I have no idea when this is gonna get posted as I’m writing it while traveling for the holidays. I overestimated badly my ability to get shit done before my trip. Because I’m traveling and writing this piecemeal, I’ve decided to do the reader question thing. I didn’t want to go back to that well again before year’s end, but here I am.

If you’d like a question answered in a future Ask Angry column, send it to ask.angry@angry.games. Give me explicit permission to call you whatever you want me to call you and keep your e-mail — yes, the whole e-mail — short, sweet, and to the point. And don’t ask a question you need an urgent answer to. I’m changing up how I do the Ask Angry thing next year, so it might be a few months before I post one.

George asks…

Do you have any advice on handling the wish spell (whether it be system-specific for 5e DnD or in general) due to its particularly broad scope? As a game designer, how would you handle the wish spell and spells of similar power in your game?

Thanks, George, for getting this shit right. Your ability to follow basic-ass instructions puts you ahead of something like 98% of my correspondents. How effing sad is that?

As a side note and not to give you a hard time, certain phrasings just frost my ass and this is one of them. I hate “Do you have any advice…” questions. I don’t know why; I just do. I’m always tempted to say, “Yes, I do have advice; thanks for asking” and then sign off. There’s just something wishy-washy and indirect about “Do you have any advice…” I like direct questions like, “How do you adjudicate wish spells,” or, “How should I adjudicate wish spells?” Something strong and direct and ballsy.

And don’t even get me started on, “What’s your opinion.” I hate that shit.

But I digress. George, you’re above reproach because your e-mail didn’t give me an instant aneurysm. You get a cookie for doing the bare-ass minimum. Nice work.

Truth be told, I think modern Dungeons & Dragons — basically post Y2K D&D and both Pathfinders — handle the wish spell as well as any game could. Each spells out a bunch of safe uses of the wish spell and most of those are down to duplicating other spells without cost or restriction or producing similarly-powered effects. The 5E list of safe uses is actually pretty good.

Each description then goes on to say you can wish for anything you want, but the farther you go beyond the safe uses, the more likely you’re going to get the Monkey’s Paw middle finger. Not that most of you kids even know what The Monkey’s Paw is. This is a case where the trope outlived the original work.

Seriously, though, read The Monkey’s Paw. It’s a short horror story published in 1902 in Harper’s Monthly and it’s the original be careful what you wish for story. And before any of you mention djinns, do remember djinns are benevolent spirits and most didn’t grant wishes but instead offered nasty bargains. They only became jerkasses when they’d been enslaved by evil sorcerers.

But I digress…

My point is, whether you’ve got a PC casting the spell or an item or creature is granting a wish, go by what the book says and your own good judgment. The key is to follow the spirit behind the myths and stories of twisted wishes. If the player’s being ruthless or greedy, screw ’em. If they’re trying to avoid the costs or consequences of their actions, screw ’em. If they’re trying to reverse an event they don’t like or can’t accept, teach them to accept life as it comes. The less self-serving or consequence-avoiding or fate-undoing the wish, the lighter the cost, but a cost there must always be. Roleplaying games are all about costs and consequences.

That said…

From a literary standpoint, I love the idea of the twisted wish word game. You know, the one that ends with the genie saying, “I technically gave you what you asked for…” But gamers have turned that shit into a terrible metagame of spelling out paragraphs of legalese for the GM to spot a loophole in. If that’s fun for you, more power to you, but I’m not a fan. And it misses the point. Wish spells aren’t about technical pedantry, they’re about the morality of the wish itself.

But, thanks to that bullshit in gamer lore, most players won’t touch a wish-granting item or creature with a standard-issue ten-foot pole these days. Just like no one will willingly pull a card from the Deck of Many Things.

And that’s why I hate that wish is even a spell. I don’t think it belongs on the spell list. Neither does the divine version, miracle. I think it’s one of those things that doesn’t need rules, just judgment, and I think wishes should only ever come from rings of wishing or imprisoned djinns or grateful angels or magical monkey’s paws.

Forehead asks…

I’ve found AI useful for preparing content and I’ve even used it mid-game as an alternative to loot tables and name generators. Is this fine and just another tool or do you reckon I’m doing myself a disservice by relying on a crutch? Have you used AI for GM tasks? What uses would you permit or encourage?

