Team17: From Amiga Shooters to Worms Forever

Team17 is the only British studio of the Amiga generation still standing, still listed, and still shipping. That’s the headline fact and it deserves better than the shrug it usually gets, because the survival rests on two structural decisions taken about twenty years apart, and the second one is more interesting than the first.

The first: in 1990 a Wakefield public-domain software library called 17-Bit Software merged with a Swedish demoscene group called Team 7 and started making games. The second: somewhere around 2010 the company stopped being a developer with a publishing arm and became a publisher with a development habit. Between those two decisions sits five years of the best action software on the Amiga and thirty years of worms.

A library and a demo crew

17-Bit Software was Michael Robinson and Martyn Brown selling disks of public-domain and shareware Amiga software by mail order. This is worth sitting with. Before they made a single game, the founders’ actual expertise was distribution — knowing what people wanted, what they’d pay, and how to get a floppy into a bedroom in Barnsley.

That’s an unusual education for a game studio and it shows in everything that follows. I’ve written about the covertape as a distribution model and the PD library was its scruffier cousin: a network that ran on postal orders and word of mouth, entirely outside the retail chain the big publishers controlled.

Team 7 were Swedish demosceners — the culture that produced people who could make an Amiga do things Commodore’s engineers hadn’t budgeted for. I’ve argued that the demoscene’s shadow over game design is longer than anyone admits, and Team17 is the cleanest single proof. Merge a distributor with a group of people who optimise for spectacle-per-cycle and you get a company that knows exactly what a sixteen- year-old wants and can technically deliver it.

The Amiga years

Full Contact came first in 1991, and then Alien Breed in the same year, and the run that follows is about as strong as any British studio managed on the machine.

Alien Breed is a top-down run-and-gun that is, with total transparency, an Aliens fan film you can play — I’ve written the whole case. What matters for the studio’s story is the execution. It’s fast, it’s loud, the two-player co-op is real, and the intercom text and the self-destruct countdown do more atmospheric work than the polygon budget of games ten years its senior.

Project-X (1992) is a horizontally scrolling shooter of genuinely unreasonable difficulty and genuinely unreasonable presentation — the studio flexing its Swedish half. Superfrog (1993) is a platformer built to prove the Amiga could do a mascot game while the mascot platformer boom was eating the console market. Body Blows (1993) is a Street Fighter II answer, and it’s the game where you can see the limits of the approach — technically fine, mechanically thin, because a fighting game lives in frame data and Team17 were spectacle people.

The pattern across all of it: Team17 games looked and sounded better than they had any right to, played fast, and rarely had a deep system underneath. That’s a demoscene inheritance, and it’s a real limitation. Alien Breed’s rooms don’t develop the way a Doom level develops. Project-X is a wall of stuff. The studio’s craft was in the delivery.

Which makes what happened next genuinely funny.

Total Wormage

Andy Davidson wrote a turn-based artillery game in Blitz Basic on an Amiga and called it Total Wormage. He entered it into a competition run by Amiga Format. It didn’t win.

He then took it to the European Computer Trade Show and showed it to Team17 directly, and they signed it. Worms shipped in 1995.

The thing to understand is that Worms is the exact opposite of every game Team17 had made to that point. It’s slow. It’s turn-based. It has no technical showpiece in it — the graphics are cartoon worms on a lump of destructible terrain, and any of the studio’s existing programmers could have drawn it in a week. What Worms has instead is a system, and the system is the best thing the company ever put its name to.

Here’s why it works. Artillery games — Scorched Earth, Tank Wars, the whole shareware lineage — are about solving a parabola. Worms adds three things that turn arithmetic into drama. The terrain is fully destructible, so every shot rewrites the problem for every subsequent shot. The wind changes each turn, so the solution you found is void. And your worms can move, which means position is a resource you spend time acquiring rather than a fact you’re dealt.

Put those together and you get a game where a comfortable lead evaporates because you blew a hole under your own team. The comedy is emergent — the game is funny because the physics are cruel, and the cruelty is a designed property. That’s a much better piece of design thinking than anything in Alien Breed, and it walked in the door from outside.

The turn structure earns its keep too. A turn timer forces you to commit to a shot before you’ve finished thinking about it, which converts a solvable geometry problem into a pressure test — the same conversion the Bitmap Brothers made by putting a shop on a moving bar. Worms without the clock is a maths exercise. Worms with it is a party.

There’s a lesson in that about where studios’ good ideas come from, and Team17 to their credit learned it. Slowly.

Worms forever

Then came the franchise, and the franchise is the story for the next fifteen years. Worms 2, Worms Armageddon (1999, still the one the competitive scene plays), Worms World Party, and a long detour into 3D — Worms 3D, Worms 4: Mayhem — that demonstrated comprehensively why the game is 2D. A parabola over a destructible cross-section is legible. A parabola through a volume is a guess. The series eventually retreated to 2D and has stayed there.

I count something over twenty Worms releases. Most are fine. Armageddon is the one with the deepest ruleset and the one that has been patched, by fans and the studio, for over two decades — a genuinely unusual afterlife for a 1999 game.

The reason Armageddon held and the sequels didn’t is worth naming, because it’s a design point rather than a nostalgia one. Armageddon shipped with an absurd surplus of options: weapon schemes you could rewrite wholesale, rope physics you could exploit, a level editor, and enough exposed parameters that the competitive scene could design its own game inside it. The later entries tidied that up. A tidier Worms is a smaller Worms, because the whole appeal is a system loose enough to be abused.

The honest read on this period: Team17 survived the collapse that took DMA into other people’s hands and Core into annual death by owning an IP that could be re-sold every eighteen months. That’s the whole trick. It’s also a creative dead end, and by the late 2000s Team17 was a company whose only asset was a joke about explosions.

The pivot nobody predicted

Around 2010 Team17 reinvented itself as a publisher for other people’s small games, and it worked so well that it’s now the main business.

The Escapists (2015). Overcooked (2016, by Ghost Town Games — two people). Yooka-Laylee. Blasphemous. Moving Out. Hell Let Loose. Golf With Your Friends. The company floated on AIM in 2018 on the strength of it.

Look at what that catalogue has in common: small teams, sharp single mechanics, systems that generate stories. Overcooked is a kitchen where the design problem is that your friends are in the way — the comedy is emergent from a cruel system, which is Worms exactly, at a different table. Team17 spent thirty years learning what Andy Davidson handed them in 1994, and then went and found other people who’d worked it out independently.

That’s a real editorial position, and it deserves the credit. A publisher with taste is rarer than a studio with a hit.

The record isn’t clean. In 2021 the company announced an NFT project built on Worms art, faced immediate and public objection from developers on its own label, and cancelled it within about forty-eight hours. It was a bad idea, publicly reversed, and the speed of the reversal is the only defensible part.

What the career actually shows

Team17’s arc is an argument about where design value sits. Their in-house craft — the demo polish, the speed, the noise — was real and it was replaceable, and every one of those Amiga games is now a fond memory rather than a live thing. The value that compounded came from systems: one unsolicited artillery game in 1994, and then a decade of buying other people’s systems and putting them in front of an audience.

For a company that started by selling other people’s disks out of Wakefield, that’s a career that ends exactly where it began. The distribution instinct was always the asset.

Where to start

Worms Armageddon, four players, one sofa, no time limits off. It’s the version the systems argument rests on.

Alien Breed on the Amiga for the other half of the company — an hour of it tells you everything about what the studio was good at and what it wasn’t reaching for.

Overcooked if you want to see the Worms idea running in someone else’s hands, thirty years later, at speed.