Social deduction games are, at their core, contests between an informed minority and an uninformed majority. The uninformed majority attempts to solve a deductive puzzle: identifying who is part of the informed minority. Conversely, the informed minority seeks to disrupt this process by confusing the majority into misidentification. Typically, we assign positive or neutral terms to the majority—villagers, crewmates, the Good Team—and negative terms to the minority—mafia, imposters, the Evil Team.
In traditional social deduction games, the mechanics operate under a theory of total elimination or identification. In Mafia, villagers must identify all mafia members before the mafia eliminates the villagers. The same applies in Among Us. In Blood on the Clocktower, however, the Good Team only needs to identify one member of the Evil Team: the Demon. This can be called a "flagbearer" mechanic.
On the surface, the flagbearer mechanic seems to favor the Good Team—they only need to identify one Evil Team member, and they can do so through various means. In Mafia, they must catch every mafia member; a single well-blended-in mafia member can thwart the villagers’ victory.
However, Blood on the Clocktower is notably more balanced in favor of the minority. While villagers win about two-thirds of the time or more in Mafia, the Good Team wins only about half the time in BotC.
How does the Evil Team achieve this success? And how does Evil leverage this to increase its chances of victory?
This essay will focus on how the Evil Team can utilize the flagbearer mechanic to their advantage.
A. The Flagbearer Mechanic
The flagbearer mechanic implies that the only thing that truly matters for the Evil Team is ensuring the Demon token reaches the end of the game safely. I refer to it as the Demon token because BotC scripts generally include ways for the Demon token to move from one player to another—for example, the Scarlet Woman and the Imp in Trouble Brewing; the Barber, the Pit Hag, the Snake Charmer, and the Fang Gu in Sects & Violets; the Hatter, Lil Monsta, and Legion in custom scripts. Even in Bad Moon Rising, where the Demon token cannot move, the Mastermind effectively makes all Evil players equivalent to the Demon on the final day (and all Good players equivalent to the Saint).
The main effect of the flagbearer mechanic is that it’s not enough to keep the Demon from being executed today; the Demon must also survive tomorrow. Evil teams lose when they prioritize executing a Good player today over safeguarding the Demon’s survival in the future.
The fundamental task for the Evil Team is to identify and cultivate a plausible alternative explanation to the Town's logical puzzle—one that does not rely on the Demon being the Demon. This involves creating and nurturing particular players as framing targets.
If the Evil Team frames a player too early, problems can arise. Consider a typical eleven-player game of Trouble Brewing. Alice the Imp bluffs the Fortune Teller with a Yes on Bob and Carla. David the Spy bluffs as Empath with zeros on Carla and Eddie. Francine the Baron bluffs the Washerwoman, confirming David. Based on these bluffs, the Town executes Bob, the beleaguered Mayor, on Day 1. This seems advantageous for Evil!
But here’s the issue: Evil has invested substantial effort into framing Bob, but now Eddie, the Undertaker, learns Bob is the Mayor, confirming him as Good. With Bob confirmed innocent, suspicion shifts to the Fortune Teller, Empath, and Washerwoman—who conspired to frame him.
A good frame must be nurtured over time, allowed to develop, and saved for the endgame. It’s in the final stages that you want to ensure the execution targets a plausible non-Demon.
Each day in BotC, the Evil Team should consider which execution makes it least likely that the Demon will be killed that day or in the future. Evil loses control when they execute the wrong Good player at the wrong time—as this removes someone they can frame later and makes the Demon a stronger candidate. The Evil Team should often sacrifice minions to prevent frameable townsfolk from dying too early. Otherwise, the Demon becomes more likely to die.
Similarly, when the Demon or its minions select a target at night, they should avoid closing paths that leave only the Demon as a plausible candidate. For example, in a nine-player Trouble Brewing game, with the following players:
Anna the Imp
Bob the Butler
Cindy the Fortune Teller
Dan the Baron
Evelyn the Saint
Frank the Recluse
Genevieve the Drunk (Monk)
Harold the Undertaker
Ivana the Librarian (who saw Genevieve and Dan as the Drunk)
On Day 1, Dan is executed, and Anna kills Cindy, who claimed No on Genevieve and Harold. At first glance, this seems good—the powerful Fortune Teller is gone. However, Harold then confirms Dan as the Baron, which also confirms Evelyn, Frank, Bob, and Ivana’s claim of seeing Genevieve and Dan as the Drunk. Furthermore, Harold is confirmed not to be the Demon by Cindy and will be difficult to frame. This situation leaves only Anna as a strong Demon candidate.
