed when they began the phase. I had the 14 phases laid out on the table, and the two cards were placed in columns on either side of each phase.

In addition to the face value of numbered cards, jacks were worth 11, queens 12, kings 13, and aces were wild cards, worth either 1 or 14, at the players’ discretion. The gap between the numbers on the two cards indicated the danger of the current phase. A difference of 0-2 indicated minimal danger; 3-5 signified moderate danger; and 6 or higher suggests extreme danger. I had concepts planned for how danger might manifest in each phase, as well as some general danger-escalation tools that could be used at any stage.

To determine how well their efforts to deal with a dangerous situation were going, they could draw a card from the top of the deck and add it to either side. So if a 3 and a 9 were revealed for either side of the phase, creating a danger score of 6, and a player drew an additional card and revealed a 5, they could add it to the side with the 3 to bring that side up to 8 and the stage's overall danger margin down to 1. The six characters in play had three chances each to add a card in this manner, for a total of 18 among the group. To use one of these draws, the PC had to “flash forward” to one of three epilogue moments – one day later, one month later, and one year later.

Here’s the catch. In addition to resolving the danger of the present stage, the running totals of each column will determine if they players achieved one, both, or neither of their overall goals (separate from the danger in each phase). 

So the players needed to balance the phase-by-phase spread against the running totals for their two goals. These two incentives could easily conflict, where they needed to "make up" a difference in the overall goal totals by leaning into danger in some of the stages. The players in my game were actually forced to make the last stage much more dangerous for themselves in order to get their goal numbers to where they wanted them to be.

We handled all of this narratively, using NPCs, abilities, resources, and the flash-forward epilogue moments to interpret the results of the revealed cards. No dice were rolled during this session. But there’s no reason (besides time) that something like this couldn’t be used as a “progress clock”-style overlay on regular game action.

Difficulty 

I decided that a final difference of 4 or less would indicate both goals has been completed successfully, while 5 or more would involve compromising or partially failing at least one of the two goals. I didn't go too deep into calculating the odds here, but I think I was lucky that the cards provided some tension that forced them to act, as there is a non-trivial chance that the numbers on the cards could balance themselves too easily and let the players move forward passively.

Fortunately, it is easier to make this harder in one of several ways:

• Decrease the margin for success on both goals; for example, say that a margin of 2 or less is required to succeed on both goals, with 3-5 succeeding on the higher-value goal only, and a difference of 6 or more indicating that both goals fail.
• Set Aces to 14 only, rather than allowing them to be wild cards. 
• Remove the 2s, 3s, and even the 4s from the deck. The game will become swingier without the low cards, with more phases showing high levels of danger. 
• Allow for fewer interventions to add cards to balance out danger.
• Instead of measuring the success of both goals by the final margin, give the two goals separate (and conflicting) success states. For example, one column could succeed on a high score, while the other could succeed on a low score. This would essentially force the players to choose one over the other, partially influenced by the randomness of the cards. 

But Why Tho?

To some degree this is just a different way of getting to what is really just a progress clock or skill challenge. Dice could simulate this in a way that is closer to the conventional game action. 

But I think there is something for switching systems entirely to frame an epilogue. After spending an entire campaign rolling dice to resolve moment-to-moment action, the cards signposted that the resolution system here was going to be different. 

The cards also provided a good visual centerpiece for the session. When I run games in person, I'm mostly happy to do it with few maps and props. But I have noticed that players benefit from having a shared visual thing to look at, particularly when the session has a very specific shared objective like this. 

The Book Club RPG

A novel TTRPG.

Everyone brings a book to the game. It must be a physical copy that you own and are willing to mark up.

You will need a random number generator set to the number of pages in your book. You may be able to do this with dice, but an online random number generator is easiest, as you can set a range that directly corresponds to the number of pages. Follow the book’s pagination and ignore introductions, appendices, or anything else outside the normal page number range. 

Character creation

Name

“Roll” a random page. Start from the top of the page and grab the first name you find. You can use your own judgment if something “counts” as a name, but generally, err in the direction of choosing the first word that seems reasonably name-like. This is your character’s name.

Mark the word so that you remember that you already used it. A highlighter may help, if you want to keep the book readable. But blacking it out with a sharpie has a powerful energy to it.

Try to stick with this character name, even if you don’t like it, or it isn’t the sort of name you would normally choose for a character (especially if it isn’t the sort of name you typically choose). If you really want to change it, you can reroll, but you must mark one “dogear” on your character sheet. More on that later.

Appearance

Roll three random pages. For each result, choose the first adjective that appears on that page. When you have three adjectives, that’s the description of your character. 

You may get an adjective that doesn’t really make sense for describing a person. Try to use it anyway, even if you think of it as more of a figurative description than a literal one. Mark each word in the same fashion as you did the name.

If you genuinely can’t find any way to make the adjective work, or just dislike the adjective so much that you don’t want it associated with your character, reroll and mark a dogear on your character sheet.

Skills

Roll three random pages. For each result, choose the first verb that appears on that page. These three verbs are your character’s skills.

Some of these will be easy to interpret as player skills, like running or fighting. Others will require some interpretation. What does it mean to have a skill at “representing,” “isolating,” or “facilitating.” Or a verb that is not normally associated with a person, say, “flooding”? You will have to employ some lateral thinking. 

As before, mark the words you use. Basically you’re going to mark a word every time you roll for a word. If you forget every once and a while, that’s fine, but try to do it when you can.

If you really can’t make a verb work, that’s OK, but you know the drill; take a dogear.

Equipment

Roll for three more pages. For each one, take the first concrete noun that refers to a physical object that a person could carry in one hand. It doesn’t have to be something that a person would ordinarily carry. Use common sense.

Ethos

This one is optional, but may be useful to your game if your character still feels fuzzy. You don’t need this for a one-shot, but it may come in handy for a longer game. 

Roll for three more pages and choose the first three abstract nouns you see. Your character has strong feelings about these ideas, but they don’t have to be affirmative. For example, if you get a word like “slavery,” the character could simply be an anti-slavery advocate. Same rule as above with dogears. 

