Blades in the Dark is Bad at Running Heists
My friend ran Scum and Villainy for some friends and I for about six months now, and I haven't been enjoying myself as much as I thought I would've when he posted the blurb. Running heists in a fantasy setting sounds like something I should've loved. Firefly, Cowboy Bebop, Star Wars and the good bits of the Expanse. Imagine my surprise when I really wasn't enjoying myself all that much.
This is particularly galling because Blades in the Dark is held up as this indie darling - a really well-designed, tight game that solved fundamental problems with traditional RPG design. It gets praised constantly for its elegant mechanics and innovative approach to heists. That has not been my experience. It's been frustrating and tedious and more than once I was on the verge of bowing out because I thought it wasn't for me.
I had major problems with it. Major ones which made me at times zone out in sessions. At first I thought it was just some wonky mechanics; an iffy skill system, a character advancement system entirely disconnected from the goals of the players, lack of referee direction which made sessions grind to a halt while my friend tried to come up with another complication caused by a third failed roll in a row. But then the realization hit: the core concept of the game is ultimately and fundamentally flawed in a way which prevented me from ever enjoying it. Blades in the Dark is bad at running heists.
You're not actually being clever
The no-preparation heist concept is fundamentally broken. When you're sitting at the table, it feels like being along for the ride, rather than being clever criminals in control. Flashbacks to find solutions are retcons, not clever preparation. The inventory system suffers from the same issues - retroactively deciding what you brought removes any satisfaction from actual forethought.
This removes the satisfaction of actual planning and preparation paying off. You're not playing criminals - you're playing improvisational storytellers who happen to be wearing criminal costumes.
Here's what this looks like at the table. I was running a brawler - effectively just Brock Sampson with the serial numbers filed off. We'd just fought our way off a station where we'd stolen some thing or another. We were getting shot to hell and the ship we stole as cover was barely limping along. I had the idea to use a flashback to say we'd programmed our ship to autopilot under stealth to an orbit around the station so we could use that to get away while overloading the reactor of the stolen one.
It was one moment of "huh that's clever" but it didn't feel like some kind of machiavellian planning. It felt more like having a magic lamp we'd rubbed and said "wouldn't it have been nice if we'd thought to do this." There's no satisfaction in that kind of cleverness - it's just narrative convenience disguised as criminal competence.
The system is trying to emulate Ocean's Eleven cinematography - specifically the 2001 remake, not the original. But what works on celluloid doesn't work in RPGs. OK, yeah - they thought planning a heist at the table would be a huge waste of time and probably take up most of the session. But that misses the point entirely about what makes heist games fun.
This stylistic film choice becomes terrible game mechanics. It creates fundamental player agency problems. The fiction says you're masterminds, but the mechanics say you're reactive improvisers. You get told you're brilliant criminals, but the system never lets you actually be brilliant.
Directly copying Ocean's Eleven at the table doesn't work
When you watch Ocean's Eleven, you're not impressed by Danny Ocean's ability to make up solutions on the spot. You're impressed by the elaborate preparation and how all the pieces click together. The flashback reveals show you the cleverness that was already there, not improvised cleverness being created in the moment.
The system prioritizes narrative convenience over player investment. It confuses cinematic pacing with player engagement. The designers treat preparation as boring instead of recognizing it as the core criminal fantasy. They value improvisation over competence and planning, which completely misunderstands what makes heist stories compelling. It's effectively a confusion of cause and effect; thinking that the stylistic choice of showing preparation in flashbacks is the important part, rather than the actual act of preparation.
What might actually work
So what would a fixed version of Blades be like?
Firstly and most obviously: remove flashbacks (and the inventory system) entirely. They fail at their intended goal. But let's do a nod to some of what the designers seemed to have thought, and something I'd generally agree with. Preparing in session is a huge pain and a huge waste of actual play time.
Instead, let's repurpose downtime for actual heist preparation. Instead of "I flash back to when I prepared for this," it should be "In downtime you attempted to get a floor plan of the casino; you paid off guards for a 10-minute break in the rotations." This would reward forethought instead of retroactive cleverness. Let players actually plan and see their plans pay off - or fail spectacularly because they missed something or got unlucky. That's where the drama comes from, not from whether you can come up with a clever story about what you supposedly did earlier.
Here's how this could work: your crew wants to infiltrate a corporate data center. During downtime, the hacker uses their Tinker skill to research the facility's security protocols, the face uses Sway to cultivate a contact among the cleaning staff, and the infiltrator uses Study to memorize guard patrol patterns from publicly available security footage. But when you arrive, complications emerge from your preparation - the floor plan you acquired is six months outdated and there's a new server room, your cleaning contact mentions they've switched to biometric locks since last week, and those patrol patterns were deliberately varied after a recent break-in attempt. Now you have real advantages from your work, but also real complications that your foresight couldn't have predicted. It's like the "older code, that still checks out" in Star Wars that tipped off Vader - your preparation is real and useful, but it also creates new problems you have to solve on the fly.
The best heist moments come from when your preparation meets unexpected complications, not from making up preparation after you already know what the complications are.
Another take on a TTRPG heartbreaker
This could have been a great heist game with different design priorities. Maybe even more strongly: it would've been a system that I'd be happy to play years of returning sessions in, while now I'm fine with us shutting it down after six months. It's a heartbreaker rpg. Not the kind where the creator will get his heart broken because no one plays it after he put his heart and soul into it, but because the players see the pile of disappointment and broken promises that make up the system.
The bones are there - stress as a resource, retirement mechanics, crew advancement, hell, it even has great mechanics for forcing players to play a stable of characters! All of that works. But the core flashback system undermines everything else. It shows how chasing cinematic feel can destroy the actual gameplay experience, turning what should be a game about criminal masterminds into a collaborative storytelling exercise.
This isn't just about one game being disappointing. It's about a design philosophy that prioritizes narrative aesthetics over player engagement. When we mistake the trappings of a genre for its essence, we end up with systems that look right on paper but feel hollow at the table. Blades in the Dark could have been the definitive heist game. Instead, it's a cautionary tale about what happens when you let style override substance.
Heist stories are one of the major genre fiction staples, and they deserve a game dedicated to them. Sadly, they deserve something better than this.