Alternatively, if Anna kills Genevieve and Cindy reports a Yes on Ivana and Evelyn, more potential worlds open. This example illustrates that the Evil Team’s night actions should consider how to keep their options open and how their kill choices influence the balance of potential Demon candidates.
I often prefer to frame Fortune Tellers rather than kill them, but I do kill them when it makes sense to do so, namely when they are trusted and their information doesn’t point back to me. The key point is that the Evil Team must evaluate which approach offers the best chance for victory—not just always kill the Fortune Teller or always frame them. If a Fortune Teller is heavily trusted and has likely hit a red herring or the Recluse, it might be better to kill them. Conversely, if they have identified the Demon, leaving them alive can be advantageous. The decision depends heavily on the overall game state and what roles are in play.
The main goal for the Evil Team is not merely to eliminate powerful Town players but to keep the Demon alive. For example, in Baron games, if it becomes clear that Outsiders are trusted and the Baron is in play, it’s usually better for the Demon to eliminate Outsiders and leave powerful Town roles available to frame. If the Poisoner is present, poisoning a trusted Fortune Teller might be more beneficial than outright killing them. Still, if the Fortune Teller has correctly identified the Demon and caused their execution, the Scarlet Woman should generally kill the Fortune Teller.
In summary, whether to kill, frame, or poison information roles largely depends on the social and informational landscape, all in service of the ultimate goal: keeping the Demon alive, not necessarily eliminating information.
Night kills and supported executions should focus on preventing the Demon’s death, rather than winning individual battles. Evil must craft their frames to survive into the endgame and target players at night who are difficult to frame.
B. Minions: Death is My Gift
Minions are arguably the most fun roles in the game. While this is subjective, many players find them highly engaging due to their high agency, significant impact on the game, and their varied and often creative abilities. This is because minions operate in a loosely constrained design space: Demons need to fit the pattern “each night, choose a player to die,” with some added bonus abilities and drawbacks. Townsfolk, on the other hand, need to be part of the deductive puzzle. Outsiders face numerous constraints—they can’t be just killed or must have exploitable drawbacks. Minions, however, are generally freer to push the boundaries of what makes for fun and interesting gameplay.
Because of this freedom, there’s no single comprehensive guide to playing minions—each role, such as the Poisoner, Scarlet Woman, Goblin, Baron, Wizard, or Wraith, has its own incentives and strategies. Nonetheless, all minions share an essential principle: death in service of the Demon is always beneficial.
Referencing the execution budget discussed in my introductory strategy guide, in a typical ten-player game, there are four opportunities for the Good Team to execute players. If the minions can arrange to be two of those four, they effectively halve the Good Team’s execution options, increasing the chances that the Demon survives into the endgame. Since Day 1 executions are largely unpredictable, being the second or third to be executed means the Demon has a higher likelihood of avoiding early death, leaving only the final execution as the critical moment.
This principle applies to all minions, including those who might not want to die. For example, I once played as the Goblin and chose not to claim it to avoid execution—ensuring that my Demon remained unscathed and preventing the Good Team from killing the Demon prematurely.
This isn’t new insight. The Good Team, even in inexperienced groups, quickly learns they should avoid executing minions in the mid- or late game unless necessary. Evil, however, often fails to embrace this: players are reluctant to vote out their own minions, even if doing so improves the Demon’s survival odds. The key mantra is: death is the gift of minions. Dying at the right time typically helps Evil on average.
This principle also extends to nighttime kills. If a minion is trusted and bluffing convincingly, killing them at night can further Evil’s goals more than leaving them alive. For example:
Most of Town trusts Abe, the Poisoner, who is bluffing Ravenkeeper, and Bianca, the Soldier. Carl, the Imp, and Denise, the Mayor, are the only Demon candidates.
Carl has to choose who to kill. Abe will likely poison Bianca, as a trusted Soldier who cannot otherwise die. Carl could kill Bianca, but this will make it look like a Poisoner is alive, potentially undermining Abe’s status as a trusted player. If Carl kills Abe, Abe can claim to see Carl as his bluff or see Denise as the Imp. This significantly increases the chances that Denise will be executed. Additionally, it makes sense that the Ravenkeeper would die when the other players are Demon candidates and the Soldier is unkillable.