Content

If you ever roll a word that violates the table’s agreed-upon content and vibe for the game, you can reroll such results without taking a dogear. For example, if you use lines and veils, a word that is strongly associated with a subject covered by a line or veil can be rerolled for “free.”  

Action Resolution 

Whenever you do something uncertain, roll a random number and flip to that page. The first letter of the first complete word determines the nature of the outcome.

• A-B: Yes, and 
• C-F: Yes
• G-L: Yes, but 
• M-P: No, and
• Q-S: No
• T-Z: No, but

If you have a relevant skill or piece of equipment, you can look at the first two words with different letters and choose an outcome from between them (for example, if the first series of words on a page was “all actionable attributes have been accounted for,” you would choose “A” and “H”). Mark any words that you “use” in this way. 

The math doesn’t work out perfectly with these letter divisions, but it doesn’t need to; if anything, I kinda like flat “no” as the smallest category and “no, but” as the largest one.

Dogears

You may acquire a few dogears during character creation or play. This is a metacurrency that a player can spend to compel the player with a dogear to roll in a situation that otherwise wouldn't require a roll; or to make that player reroll a result that they just rolled.

When you call in someone's dogear, the affected player erases it from their sheet and you add it to your sheet. If a player cashes one of your dogears during a session, you can’t cash one on them until the next session. So only a few dogears should change hands each session, and they should always be circulating among the players.

Focus and Genre

While you can do this game with very different books, if you want to concentrate the tone a bit more, have everyone select books from the same genre. For an even more focused experience, have everyone use the exact same book. Novels are an obvious choice, but remember that anything full of text made up of full sentences can work pretty well. There's no reason this game wouldn't work for a group all using copies of an air conditioning unit's operation manual.

Studying Noita to Build Deadly Dungeons

I have played 137 hours of Noita, because apparently I enjoy suffering. Noita is a roguelike videogame in which you explore randomly generated biomes, find customizable magic wands, and use them to kill enemies. Watch a few minutes of any video on YouTube and you’ll pretty quickly get a sense of the vibe. Like many videogames, it can be mined for TTRPG ideas.

Danger from Above

Like D&D, Noita is (mostly) a game about going down, down, down into the depths of what is essentially a dungeon. Many (but not all) enemies attack more effectively downward than they do upward. You can clear them as you descend, but both enemy behavior and level architecture become more complex and challenging. Many deaths begin when an enemy attacks laterally or unexpectedly from above, or when multiple enemies apply pressure from different directions. 

And it is not just the enemies. Noita is also famous for simulating liquids pixel by pixel. A lot of this liquid is pretty dangerous. That looks like this.

A stray shot, trap, or an enemy-induced explosion can send liquid pouring down on your character. And enemies who have the high ground will often knock your character into toxic, acidic, or flammable substances. Many Noita runs end not exclusively through enemy attacks, but through some combination of enemy action, environmental hazard, and player error. 

So ask yourself, what materials or substances are abundant in the dungeon, and how could they be exploited by the dungeon denizens (and, in turn, the PCs)? Baking this question into the dungeon’s design means that players are rewarded for understanding and anticipating what the dungeon is about.

Danger from Below

Again as in D&D, Noita monsters get stronger and hazards grow more dangerous with each level you descend. But the deeper you go, the better the loot you find.

Dropping an explosive on an enemy's head can be a relatively safe way to defeat it. But if that enemy is near some fragile treasure, like a potion, you might easily destroy it (and flood the area with the potion's potentially dangerous contents). And good luck getting the gold the enemy dropped if you also set fire to their surroundings.

Some monsters will fight other types of monsters, and it’s pretty typical to hear combat well below your character, out of sight. Listening carefully will alert you to the presence of something that can kill you on sight.

An attentive DM will often telegraph noises from another room on the dungeon level. But how often do you telegraph sounds from another floor? When can you give PCs a clue (or warning) about what is going on under their feet?

Danger from Yourself

Noita’s customizable wand system makes it quite easy to build wands that are as dangerous to the player as to the enemies, and you won’t have to play the game for long before ending a run with “Minä” listed as the cause of death; which is to say, you accidentally took yourself out.

I find this to be a particularly rich vein to tap for magic items in roleplaying games. Some of my favorite items of all time have been things that my players were too scared (too wise?) to actually use them as much as I would like. I can't exactly blame them. They can’t respawn and start a new run as easily as I can in Noita. But taking the safeties off the magical items really sells the danger and opportunity of the game's world.

Rolling for Shoes and Quantum Gaming

I greatly enjoy "quantum" game mechanics. In this context, all that means is some aspect of the game that would typically be predefined is left undefined until its definition is specifically required by the events of the game. Various games use quantum mechanics without naming them, but one of my favorite articulations is this one by Luka Rejec. 

In a pinch, I would probably use Roll For Shoes by Ben Wray, or freeD6 down the line by Liche’s Libram for a quantum chassis. I particularly like the idea that basically anything can be a character skill (allowing for quantum character construction and a negotiated understanding of the game scenario), but that the randomness of the dice ensures that characters have choices and risk/reward calculations; they can’t just add skills indiscriminately, or they’ll end up with low ones that the GM can force them to test. These games also require nomic negotiation between the players to decide how to interrupt unclear overlaps and edge cases.

I do want to try running Roll For Shoes as-is, but I have also put some thought into a slight variation on it that makes it less of a d6 die pool system and lets it tack closer to D&D abilities and d20 resolution. All you need are 3d6 and 2d20, as well as a way to take notes. So you can run this game entirely from your phone, if needed. The game requires at least one GM and at least one player, although it could be played solo with an "oracle" replacing the GM.

