My best villain died of neglect.

Nobody killed him. There was no duel, no clutch nat 20, no tearful monologue on a collapsing tower. He just sat in my notes for eleven weeks while the party ran a fishing subplot, and when I finally brought him back for his big reveal, my rogue squinted at me and said, “Wait. Who?”

Who. WHO. I spent four hours on that man. I gave him a dead sister. I gave him a signature coat. I gave him a hat, and my players looked at me like I’d wheeled in a stranger.

So let’s get the uncomfortable part out of the way early. Your villain is probably fine. The motive is fine, the backstory is fine, the voice you do is fine (it’s a little much, but it’s fine). The problem is that your villain only exists on the nights you say his name out loud. The other six days of the week he’s in a drawer, and your players are living their actual lives, and a guy in a drawer cannot compete with a guy’s kid’s soccer practice.

Players don’t forget villains because the villains are boring. They forget them because the villains are absent.

Good news: that’s a scheduling problem, not a writing problem. Which means you can fix it in about ten minutes a week. Stay with me.

The number that rewired how I run bad guys

Here’s the stat that ruined my whole approach, in a good way.

An EN World discussion on campaign attrition cites something brutal: 37% of players basically never finish a campaign, with around half rarely finishing games at all, implying that fizzling is closer to the norm than the exception. Wizards’ own player data backs it up from another direction: 11% of players never progress past level one, 9% reach level two, and 15% reach level three.

Read that again. Most tables never reach the boss fight.

Every piece of villain advice on the internet is built around the finale. Foreshadow him! Build dread! Make the last battle land! Cool. Love it. But if your campaign is a coin flip to reach session twenty, you’ve been saving your best material for a party that may never show up.

The villain has to pay off now. In the gaps. Between the sessions, where the campaign actually lives in your players’ heads.

Why most villain advice can’t help you here

Go read ten articles about D&D villains. I did, so you don’t need to. Nine of them are character creation guides wearing a fake mustache. Motive. Flaw. Pet-the-dog moment. Sympathetic goal, monstrous method. All solid, all useless for our problem, because they answer “how do I build a good villain” and we’re asking “my villain already exists, and my players stopped caring in March.”

The procedural advice gets closer. Dungeon World’s Fronts are excellent. They’re collections of linked dangers built outside of active play, the solo fun you get between games while you craft the foes who’ll challenge the party. Mike Mearls put it about as strongly as you can: “Dungeon World’s concept of fronts improved my D&D games immeasurably.”

Blades in the Dark does it with clocks. Every faction owns a long-term goal, and during downtime the GM ticks forward the clocks they care about, so the world shifts in ways the crew isn’t directly connected to.

Both are terrific. Both are also written for your notebook. And here’s the trap I fell into for years: I ticked my clocks, I advanced my fronts, I knew exactly what my villain was doing on any given Tuesday, and my players still forgot him. Because the villain was doing all his best work in a Google Doc nobody was allowed to read.

Blades actually spells out the missing half, and I skipped past it maybe forty times. When factions do things known in the underworld, tell the players through a friend, contact, or vice purveyor, and let those rumors become future scores.

That’s not a prep rule. That’s a broadcast rule. The tick is half the job. The transmission is the other half.

So here’s my method. Three tricks. Ten minutes total. I call it the Off-Table Turn, because naming things makes them sound official and I am nothing if not a fraud with good branding.

Trick 1: The Subtraction Log (about 3 minutes)

Your players cannot feel your villain gaining something. “The cult acquired the third shard” means nothing to them. It’s a number going up on a scoreboard they can’t see, in a game they didn’t agree to play.

What they can feel is losing something.

So flip it. Between sessions, write one line in your notes: what did the party lose this week? Not gold. Not HP. Something small and warm that they liked.

Real example from my table. My party adored a bridge toll collector named Ostrick who charged them in riddles instead of coin. Total throwaway NPC. I invented him in four seconds because I needed a bridge to feel like a place. They loved him. They brought him snacks. One player wrote him into her backstory unprompted, which is the highest honor a random bridge man can receive.

So when my villain’s grip on the region tightened, I didn’t announce it. The party crossed the bridge and there was a new guy. Uniform. Ledger. No riddles. When they asked where Ostrick went, the new guy said he didn’t know the name.

That silence did more work than three pages of villain lore. My paladin went quiet. My rogue said a word I will not repeat here. And nobody needed a lecture about the villain’s expanding influence, because they’d just paid a toll to it.

If you play video games, you already know this feeling. It’s walking back into the hub town and finding the vendor’s stall boarded up. The game didn’t tell you the war was going badly. It just took your shopkeeper.

How to run it: pick one thing the party liked. An NPC, a road, a tavern, a shortcut, a rumor mill, a friendly dog. Remove it, downgrade it, or replace it with something colder. Don’t explain. Let them ask.

One rule: never subtract a character’s stuff. Take away the world’s warmth, not the fighter’s sword. One is a story. The other is a fight with your player in the parking lot.

Trick 2: The Rumor Drop (about 5 minutes)

Now we deal with the actual delivery problem, and this is the one almost nobody talks about.

Your campaign has a channel you’re not using. The group chat. The Discord server. The text thread where somebody posts a meme about the barbarian’s terrible decisions. Your players are already in there. Between sessions, that channel is the only place your villain can reach them.