This principle holds true in less dramatic examples. If the Baron is bluffing the Fortune Teller, their bluff is more credible if they die in the midgame, and their “No” on the Demon is more persuasive than it would be if they were alive. By dying at night, minions gain social trust to back up their bluffs. In short, well-timed deaths help Evil control the narrative.
People accuse me of being callous toward my minions’ lives. I am If you are on my Evil Team and you are a minion, I will sell you out if Town is suspicious of you and kill you at night if they trust you. Except in rare cases where a three-Evil-alive endgame seems possible, I always cringe to see a minion survive to the endgame. Minions exist to die, and failing to die opportunistically risks the Demon’s life.
A slight addendum: While death is frequently the most powerful thing a minion can provide the Demon, like everything there are trades offs. Voting and nominating power isn’t meaningless. If a minion’s death does not serve Evil’s goals, you should of course keep them alive. But Evil is more frequently tripped up by keeping minions alive when their death would serve a greater purpose (because they are trusted more than the Demon, or because their execution will lead to a endgame frame surviving), and Evil generally should let their minions die more, or kill them.
C. How Evil Bluffs
In Part 1, I discussed how townsfolk broadly fall into two categories in terms of information: confirmation and conflict. Confirmation focuses on information that identifies a player as Good and is usually resolved without executions. Conflict, on the other hand, focuses on information that identifies players as Evil and can only be resolved by executing conflicting information, using a valuable resource.
These categories go beyond strictly speaking information. Some characters can cross the line. For example, the Monk can largely (though not entirely) confirm players by protecting them at night and waking up to see no one died. The Choirboy has both a confirmation aspect (the King is Good) and a conflict aspect (John, who killed the King, is the Demon). The Village Idiot either confirms a player or creates conflict with them every night.
Outsiders exist in a slightly different space. They are frequently confirmatory, as outsider count creates confirmation opportunities, and some outsiders have effects that should be noticed by Town. On the other hand, the known outsider count creates the risk of conflict between outsider claims—if there is no way for three outsiders to be in play, but three are claiming outsider, one must be Evil.
These are important considerations for Evil when deciding on a bluffing strategy. Evil teams frequently trip up when no conflict information is generated, as they are left without Good players to frame. But Evil also gets tripped up when they lose a conflict fight and a critical Evil player (such as the Demon) gets executed as a result.
As a baseline rule, I prefer, as Evil, to bluff confirmation as the Demon and conflict as a Minion. If conflict gets resolved by my execution as a Minion, great. Confirmation puts me at less risk as a Demon because Good players have no reason to be suspicious of my correct information.
That’s only a general rule, and it has an obvious flaw: if players realize only confirmation is where the Demon is likely to be found, they will adjust their executions to target confirmations. Even worse, they’ll start with confirmation that’s easier to fake—confirmation that detects alignment, not a specific role. Over time, players who don’t die become suspicious, even if they lack this meta read.
Still, the general idea holds: Minions should create conflict with their bluffs, and Demons should generally avoid it. Most of the time, if a Minion can create conflict, it will result at minimum in the Minion being executed (which, as explained above, often helps Evil on average). At maximum potency, it can lead to one or more Good players being executed, which is even more advantageous. Less strategically important, conflict information bluffs from Minions can fill information gaps that might otherwise slow a game—for example, a Trouble Brewing game without an investigator to spark day-one conversation can drag due to lack of execution targets.
For the Demon, becoming part of confirmed information is often powerful. Players are often hesitant to allow a Minion to bluff the Washerwoman confirming the Demon as, for example, the Soldier. If anyone is likely to be executed as a result of this bluff, it’s the Washerwoman, while the Demon gains credibility. Even better, the Washerwoman can be killed at night, which supports the Soldier bluff and gives the Demon a built-in excuse to stay alive.
Meta considerations complicate this. Demons sometimes need to bluff conflict information to avoid becoming predictable. For instance, a bluff of Empath 1 on the Poisoner can make the Demon look good if the Undertaker sees the Poisoner as such. One of my first Demon wins involved sacrificing a Minion via a true fake Investigator ping, letting their Evil nature become apparent over time.