The procedure is as follows:

1. The group chooses a genre. The GM should be at least as familiar with the genre as the player or players are. 
2. Choose a goal. It should be very general, but within the confines of the genre. Don’t think too much about who the characters are yet. A goal could be “defeat the evil overlord” or “win the reality TV show” or “steal the huge diamond from the gallery.”
3. Each player chooses a very minimal character concept. This should be as bare-bones as possible. It could be an occupation, a background, or something else. Basically a one-sentence premise, just enough to explain why this person is present in the opening scene.
4. The GM begins the game. Play starts with an opening scene, usually something that will bring the characters together (if there is more than one) and either establish how they know each other or allow them to meet for the first time.
5. When a PC has to do something uncertain, they roll 3d6. This becomes their permanent ability stat (or skill, or whatever term you prefer) for that action.
6. Whenever a player needs to test that stat (including immediately after that first 3d6 roll), they roll d20. A result equal to or lower than their stat is a success. The GM may give them advantage or disadvantage on the roll.
7. If a player rolls something really low on 3d6, they can choose not to test it and can try to approach the situation in a different way. But the result of that 3d6 roll stays on their character sheet.
8. Failed d20 rolls create new antagonists, hazards, obstacles, complications, or other threats. This could be anything in the story that the GM can use to trip up the players. The GM writes it down, along with the number that the player rolled that resulted in the failure. So if the player has an Argue stat of 13, and then rolls a 16 on a test when trying to convince the studio boss to greenlight their movie, the GM notes “Studio Boss: 16.” 
9. Whenever a number associated with a threat is rolled by anyone, that threat can reappear. It doesn’t matter if a different player rolls the number, or if the PC whose failure led to the creation of the threat is present or not, or even if the roll in this instance was a success (because it is rolling against a higher stat). If the GM thinks it makes sense for the threat to appear, it can happen.
10. The story ends whenever everyone thinks it has reached a natural ending. Or…
11. Alternately, the story can end when the game reaches a predetermined number of one or more of the following criteria:
1. Successes. X successful checks against stats are enough to complete the goal.
2. Failures. X failed rolls on stats are enough for the goal to fail.
3. Threats. X threats created are enough to ensure the PCs will fail to complete the goal.
4. Stats. Each character can have only X stats, maximum. Once they’ve reached this maximum, their next failure will knock them out of the story in some fashion or another. If all the PCs are knocked out, they fail to complete the goal.

I'm keeping this in my back pocket in case I need a quick, improvised RPG. If I have an opportunity to try this, I will report back.

How Could TSR Have Made a Better Card-Based Product in 1992?

Last time: Making Sense of TSR's 1992 Collector Cards

Let’s go back to that question I posed at the start of the last post, regarding who or what these cards were for. Famously, late-stage TSR produced products with little or no consideration for market demand. The company started multiple unprofitable game lines and packed warehouses full of unsold product. These decisions nearly drove them out of business, and eventually forced the sale of TSR to Wizards of the Coast. Books and articles by Jon Peterson, Ben Riggs, and others cover this history in greater detail.

So it’s not hard to understand why these cards missed the mark. Someone had the idea of a card-based product for D&D (great!) but no idea how to develop, test, optimize, produce, and market such a product (not great).

But let’s say you’re a designer at TSR in the early 1990s and you get the assignment to create some D&D themed cards. What form should it take? 

Let’s say we don’t expect any TSR designer to predict how Magic: the Gathering (MtG) would soon reshape the market for card-based fantasy gaming products. You’re just expected to make the best-possible product with your circa-1992 knowledge.

Cut the NPCs. I didn’t go through and count how much, but much of this product line was composed of NPC cards. Obviously the pack I pulled was 100% NPCs, and in a pre-Internet world, it would be easy to assume that’s all there is to the product. Which is a shame, because the NPCs are the weak link here.

I mentioned the “let me tell you about my home game” energy in the “Cain Blizzard” entry last week, but the pack I pulled isn’t even the worst example by any means, because at least it wasn’t trying to be funny. For example, I assume this card was a joke from a designer’s home game, and while it may have gone over well in that context, it is neither amusing nor gameable as a commercial product. I have seen incredibly funny things happen in games that I can never properly convey to someone who wasn’t there; RPGs are truly one of the great “you had to be there” areas of humor.

Focus on the monsters. Monsters appeal even to those who don’t play the game. Ask someone what got them into fantasy, or D&D in particular, and there’s a good chance they will say dragons, or beholders, or some other iconic monster. They don’t need to understand what AC or HD is to understand that  the monsters are cool. Spells and magic items also fit more comfortably on cards than NPCs do, but they require a lot of context, and don’t sell the idea of the game on their own in the same way that monsters do.

Our hypothetical 1992 designer need look no further than TSR’s own history for ideas.

Update the cards with new art. No shade on the Jim Roslof art in those early 1980s cards. I like the old-school style, and I think it works well in the OG Monster Manual, or in the context of Dungeon Crawl Classics, a product with a primary audience that is mostly already familiar with RPG aesthetics. But monster cards should be, in part, a way to grab the attention of people without any previous exposure to the game. 

I mentioned how baseball cards would have been a common reference point for an American game designer in the pre-MtG world, but there was another then-recent model that could have been copied. Garbage Pail Kids cards would have played in a similar space. The whole subgrenre of trading cards from that era appealed to a sort of grotty, Mad Magazine rebelliousness. D&D monster cards wouldn’t have to go as far down that road, but some of that energy (on both the front and the back of the card) would have given the product a better chance of catching those young players that TSR was always after.

The art doesn’t have to be perfectly cutting edge for 1992; early MtG itself had a jumble of art styles, with inconsistent tone and art direction. And that included a fair share of MtG fantasy art that would have been more at home airbrushed on the side of a Ford Econoline van than on the cover of a glossy magazine at Waldenbooks. But MtG mixed in enough contemporary art, with a wide range of artists, so it never felt stuck in the past. Compare MtG to some of the CCGs that sprung up in its wake; whatever the gameplay merits of Wyvern, that art style was always going to feel like something off the cover of a yellowed Del Rey paperback, not a cutting-edge 1990s product.

Make the information on the back more gameable. There's room to build on the 1980s TSR cards beyond just art direction. As noted here, the cards from the early 1980s didn’t feature all the monsters, so a new product could serve in part to better capture D&D's ever-expanding bestiary. 