Mike Shea suggests flash fiction for this: 100 to 500 words of in-world narrative, emailed out or shared on a campaign Discord, often following a villain or an important NPC and covering parts of the story the characters never witness.

Love it. Also, be real, I am not writing 500 words of prose on a Wednesday. So I shrank it.

Three sentences. One NPC voice. Posted midweek. That’s the whole trick.

Not narration. Not “the DM speaking.” Somebody in the world, saying something to somebody else. Here’s the format I use:

Overheard at the Copper Kettle, three days after you left: “They came for the granary at dawn, and they didn’t burn it. Counted it. Wrote it all down in a little book and left, polite as anything. My mother says that’s worse.”

Twenty seconds to read. Zero mechanical weight. And it accomplishes three things at once: it puts the villain in your players’ phones on a Wednesday, it makes the world sound like it kept running without them, and it seeds a hook you can cash in whenever you feel like it.

Know Direction has been quietly right about this for years. Build a villain’s infamy through gossip in settlements, and let some of it be exaggerated or false, because that keeps players asking questions and builds mystique. The false rumors are the fun part. Players love catching the world in a lie. Let one drop be wrong on purpose and watch them argue about it for forty minutes before the session even starts.

One rule: no reply required. The second your rumor drop feels like homework, engagement dies. Post it, say nothing, let it sit.

Trick 3: The Callback (about 2 minutes)

This one is small, and it’s the difference between your players feeling like the world is alive and your players feeling like you’re punishing them for having fun.

The standard advice is that villains should act on their own timeline. One villain guide puts it plainly: if the party spends three sessions shopping, the cult should advance its ritual by three steps, and if they chase a red herring for a week, the corrupt official consolidates more power.

Correct in principle. Awful in practice, if you deliver it wrong.

Because “the cult advanced three steps while you shopped” reads as: you had fun incorrectly, and now you’re being fined. Nobody wants that. That’s a parent, not a story.

So make it specific. Don’t advance the villain’s plan in general. Advance it on the exact hook the party walked away from.

Session six, my group got a lead about a missing shipment of surveyor’s tools. Boring. They ignored it because it was boring, because I wrote it badly. Fine. So the villain took it. Three sessions later they found out the tools were being used to map the aqueduct, and the aqueduct feeds the city they were currently standing in.

Same mechanic. Totally different feeling. Not “you goofed off.” More like “oh. Oh, that thread was load-bearing.”

This is what Dungeon World calls grim portents, and the framing is worth stealing wholesale: when a grim portent comes to pass, the impending doom grows stronger, more apparent, and more present in the world. More present. That’s the word doing the heavy lifting.

How to run it: at the end of each session, jot down one hook the party declined. That’s your villain’s next move. You’re not inventing anything. They already handed you the material.

“Hang on, isn’t this just railroading?”

Fair question. Somebody in your group is going to ask it, probably the one who reads game design blogs, possibly this one, hi.

Is it railroading? Nope.

Well. It can be, if you do it lazily. Let’s fix that.

The failure mode is deciding in advance that the villain’s move succeeds. That’s when off-screen action turns into an unstoppable cutscene, and players learn their choices don’t matter, which is the fastest way to kill a campaign that isn’t scheduling.

So prep the attempt, not the outcome. The villain tries to take the granary. Maybe the locals hid half of it. Maybe a rival faction got there first. Maybe he took it and it cost him a lieutenant, and now that lieutenant is furious and available. Give the move a real failure state and it stops being a railroad, because a railroad only goes one place.

The Alexandrian has been beating this drum for over a decade, and he flags the other trap too: background events are wonderful until they never become anything except background. If your villain’s off-screen turn never touches something the players can see, poke, or lose, you didn’t run a villain. You wrote fan fiction about your own campaign. (I’ve done this. Repeatedly. My notes from 2021 are a war crime.)

The ten minute budget

Let’s talk about why these three are all so small.

Sly Flourish polled 3,663 DMs on prep time for a roughly four-hour game. 10% spend 30 minutes or less, 33% spend about an hour, 28% spend about two hours, and 29% spend three hours or more.

So 43% of us get an hour or less. Total. For everything: encounters, maps, stat blocks, the tavern name you’ll panic-invent anyway.

Any villain technique that costs more than a few minutes is going in the bin, no matter how clever it is. That’s why the Off-Table Turn is three minutes, five minutes, and two minutes. It fits in the cracks. You can run the whole thing on your phone in a parking lot, which is where I have run it, more than once, five minutes before a session, like a naughty boy.

Do one of them this week

Don’t do all three. You’ll burn out by Thursday and blame me in the comments.

Pick the Subtraction Log. It’s the cheapest and it hits the hardest. Before your next session, look at your notes, find one small thing your party liked, and take it away without saying why.

Then watch their faces when they notice.

That’s the whole thing. Your villain doesn’t need a better backstory or a scarier stat block. He needs to be somewhere in your players’ week other than your notes. Ten minutes buys you a bad guy who exists on Wednesdays, and a bad guy who exists on Wednesdays is a bad guy your table will still remember in session forty.

Assuming you get to session forty. Statistically, uh. Let’s not talk about that.

Now go take somebody’s riddle man.