People have been after me for an opinion on AI for a while now and I’ve been putting off offering it. Partly because the question requires a measured response and I don’t measure responses. And partly because I’m in a different boat than most of you are sailing. And partly because there are a bunch of moral, ethical, and legal questions to examine.

First, let me say that Generative AI is a tool. Like any tool, you can use it for good or evil. And, like any tool, it’s good for some things and not for others. Knowing how and when to use it is a big part of using it well.

Second, let me say as a content creator who gets paid to make content and believes that’s right and good and proper, I do not use — and I will never use — Generative AI to create content. Every word you word is a word I’ve personally written. Every piece of art and every map that accompanies my work is a piece I’ve created myself or that I’ve contracted someone to create for me. Without the use of Generative AI.

Part of that’s just a quality issue. I can’t predict the future, but for right now, I can say Generative AI just can’t create the same quality of art as a human. But there are also legal and ethical issues to consider. As a published content creator, I know Generative AI programs have been trained using my work without my express permission and I do not appreciate that. The ethical and legal and moral issues must be ironed out, but it’ll be years before that happens and, in that time, we have no idea where Generative AI will go.

That said, I have experimented quite a bit with Generative AI programs like ChatGPT for my personal home use and I’ve even used it for research purposes in my work. Not extensively, mind you, and I probably won’t keep doing that in the future, but just know that I’m not talking out of my ass here.

I’m leaving aside the moral and ethical and legal shit here. I’m not going to say anyone’s a bad person for using Generative AI. That would be a dumb stance to take. But it would be equally dumb to pretend there ain’t serious issues to consider and to point out that software engineers are way more concerned with what they can make these tools do than what they should make these tools do.

Is Generative AI a useful GMing tool? Abso-frigging-lutely it is. At its heart, Generative AI is basically just an expert search engine that’s so good at summarizing shit that it can throw its results into a blender and spit out stuff that looks original. And search engines are already extremely useful tools. Hell, Generative AI is a necessary tool because the Internet is fast becoming useless to humans.

I see Generative AI as a summer intern. I can assign it a task or ask it a question and it’ll come back with something. And it’ll come back pretty damned fast. But what it brings me is questionable. So I don’t ask it for anything I wouldn’t trust an unpaid teenager to do. And I check its work.

Would I ask Generative AI for a list of names? Sure. I can pick which names to use and which ones not to. And I can even say, “These names are good, but can you make them sound more Gaelic” or whatever, and get a whole new list in seconds. Which is also something I could do with a summer intern. It’d just take longer.

Would I ask it to generate the contents of a treasure chest? Maybe, but probably not. Why? Because treasure affects game balance and the published tables are already balanced. Supposedly. I can’t trust an unpaid teenager not to screw up my game’s balance and in the time it took me to check my intern’s work, I could just have generated the treasure myself. I wouldn’t even ask it for, say, “an art object worth 500 gp” because I don’t trust it to deliver things that should reasonably cost 500 gp. I mean, I don’t consider the crystalized eyes of an extinct dragon or a clockwork songbird robot to be worth a mere 500 gp, but ChatGPT does. Trust me. I asked.

I ain’t saying that I’d never use it to generate treasure or that doing so is wrong, I’m just trying to show you how to think about Generative AI. It’s a very eager unpaid summer intern. It’s got zero experience and zero judgment, and it just wants to get you an answer as quickly as possible, so it ain’t digging deep or thinking about context or consequences.

It’s also worth recognizing that if you rely on Generative AI too much, you’re not building your own skills. I would never use Generative AI to write flavor text because, as a GM, I have to be able to narrate on the fly. And I have to do it faster than I can ask a Generative AI to do it. If I can’t do that, I need all the practice I can get.

And the same’s true when it comes to generating ideas. I know some GMs love Generative AI for its ability to spit out ideas for plots and hazards and encounters because they find it hard to produce ideas on the fly. As a professional creator, I’ve had to learn — as all creators do — that idea generation isn’t hard; it just takes practice. Every content creator has to learn how to write a story or article or module or video or whatever even when they’ve got no ideas or inspiration. The more you do it, the easier it gets. I can sit down at the computer with no ideas in my head and still have a 5000-word article draft six hours later. And so can you.