The early ‘80s monster cards are more gameable than the 1992 collector cards reviewed here, but there’s still plenty of room for improvement. They hew closely to the style of the 1977 monster manual. The Gygaxisms of that era have a certain magisterial authority that reads well in the context of a big rulebook, but a card needs to be much more concise.

As with the 1992 collector cards, the wording in the 1980s cards is indifferent to the utility of the product, wasting words describing physical features that are already obvious from the art, or providing deep lore that’s unlikely to come up at the table. That kind of thing may be fine for a monster manual, but not for a card that is primed for table use.

But someone iterating on that model could have really focused on immediately usable information, even if we can’t assume that they would have foreseen the lean efficiency of the OSE house style for presenting monster information, or the impeccable utility of the Monster Overhaul, my current preferred monster manual. 

Custom reaction roll results? Typical behaviors and desires when first encountered? Tactics employed? A sample treasure hoard? It’s easy to imagine a lot of things that could go here, even limited to a 1992 design lexicon. The same blog I linked to earlier notes that D&D still hasn’t cracked this nut in the modern age, so perhaps the ideal D&D card product is still yet to be discovered.

Making Sense of TSR's 1992 Collector Cards

A friend of mine (and player in several of my games) gave me an old pack of TSR Collector Cards for my birthday. Many thanks to him, as I found these interesting enough to write about. Let’s dive in.

First Impressions

These are cheap. They are labeled as “collector cards,” but they don’t have any of the slickness you might expect from more modern products. That makes sense, as these came out in 1992… the year before Magic: The Gathering (“MtG” henceforth) completely changed how card-based game products were made and sold.

These seem to be made of a rough cardstock, without any lamination or other finishing. They’re sized at 2.5 x 3.5 inches. The pack begins with several collector checklists, and each card includes “trading cards” near the logo at the top. I suppose the idea is that people would swap what they were missing with others to complete the full set. Pre-MtG, baseball cards would have been the most likely point of reference for US-based TSR, so I assume that was the model they were working from. The format is picture on the front, stats on the back.

Granola OatTreat

The contents of one of the cards immediately raise the question: what are these cards actually for? The card provides some information about the pictured character, but it’s not really enough detail for game use, either as an NPC combat encounter, or as a pregen PC. Nearly half of the card is given over to a three-sentence background, which is painfully inefficient in using its space, with loose and flowery language (“He had no love for armor, though, or the making of it…”)

Hey look, we pulled both of the Oaktree brothers. I picture the Oaktree brothers as second-run rivals to the Oakridge Boys. 

The cards in this pack are all NPCs, and they have big “let me tell you about my home game” energy, a trap that TSR and other RPG publishers often stumbled into in the earlier days of the hobby. It’s hard to blame them; I imagine everyone thought they were on the verge of creating the next Dragonlance-style hit. Everything on the back of the “Cain Blizzard” card could fly in a home game, but is pretty embarrassing as a product intended to be sold to the public, from the goofy character name to the excessive multiclassing to the confusing backstory (he finds mixing with others difficult, but was still chosen as an emissary to the world?) 

Jadethread!

The art on those first few Greyhawk cards is pretty bland – lacking the roughshod, outsider charm of very early D&D, but not professional enough to hang with commercial fantasy art, even by early 1990s standards. But I’m kinda charmed by Jadethread from the Forgotten Realms. He looks like Phillip Seymour Hoffman in a heist movie. Why is his mask so large? Was he working a blowtorch as part of the heist? I do like that he appears to be doing a little fist pump here. Way to go, Jadethread. You did it.

These cards don’t credit the artist, which is another knock against them. MtG attributed the artwork on the cards from day one.

What I’m also just now noticing is that some of these cards include the superscript “™” after the character’s name, while others don’t. This is probably just an oversight, but I prefer to think that someone at TSR thought that Jadethread was important IP for D&D’s future, while the Oaktree brothers were yeeted into the 1992 equivalent of Creative Commons.

Jobinov and Thiawskeen

Oh, the pain of unpronounceable fantasy names. It might be… THEE-aw-skeen? 

I was going to make fun of Jobinov’s art, but then I turned over the card and saw “Equipment: Lasso.” Jobinov does one thing and he does it well, I have to respect this. I must still question the choice to draw a halfling – a creature etymologically most known for height relative to the height of others – without any point of reference, not even a piece of furniture, just a hazy blue background. 

Thiawskeen and Jobinov are the only two Spelljammer cards in this pack, but there is nothing on either card that conveys "this is Spelljammer." That's another failure of this product line; if it is not going to communicate something about each TSR world and get players interested in buying products from those settings, why feature characters from each one?

Dappledref

“[H]e just hates working for a living.” Dappledref sounds like he stepped right out of a Stan Kelly comic.

Wethilion

For some reason, this pack is chock full of high-level rogues with ordinary stats and minimal equipment. The waste of space here is off the charts. Think about how much you could consolidate this and make it more gameable if it went something like this:

Wethilion (CN Gnome Rogue 10)
AC 6 TH 16 MV 6 HP 33
Has 2d4 automatic copters, clockwork animals, and mechanical soldiers nearby at all times; at least one will invariably malfunction, emitting sparks or crashing into the nearest fragile object. Uses Rube Goldberg-style inventions to assassinate his victims. Too smart for his own good, his schemes are unnecessarily complex. Secretly wishes to find the Sherlock Holmes to his Moriarity. Sample assassination schemes: A clock that drips poison into a particularly punctual victim’s tea; an ice sculpture that releases magic or poison as it melts; a clockwork assassin disguised as a statue, built to kill the king who shares its likeness.

The Dark Sun characters are not especially interesting, but their art does, at least, suggest some internal consistency and include details specific to Dark Sun. I don’t think they are by Brom originals (again, no credited artist), but I imagine Brom’s strong visual identity for Dark Sun gave it some implicit (or possibly even explicit) art direction that improved the quality across the board. The content on the back of each card 

Next time: How Could TSR Have Made a Better Card-Based Product in 1992?

TTRPGs Can Make Even Tic-Tac-Toe Interesting

I’ve written previously about how puzzles should be incorporated in RPGs and why tension and escalation makes them interesting. Let’s approach it from another angle and see how far a dead-simple system can go in powering a TTRPG scenario. 