Flexibility is key. Players may meta-read early, correct pushes on Minions as a Demon, so the Demon must sometimes bluff incorrectly to mislead the Town. In one game, my Demon was shot by the Slayer on Day 1, which led to the immediate execution of their neighbor, the Marionette, leaving only me, the Scarlet Woman. I had also been bluffing Empath with a 1; my neighbors were the Virgin, and the Town ultimately nominated them. The game resulted in a win, thanks to leaning into a bluff as the Drunk, which helped satisfy outsider count checks.
A note on Minion bluffing: losing all social trust and being executed is not a reason to stop bluffing. Evil players who withdraw when discovered are liabilities. In healthy playgroups, Good players will still interact with you; you should engage them, mixing truth and lies.
Minions should actively bluff information and create conflicting narratives. This can include confusing outsider counts, discrediting Monks by claiming Soldiers, or pushing active information bluffs—people follow information to guide executions. If Evil does not create narratives to counterbalance the Good Team, they will likely be overwhelmed. While risky, losing a bluff as a Minion is often beneficial.
Demons should also actively bluff to avoid meta reads suggesting only Minions manipulate the information landscape.
This may not seem like the most helpful bluffing advice because it is somewhat circular. If I were to distill my bluffing advice, it would be as follows:
As the Demon:
The best bluffs place you in confirmation chains with Good players, ideally being the confirmed player or confirming others as Good.
Don’t be afraid to make synergistic bluffs where a Minion confirms you.
Pay attention to meta reads and adjust accordingly.
If all else fails, bluff as a Minion.
As a Minion:
Bluff in ways that create conflict with Good players. If you win, you can direct a valuable execution toward a Good player. If you lose, an execution spent on a Minion is often beneficial.
If trusted, find ways to confirm your Demon.
Never stop bluffing, no matter how suspicious you appear.
Be vocal—offer information frequently and create narratives that serve Evil’s purpose.
If you play repeatedly with the same group, constantly shift your bluffing strategy to remain unpredictable.
D. Trust Problems
The final point I want to make about generic Evil Team strategy involves how the Evil Team should engage with the trust economy. As I mentioned in my introductory essay, trust itself becomes untrustworthy if the Evil Team doesn’t act into it. Here’s an example from Trouble Brewing:
Abbie, the Poisoner, bluffs as the Washerwoman confirming Benjamin, the Imp, as the Monk. Both Abbie and Benjamin sell the confirmation well to the Good Team, and no one discusses executing them for several days. On Day 4, Caroline, the Ravenkeeper, asks why the Monk and their Washerwoman are still alive, given how trusted they are. No one can answer this question, so they decide to kill Benjamin just to be safe. Good wins.
You can probably recall a game where something similar happened. Evil’s bluff was too convincing, rendering Evil players overly trusted. Their failure to die at night became suspicious, causing the Good Team to turn on them. This creates the central trust problem for Evil: actually trusted players die at night, while extremely suspicious players are executed.
It thus behooves Evil players—particularly the Demon—to find a middle ground: be suspicious enough that not dying at night is not weird, but not so suspicious that the Town wants to execute you. This is also the space where Evil wants their frames to live.
Importantly, this ambiguous-trust zone also allows several Good characters to maximize their abilities. I’ve previously written about how a Monk or Slayer who lands in this space can extract maximum power in the late game. Fortune Tellers and similar recurring information roles can obtain maximal information. Mayors who operate in this space are least likely to die at night for balance reasons: they are less likely to be targeted initially, and less likely to become so trusted that Evil cannot win.
Mid-trust strategies can be risky for Evil, but they are essential: they allow frames and bluffs to survive into the endgame and give Evil the ability to manipulate opposing votes in final stages. Evil should also identify and frame Good players who land in the mid-trust zone because these players:
Are unlikely to be executed before the final three, ensuring bluffs and frame efforts do not fizzle.
Can generate opposing votes in the final three.
Conclusion
This essay is meant as a generic introductory guide. I plan to cover bluffing in the various base-three scripts and as specific Evil characters in future essays. It is necessarily non-comprehensive, but it should help you begin thinking about how to play as Evil, or at least demystify playing Evil and get it to a place where you don’t freeze up during gameplay.