Creativity on Demand is something I’d call a vital Game Mastering skill. And it atrophies fast. I so don’t recommend you make a habit of going to the Generative AI well instead of, say, brainstorming. That said, I’ve used Generative AI to generate plot points, outlines, obstacles, and hazards for my personal games. Mainly, again, to experiment with the tool. It works and, in a pinch, it can save your ass, but it’s a really bad habit and probably very addictive.

One thing GMs overlook though is the ability to use Generative AI for research. It’s an excellent way to get an overview of something you know nothing about. Especially if you’re willing to have a conversation with the AI and then do some follow-up research to flesh out what you learned and confirm the information is good.

Say, for example, I wanted to write a police-procedural game but I don’t actually know anything about how detectives really investigate murders. And I always assume I don’t know anything about anything I’ve never done myself. Which is just good practice. I might ask ChatGPT to outline the basic process by which murders are investigated. And it might then give me a good step-by-step overview of the process. And then I might ask it to expand on various points. For example, I might ask how witness interviews are conducted.

It’s important to note, by the way, that ChatGPT and other Generative AI programs do generate their conversations based on the content of the conversation. So you can ask follow-up questions and seek clarification or ask it to modify answers or rephrase things.

As a side note, ChatGPT is also a better search engine than most search engines nowadays. Seriously.

As a helpful example — and to make sure I’m not talking out of my ass — I did have a conversation with ChatGPT about murder investigations. I didn’t take it too far — I just wanted to illustrate my point — but you can read my conversation with ChatGPT about murder investigations if you’d like.

If I were running a home game, I might stop there and trust ChatGPT’s answers as is. If I were writing a book or module for publication, I’d conduct some interviews with professionals and do a lot of follow-up research to make sure ChatGPT was giving me good answers.

It’s important to note that you get from Generative AI what you put into it. If you just type, “ideas for wilderness hazards,” you’ll get a bunch of random crap. That’s fine if it’s all you want, but the more detailed you are and the more qualifying questions you ask to refine the answers, the better the answers you’ll get.

To illustrate that point, I asked ChatGPT to provide a dozen possible wilderness encounters I could fling in front of a 3rd level party in a boggy marsh in the forgotten realms. One encounter it suggested involved ghostly duplicates of the party, so I asked it what game mechanics it might use to build that encounter. Another encounter involved a sinking dungeon in the mire, so I asked it how to track the dungeon’s sinking and how I might map it. You can check out my conversation with ChatGPT about boggy marsh encounters if you want.

Again, the key is to think of it like an intern: give it detailed tasks, ask it to dig up more information on whatever it finds, and use whatever crap it hands you very carefully.

And for the love of fuck, if you use Generative AI to make content you intend to sell or publish, at least disclose very clearly that you did so and make it clear where your work ends and the robot’s begins. Can we at least agree on that much?

Summer asks…

Are you still developing a roleplaying system of your own? If so, what are optimistic and pessimistic estimates for when it will be available for purchase?

Yes, Summer, I am definitely developing a roleplaying game system of my own. In fact, I’m developing two.

It’s been a goal of mine for several years now to publish my own roleplaying game system. And even if it weren’t, it would have been after January of 2023 given that Wizards of the Coast made it clear that they’d like to be the only ones publishing D&D content forevermore. And if you think they dropped that goal just because they backed off and slapped a Creative Commons license on parts of the current edition of D&D, I predict you’re in for some nasty surprises in the next eighteen months.

That said, because of the glut of Dungeons-&-Dragons-likes that we’re going to see choking the space for the next year-and-a-half and that WotC also has a very shiny, highly polished turd of a new D&D revision coming soon, this ain’t the best time to throw a big project at the market. So I’ve got something smaller to hold the fort coming next year.

I have a small-scale fantasy adventure roleplaying game engine in the testing stage right now. In the next two or three months, select members of my supporter community will get to play their first adventure using the SLAPDASH Engine. And by year’s end — 2024 — I plan to have a core book available for sale. If you like fantasy adventure but think D&D needs some serious improvements of the kind I’ve been writing about for years, you’ll like the SLAPDASH Engine. And you’ll get a few years of fun out of it at a bargain price. But it ain’t a D&D-killer by any stretch. Just an alternative to keep my fans happy while they wait for…

I have a longer-term and much bigger project that’s a few years from completion. Again, it’s a fantasy adventure roleplaying game, but it’s a completely different take on tabletop roleplaying than anything currently on the market. I meant to move forward on it years ago, but there were a few obstacles and I’ve had to make a sound-business-type decision to put it on the back burner. I can’t really say any more about it, though. But it is coming someday.