Take, for example, tic-tac-toe*. Tic-tac-toe is a game many children learn to play. But they do not play it for long. They learn the rules, optimize their play, reach an equilibrium where no one can win, and then stop playing the game. You would not ask anyone (besides a young child who has not yet learned the rules) to play tic-tac-toe with you. It would not be a fun experience for either player. In a way, tic-tac-toe is like a very simple lesson in how to enjoy games, and by extension, how to design them.

Now think about Hollywood Squares**, the game show that has been on and off television for nearly 60 years (revived once again this year, Wikipedia tells me). It's a quiz show where players win by connecting a line on a three-by-three grid of celebrity panelists. 

Nobody is watching Hollywood Squares for that hot tic-tac-toe action. But everyone watching brings their understanding of tic-tac-toe to the experience. It provides a framework for the real entertainment. Tic-tac-toe is not in itself fun, but the show wouldn’t have lasted as long as it has without the structure tic-tac-toe provides.

Apply this idea to TTRPGs. You could play a “game” where both players roll a d20, and whoever gets the highest result wins, and there are no other mechanics. This would obviously be a boring game. But a lot of action in a TTRPG session basically boils down to “two people roll, and the better result prevails.” It’s interesting because we care about the fictional events the roll is resolving.

So a mechanic can be too flimsy to stand on its own, but can still serve as a tool in a larger game.

Imagine the following scenario (set aside for a moment the artificial premise and how or why this puzzle would exist in-fiction; this is an example only). A room is divided into nine sectors. Two opposing sides are in conflict to control it – say the adventuring party and an antagonistic faction. Creating a tic-tac-toe streak “wins,” allowing one side to prevail over the other. What constitutes marking a square is up to the DM; maybe just physically controlling a space, maybe overcoming an opponent in one-on-one battle, maybe something else. 

The players will quickly recognize that they’re playing tic-tac-toe, but the scenario is obviously more interesting than “regular” tic-tac-toe. For one, as described above, tic-tac-toe is facilitating something larger than itself, rather than carrying the entire weight of being fun through its rules and mechanics alone. And two, the open-ended nature of TTRPGs means that tic-tac-toe is only a loose framework, not a rigid structure. 

Can a marked square be returned to its blank state, or flipped by another team? Can PCs mark more than one square at a time, or otherwise break the turn order? Can they stop the opposition from marking a square? How can their characters’ abilities and equipment and knowledge be adapted to this unexpected scenario? 

Remember that many games and game systems can be combined in unexpected ways to create something that is more than the sum of its parts.

*AKA Noughts and Crosses

**AKA Celebrity Squares

When Is a Nerf Actually a Buff?

When is reducing the power of a character’s ability (“nerfing” it) actually making it stronger, in the context of a specific campaign or game concept? When the original character ability solves a challenge so completely that GMs simply stop including that challenge in the game. A nerf to such an ability returns that ability to relevance.

I ran a multi-year D&D 5E game from first level to well beyond 20th level. D&D 5E uses banded accuracy to control the runaway escalation of bonuses and modifiers that plagued previous editions. Banded accuracy works pretty well through the first 10 levels, but D&D 5E kinda falls apart beyond that point (although that can actually make the game more interesting). The following examples are all based on the 2014 version of 5E; the principle is less likely to occur in OSR or story games, where there’s less of a presumption of power balance. But the underlying idea is worth keeping in mind for all games.

For example, rogues at 11th level get an ability called Reliable Talent, which effectively puts a floor on their proficient skills. With Reliable Talent, the Expertise ability, a pair of thieves’ gloves, and a maxed-out Dexterity score, the party rogue in this game rolled dexterity-based skill checks ranging from the high teens to the mid-30s. 

For skills like stealth, this was manageable, because there was always room for enemy countermeasures. The highest stealth roll in the world won’t negate the effect of Alarm, a first-level spell that can be cast as a ritual. And an individual character with very high stealth tends to separate themselves from the group in ways that create interesting situations. 

But what about skills that have less prominent counter-play? What about picking locks? When designing locations for the party, I struggled with how to lock the damn doors in a way that was ludologically meaningful. This was the theme of several of my earliest posts, about how quickly a mundane locked door ceases to be a meaningful barrier in D&D.

Lockpicking in Hillsfar (1989) 

I had the following options, none of them great.

1. I could leave locked doors out of the game entirely. Not realistic, and would negate the ability the rogue had invested in building up.
2. I could inflate the DCs, preserving the challenge. But it would be transparently obvious that I was doing so artificially to “counter” the unbounded scaling of this particular ability.
3. I could leave locked doors in the game, but bypass the wasted time of rolling, and just say the rogue succeeds unless there is a trap or other complicating element. 
4. I could nerf the rogue’s skill to bring it back into line with the bounded accuracy system.

At the time, I went with a combination of options 1 and 3. I was never entirely satisfied with how those played out. I think option 2 is simply wrong. Option 4 was hard to consider for an ongoing game, but was… interesting.

Let’s consider another example. Paladins in 5E can easily cure non-magical diseases. On paper, this is evocative and makes sense. In practice, this means that most 5E games don’t bother to include non-magical diseases in their scenarios. I had a paladin in that 1-20+ 5E game as well, and the ability to cure non-magical diseases came up… two or three times? Not nothing, but it was very rare.

So what if we nerfed the paladin’s ability? How could that possibly be a buff?

Because if the nerf allows disease to matter in the game, the paladin will actually have opportunities to use the ability.

Imagine a campaign setting dominated and defined by a widespread, highly contagious plague. Even if the GM nerfed the paladin’s ability to cure diseases, relative to the 5E base rules, a player might get a lot more use out of that ability in a situation where the challenges include the presence of disease everywhere. 

There are more possible examples. The 2014 5E ranger famously has exploration- and survival-oriented abilities that, uh, basically trivialize traditional RPG exploration and survival. Getting lost and getting hungry are uninteresting challenges in a 5E group that includes a ranger (or to some degree, a character with the outlander background). Weaken that ability… and the ranger, counter-intuitively, becomes more useful and interesting, because navigation and survival are real dangers in a game where the ranger doesn't trivialize these challenges.

There are probably other examples, but this is enough to say – consider instances where nerfing an ability will actually make it more relevant to gameplay. If this is session zero, discuss it with the players. If you’re hacking a game in progress, talk to the players and see what abilities they have that they rarely use. You might be surprised how readily players will sacrifice a bit of on-paper power to ensure the DM can actually create opportunities for them to use those unique abilities.

Deckbuilding in the Stygian Library: The First Layer

Last week: Exploring the Stygian Library as a Deckbuilding Game

Before we get into our first location, we have one important unanswered question. What are Felix and company looking for, exactly? Let’s stick with Knave 2 (“K2”) and roll on the Books table on page 40. We get 47: hunting. They’re after an ancient tome that details ways to hunt terrible, long-extinct monsters from deep under the sea. At least, everyone thought they were extinct. Lately they have revealed themselves to be dangerously non-extinct, so the value of this previously obsolete book has gone through the proverbial roof. Under The Stygian Library’s (“TSL”) distinctions of how hard a book is to find, we’ll put this at 30: “Obscure information, the sort of thing known only to a few scholars and jealously guarded.” Of course, Felix is happy to grab anything else that looks valuable; but this book is the specific reason his patron sent him into the library.

The Display Case

To begin, we’ll generate rooms strictly by the TSL rules. We certainly could create cards for every option on the primary roll tables, to populate rooms through a deck. And we may decide we need DM-facing deckbuilding components in the future. But in the interest of keeping it simple and iterating quickly, I’m going to limit the deckbuilding to immediate adjudication, and keep it away from elements that might be handled during prep in a non-procedural dungeon crawl.

Felix, Clotilde, and Guinevere crawl through the utility panel and tumble into the library’s first room:

• Location: The Display Case. Interesting Shoes.
• Details: Candle Sticks.
• Random Events: Something turns up - it’s unfriendly.
• An Ink Elemental and d4 Inkblots.

No easing our adventurers into this one. We immediately enter a potentially dangerous scenario, with “unfriendly” monsters present in the space. A few questions immediately present themselves:

• The random encounter is “unfriendly” -- how unfriendly?
• How far away? 
• Who sees who first?

Of course, these are the same questions handled by the traditional D&D rules of reaction and distance. Let’s try handling those with cards. For purposes of testing, I’m using dry erase playing cards, which are readily available online. They can smudge with shuffling, but are easy to erase and reuse. Index cards also work fine as a cheaper option, for those less particular about shuffling hand-feel.

We already know from the TSL result that the ink elemental and its blots are “unfriendly,” so we’ll limit reaction results to the negative end of the spectrum. Using the K2 reaction rules, that includes 2-7, everything from (gulp) “kill the PCs” to “Ignore the PCs.” We label the cards, shuffle them, and draw, for…

Gif by Shivur

“Ignore”! The best possible result. Phew. A TPK in the first room would have been underwhelming.

What do we do with this card, as well as the other reactions that we didn’t draw? Let’s set that question aside for now. Deckbuilders can incorporate the results of a scenario into the fiction in a number of ways. We could shuffle the “ignore” card into our player’s deck, and perhaps interpret its re-emergence later as the return of the original monster. Or we could preserve only the remaining reaction cards, suggesting that future encounters will face a dwindling pool of options. We’ll see if the answer reveals itself as go forward.

How far away are the creatures when the party encounters them? Traditionally 2d6x10 would give us a distance in feet for a dungeon encounter, but I don’t think we need to label 11 cards to resolve this. In many cases, cards are going to want to condense options to relatively fewer, broader choices, relative to dice.

For now, let’s just label three cards as close, medium, and far. We flip a card and get “medium”; the inklings are neither close nor far; they’re across the room, within a stone’s throw, but not right on top of our party.

Finally, is either side surprised? K2 treats surprise a little differently, with an opposed wisdom check (interestingly, this is a lot closer to modern/5E D&D than typical old-school rules). Felix has a +2. The Ink Elemental requires some conversion. We’ll halve it and round down, giving it a +2 on this check. There’s some more nuance suggested by K2 that we may or may not want to use later. For now, let’s just keep it simple and make four cards; monster surprised, party surprised, both surprised, neither surprised.

We draw and get… "monster surprised." The cards are really favoring the party so far. The adventurers spot the ink elemental and its satellite ink blots before the monsters know the party is there. Felix holds a finger to his lips and trio lays low. Since the monster’s default is to ignore them anyway, we don’t need to get into any kind of stealth adjudication. Our party also doesn’t need to spend much time in this space, as Felix quickly surveys the collection of unusual shoes and determines they aren’t worth stealing (per TSL, the value is 100 silver times the layer; since this is layer zero, these shoes are interesting, but worthless). The party uses the candlesticks here to light a torch, which Clotilde will carry. 

With no desire to linger here, the party delves deeper into the library. We’ll roll d20+1 (the layer they’re going to) for a location result of 5 (Chained Lectern) and a details result of 8 (Lamp-Post). We’ll pick up their adventure next time.

OK, So Where Is the Deckbuilding?

So far, all we have really done is replace die rolls with drawn cards. This is an intentionally slow start, because we want to gradually discern mechanics from gameplay, not dictate mechanics to the game and assume they'll just work out well. We'll build in more deckbuilding in future installments of this series.

Exploring the Stygian Library as a Deckbuilding Game

I’m a fan of deckbuilders, whether as video games (Slay the Spire, Inscryption, and others) or physical games (Dominion, Star Realms, and many more). I’ve always wanted to see deckbuilding in TTRPGs, but if that game exists, I haven’t found it yet. The closest I know of is Meromorph’s Atma card-based RPG. That game does a nice job of using cards to inspire unexpected scenarios in a TTRPG format; but it’s not a deckbuilder.

I’ve mocked up several different TTRPG deckbuilding systems, but finding time to test games with other people has been a challenge. My players would be happy to test something if I asked them. But any such ask comes at the expense of the other TTRPGs and board games we could be playing when we get together. That's a high bar to clear.