Max asks…

It is possible that, in the New Year, to stop a “West Marches” style game from dying, I will need to co-GM with the founding GM (and be the first additional GM). By co-GM, I mean only one GM at the table at a time, but the same canonical world, events, and characters. Any advice or warnings?

Thanks for the clear, concise question, Max. Man, this Ask Angry thing is so much smoother when I do actually delete all the questions that don’t meet my very basic standards.

First, I should probably explain this West Marches shit for any readers not in the know. A West Marches game is a kind of open-world campaign that lets large numbers of players or players with chaotic schedules adventure together. There’s a big ole wilderness full of adventure to explore and there’s usually a homebase town in the middle. Players can show up to any scheduled sessions they want. Whoever shows up ventures forth together to explore the world in whatever direction they want and heads back to town when the session’s over. And that’s basically it, though there are lots of variations on the theme. The key is an open world without an overarching campaign goal and players dropping in and out for single-session expeditions. And most West Marches games are run by several Game Masters working together.

West Marches games are useful for accommodating school gaming clubs and lots of busy, adult gamers with hectic schedules consider them to be the shit. And I ain’t going to crap on West Marches games. They’re a useful way of overcoming certain real-life obstacles to running a campaign. That said, I see West Marches games the way I see Running Online. If it’s the kind of thing you have to settle for, it’s fun enough, but if there’s any other way to get a campaign going, you should do that instead.

West Marches campaigns aren’t terribly hard to manage provided the Game Masters keep good records and communicate well. The GMs in most successful West Marches games have regular, weekly meetings and keep good records. There’s also usually one GM that’s recognized as “in charge” even if only as the first among equals. That’s because people suck at working in groups. Someone’s got to keep everyone else organized, on task, and accountable.

If you said you were starting a new West Marches with a co-GM or two, that’s where my advice would start and end: pick someone to be in charge, set weekly meetings, establish a good record-keeping system. But that ain’t what you’re doing. You’re saving a dying West Marches game and that sets off all sorts of alarm bells in my skull. But you’ve given me no other information to go on.

The thing that tends to kill West Marches games is their high attrition. Gradually, players drop out or stop showing up and new players stop replacing them. Eventually, you’ve got sessions where too few players show up to even support an expedition. And that’s down to the drop-in-and-drop-out nature of things. It leads to inconsistent attendance, low commitment, and low overall investment. That’s just the nature of the beast. For every successful West Marches game with a stable base, there are at least twenty that just sort of wasted away and died.

But, given you’re stepping up to help run the thing, I suspect that’s not the problem. If there are enough games to need a second GM, it ain’t dying from attrition.

There’s really only one problem an extra GM can help solve. That’s that the game’s gotten too big for one GM to handle. And that usually isn’t what someone means when they say the game’s dying. If the Founder did invite you to help run the thing just because it’s too much to manage, you’ve probably got nothing to worry about. But if you volunteered because you — or your fellow players — saw a problem in the game and want to save it, well…

You might not want to get too emotionally invested. Give it your all, of course, and do your best, but also be prepared for the game to fail despite your efforts. Emotionally prepare yourself for the game to die. Which, frankly, is good advice for all Game Masters. Accept that games die sometimes.

Beyond that, there’s only one other general piece of advice I’ve got and it’s the same advice I’d give anyone when they’re joining a team as a new junior member. Do not go in thinking you’re going to change anything. Especially not right away. Accept that, for at least the first month, your job is to help the Founder run the game. To learn how the Founder does things and to follow suit. Until you’ve done that for a while, you’re not qualified to fix anything. Even if you see a really obvious problem you know exactly how to fix on day one, you still need to sit on that shit for at least a month before you open your mouth. That’s how it be.

Keep in mind, too, that this is the Founder’s game. The Founder may feel very attached to the game they’ve built and you’re a guest in that game. They’re trusting you to help run the world, but it’s still the Founder’s world. For all intents and purposes and for a long time to come, you’re an assistant. Even if the Founder has insisted that you are a full partner and that they want all your input, you are a helper and they are in charge. If you can’t function with that understanding, don’t take the job. Because no matter what the Founder says, that’s how they’ll probably feel.

Because I’ll tell you the one thing that definitely won’t save your game: conflict between its GMs.