While I still plan to look for opportunities to test multiplayer deckbuilder TTRPGs, I’ve also decided to test a solo deck builder concept. When I see an interesting question like this generate zero replies, it suggests to me that there is untapped space that game designers should explore.

We’re going to keep this as simple as possible by only introducing deckbuilding mechanics when we really need them. The theory is that it’s easier to start with a very simple system and confirm it works ats a basic resolution system before adding really creative tweaks and neat subsystems. In short, we want to create a game that works like other solo TTRPGs, but uses the randomness of card draw instead of dice for moment-to-moment action resolution, as well as the concepts of deckbuilding for both the progression and consequences of dungeon adventure.

As I learned when my Dungeon23 project lapsed, trying to do a big project with a single output is a recipe for disappointment. A project is more robust if it can serve multiple purposes, so I made a list.

• Test if the deckbuilder model can work in a TTRPG
• Get some use out of adventures and other products I have purchased, but rarely or never run
• Experiment with solo RPGs
• Produce some blog posts

That feels pretty good, right? Let’s go!

Library of Babel gifs by Isaac Karth

The Stygian Library

The Stygian Library (“TSL” for short hereon) by Emmy “Cavegirl” Allen is a great example of a “depth crawl,” along with their similarly structured adventure, the Gardens of Ynn. Players explore a place that is generated in part procedurally; there is no fixed dungeon map.

I have used TSL before as a prep tool for my D&D games, where quite a few sessions revolved around a Borges/Prachett-style arcane library occupying a liminal extradimensional space, which could be accessed from different places. But I have never used its procedural generation tools in-session, as they’re primarily designed to work. Time to change that.

“Put enough books in one place, and they distort the world. Space bends in on itself, forming a sort of wormhole, linking the library to other libraries likewise afflicted. The space between is a sort of pocket realm, budded off from reality, maintained by the sheer power of books.”

Emmy Allen, The Stygian Library

TSL is particularly well-suited to this project because the procedural generation of the dungeon makes it easier to run as a solo exercise. The dungeon is not a pre-defined space, so in using TSL for solo play, the player has to do less work to disentangle their player knowledge from their DM knowledge. 

Delve One

We begin in The City’s largest bookshop. Someone was murdered here last night. That’s unfortunate for Someone, but it is very fortunate for us, because we want to access the Stygian Library, and an entrance will only appear in a location that contains both (1. many books and (2. a recent death. Did the patron who hired us for this job simply take advantage of an accident that happened here? Or did they create the opportunity, by orchestrating a murrr-derrr? Best not to dwell on such questions.

Investigating the far corners of the bookstore, we find nothing on the first, second, third, or 13th search of the space. On the cusp of giving up, we find a barely visible door – not much more than a utility panel – hidden behind the heaviest shelves in the place. The key to the front door of the bookstore unlocks this lock as well, which makes no sense; a warning of what’s to come. We enter the Stygian Library.

Library of Babel gifs by Isaac Karth

Our Brave/Foolish Explorers

“Tell you what, if it's a high card, I'll tell you who I am. But if it's a low card, I'll tell you who you are. Is that a deal?”

Deadfall (1993)

I am using Knave 2 (“K2” henceforth) for character generation and ideas, due to its simplicity; general interoperability with other games; and its use of slots, which are a good analogue for cards. I have run quite a lot of the original Knave, but this will be my first genuine experiment with K2. Here is my PC, with no rerolls.

Felix Digham

• A CON of 1 and a WIS of 2 (yes, I happened to roll the same results as the example in the K2 rules)
• Level 1, 0 XP
• 11 item slots
• Bag of spice
• Lamp oil
• Knife
• 2 rations
• 50’ rope
• 2 torches
• Mail shirt
• Helmet
• War scythe (two-handed, d8 damage) (slot 1)
• War scythe (slot 2)
• Poison
• 2 HP
• Careers: Merchant, Thug
• AP 2
• AC 13
• Personality: Truthful
• Appearance: Rugged
• Goal: Serve a deity
• Assets: Smuggles goods
• Liabilities: Known con artist
• Possessions not carried
• Strongbox (hidden in the bookstore and holding the below items)
• Scales
• 10 coins

I was a little confused at first by the merchant/thug and truthful/con artist dichotomies, but the assets and liabilities cleared that tension up. This is a person who has lived in the gray area between legitimate commerce and outright crime. No surprise that they are now employed in such dangerous and shady work.

Felix has spent 100 of his starting 110 coins to retain two hirelings for five days each. Should he survive one or more delves, he hopes to gather enough money to hit level 2 and retain some more sturdy companions to accompany him on future delves.

Clotilde Delamorn, hireling

• AC 11
• HP 3
• Level 1
• Attack punch d2
• MOV 40’
• MRL 4 
• 10 item slots

Guinevere Westerfield, hireling

• AC 11
• HP 3
• Level 1
• Attack punch d2
• MOV 40’
• MRL 4 
• 10 item slots

They carry some cheap sacks for loot, and nothing else. We’ll sketch out more about Clotilde and Guinevere (and Felix, for that matter) if they survive long enough to warrant our interest.

Next week: Deckbuilding in the Stygian Library: The First Layer 

Library of Babel gifs by Isaac Karth

Leveraging the Known Unknowns of NPCs

A reliable way to create an interesting NPC that provides motivation and interaction for the PCs is to measure the difference between what the GM knows and what the NPC who is controlled by that GM does not.

Too many NPCs have perfect or near-perfect awareness of a situation. It's understandable that this happens. GMing takes a lot of brain power, and the closer NPC knowledge is aligned with GM knowledge, the less mental exertion is required to make them work in an RPG scenario. But sometimes the additional work is worth it.

Imagine, for example, the following scenario. A demon is trapped in a dungeon prison. The demon wants to be free, but doesn't know that if it tries to leave the dungeon entirely, the structure will self-destruct as a last resort to keep them from escaping, trapping them underground.

The treasure hunters want to get to the center of the dungeon. They see the traps and locked doors and other defenses, but they don’t understand that they were created to keep something in, not keep someone out.

The guardian of the land in charge of the dungeon’s defenses wants to stop the demon from escaping, but doesn’t know that if the demon is crushed by the self-destruct measure, its oozing ichor will seep into the surrounding land, poisoning it for 100 generations.

Without the PCs’ intervention, this situation is going to slide toward a bad outcome for everyone. The treasure hunters die. The demon will be crushed. The guardian will fail to protect the land.

The PCs can "win" this scenario, in essence, by collecting information until their knowledge is roughly equivalent to the GM’s knowledge. Once they know all the unknowns above, they will know more about the scenario than any of the NPCs. That gives them the tools to resolve the situation favorably. Even if they merely prevent the worst-case scenario, they’ve done so by leveraging information. They figured out what they knew they don’t already knew – their known unknowns – and answered those questions.

Grading Against Gremlin Game Design

I am a game designer! The great thing about TTRPGs is that anyone who has ever had to adjudicate something or make a ruling is essentially a game designer, even to a small degree. I have run games, hacked games, and even made some stuff that people I don’t personally know have used in their own games. Indeed, I think many people become game designers long before they name what they’re doing in that way.

So it was with interest that I read the Gremlin Game Designer's Creed, via the Indie RPG Newsletter featuring it. I found the 10 rules thought-provoking, so I went through each one in turn.

Rules are toys, and the process of rules-mediated play consists of smashing their faces together like little girls making their Barbies make out. Unless a rules module is explicitly intended to be enacted solo, it should present a generous surface area for other rules to bite into. The most elegantly self-contained piece of rules design is, collaboratively speaking, also the most useless.

This sounds sensible to me, but I would need to see a positive and negative example to really judge the worth of this rule.

The principal function of "player characters" as discrete collections of mechanical traits is to furnish each player with an assemblage of shiny things to show off to other players. Mechanical abstraction is well and good, but if you abstract away the act of curating one's collection of shinies, player engagement will suffer.

I’m less sure of this one. My rules-light brain gets suspicious whenever mechanical traits are in the spotlight, rather than the interactions they facilitate. At minimum, this obviously would not apply to solo games. Even in multiplayer GM-player-interaction games, I question how much to center something like this – what if others simply fail to notice or appreciate someone’s “shinies”? Is that a gameplay fail? 

All that said, I do agree that excessive abstraction is something to watch out for.

The GM, if present, is a fellow player. Ensure that they have their own toys and shinies to play with. The failure of a game to provide these is often a major contributor to why nobody wants to run it!

Seems fair, with the same caveat that we shouldn’t lose the forest for the trees.

The most effective way of encouraging players to do what you want is to make a number go up. This applies to both to rewards and to misfortunes; a number counting up to disaster a much more visceral motivator than a number counting down to zero.

This is the first one that really conflicts with my own experience. I have always favored countdowns, particularly for misfortunes. It’s so much easier to say “when it hits zero something bad happens” rather than having some moving target of a top number, different for each situation, even if I prominently display that number for players.

Crunch is good. The defining feature of tabletop roleplaying is that rules produce stories. The act of interpreting the outputs of the rules and the act of telling the game's story are the same activity. Be mindful of what kinds of stories your rules want to tell; you may find that their opinion on the matter differs from your own!

This again immediately raises questions. This would not be true of FKR games or matrix games. Are these less capable of “producing stories” than games with more crunch? And certainly I’ve played games where the amount of crunch was an obstacle to the telling of stories. 

Beyond that it seems fine – I agree that a lot of failed games seem to boil down to the GM and the players trying to tell a story that doesn’t fit with the ruleset employed.

Actually assembling your game's rules is as much a process of discovery as it is of invention. In the course of designing and playtesting, you may find that your own game has rules that you didn't know about. Where did they come from? It is a mystery.

This seems uncontroversial and correct. 

Randomised outcomes should be made mandatory with care and restraint; randomised outcomes should be made available with delirious abandon. As far as is practicable, players should always have the option of asking the dice what unhinged bullshit should happen next. Corollary: lookup tables are your friend.

This seems right and basically in keeping with an old-school approach, where dice might not be rolled for an entire session’s worth of traps and social negotiations, but will certainly come into play when the party drinks from the magic fountain. A lot of this comes down to maximum transparency about which courses of action will require a roll of the dice, and which ones won’t. The games I played didn’t state this specifically, but once I figure it out myself, I was struck by how many other issues it helped resolve.

Players don't need your permission to depart from the rules as written; granting it is arrogant. By the same token, however, it should never be unclear to players whether they're departing from the rules as written. Let the thought process behind what you're writing hang out for all the world to see; folks will be rummaging in the game's guts anyway, so give them easy access.

This seems fine as well, with the corollary that everyone should try to run every game as close to rules-as-written at least once, before departing from those rules. In other words, don’t start hacking a game before you’ve even played it (something I’ve been guilty of doing, oops).

If your game has a default setting, explain it as little as possible, but always let the rules and presentation reflect it. Seeing an entry for "poorly made dwarf" in a table of player character backgrounds will fire a group's imagination more strongly in three words than a chapter stuffed with worldbuilding lore could in ten thousand.

This is right, and I don’t think it’s an accident that they use the poorly made dwarf from Troika as an example. Troika lacks any kind of section explaining the world and the lore; it’s all implied through indirect references in various parts of the rules. The experience I first had reading the rules was just like what’s described above.

You don't need to be good at naming things as long as you're good at puns. Wordplay, alliteration and rhyme may also serve in this capacity, as, in a pinch, may a well placed dick joke.

I am one of those terrible people who professes to hate pun-based humor, then hypocritically uses it myself. My games are loaded with alliterative NPCs (Timothy Tyranny) and dumb homonyms (like an NPC who identified themself as “the trader/traitor,” intentionally pronouncing it ambiguously between the two words). I even put a Combination Pita Hutch and Tonic Well in the last one-shot I ran. A player’s offhand description of a giant parade float rat as “Icky Mouse” inspired an entire theme park parody arc. I am ashamed of this kind of thing but will probably never stop.