Alex Nichiporchik

DUCKSIDE: An April Fools Joke which wasn't a joke

On April 1st we announced what many speculated was a traditional game industry joke, a fake meme game. It wasn't a joke though.

Alex Nichiporchik here, I'm one of the producers on DUCKSIDE, a survival game where you play as a duck. Here's the drilldown and story behind the game, where we'll take a look at some stats.

First off ,here's the announcement trailer

You can clearly tell from the comments, people started speculating if the game was fake. April 1 nonetheless.

Similar reactions across social media. We were in a bit of a pickle of how to position it, because not many games have what I refer to "honest PVP" in survival settings. This means no level locks, no RPG-style progression, you can use whatever loot you obtain, and are able to build bases, raid these bases, and a lot of the gameplay revolves around social interactions. After tens of thousands of hours in games like DayZ and Rust, the only real way was to say "Rust and DayZ but you're a duck". Sincere apologies to the Facepunch team if this was annoying at all, I love you guys and Rust in general.

But wait, isn't this just DEADSIDE?

Let's go back to the beginning. To December 2023.

Going back to late last year, the team behind Hello Engineer had just shipped their titles and we were in deep research & development on what the next project might be. We really wanted to build something with an open world, systems based design. The 1,000 hour game.

In this presentation from last year I talk about how to design thousand hour games

INITIAL PROTOTYPING

Believe it or not, our initial prototypes were Rust mods. We were exploring the idea of making a "cozy" survival game in a zombie apocalypse where you travel around on a boat. Long story short it would require a lot of content if it was a co-op game, and it would have to become in a way an extraction shooter if it was multiplayer PVP. Forcing players to build on water in a boat-base required mechanics to get players off water. We played a full wipe on a modded Rust server where tides would come in and flood the island, forcing players to get onto boats. It was fun for players who started on the hour of the wipe, and became unplayable for anyone joining later. The joy of these survival games is being able to join at any time after the game officially started (the "wipe"), and being able to dismantle bases that are decaying -- for loot and glory!

It turned into a swim fest. You'd spawn on the few pieces of land left, and have to spend lots of time swimming around. This brought up the main issue I have with survival games - it's the running. Traversing a large map as a low tier player is a little bit tiring. I play a lot of modded survival games with systems like teleports and dreaded the idea of building a game that's a running (swimming?) simulator.

"Yes but what if ducks?"

After weeks in design limbo with the boat prototype, we had an epiphany moment. What if we borrow the engine of DEADSIDE, our flagship survival game made by studio BadPixel, and have a flying character controller? What if it's a duck, goose, bird? Could this even work?

Solves the problem of running in survival games
Can have a unique hook, a twist that "lol you're a duck"
New shooting gameplay where it's all about bullet drop and aim skill, not just reactions because most of the combat would happen mid-air vs holding tight corners in traditional shooters
Deceptively cute yet extremely hardcore with base building, raiding, and tears
Players who never played or "felt" the true pain of gamesl ike Rust may not know they'll enjoy it, and won't be put off by a severe skill gap

Within the first week we had a working prototype with clunky flying, and DEADSIDE's gunplay. The very first time we got onto the server the whole team was sold. Wellp, we're making a duck survival game.

Let's announce a joke and do an actual playtest

After transitioning the game over to Unreal 5 and spending a ton of time to make the flying feel great, we had a production plan set out. We would aim for an honest, completely public playtest mid-April. We'd announce the game on April 1st with a proper Steam page and open up sign-ups. The critical part here was to get enough sign-ups so that when we open up the playtest, it would get a lot of people into the game.

The 3-week playtest

It's all about the servers. There is a reason there aren't that many games with dozens of players on a map, a world you can build in, etc. It's extremely difficult.

With games like these it's so difficult to tell when your servers will "break". The idea was to indeed break them. We launched the playtest to over a thousand online players concurrently, and very quickly found the breaking points. A combination of the amount of player-made structures, active NPCs, and player count were the balancing acts.

The playtest started off with a bang and everyone loving the game, only to devolve into a rubberband fest a couple of days after start. While working on the solution, we started wiping servers every day or two. The team worked through weekends and stabilized the build.

The biggest issue we had initially is that player made bases wouldn't decay. So you have bases on a server being built, and if abandoned, they wouldn't go anywhere. Eventually servers would reach their max capacity and start to lag. At one point I was on a team of 8 ducks that were going around different servers and "cleaning" them, At 3am we'd be going to lagging servers, farming for explosives, and destroying abandoned bases. That was fun.

Hilarious sponsored vid with Fisk

As soon as a decay system was introduced, we also lowered the player cap from 50 to 30 temporarily. This came along with server-side and client-side optimizations which made the game run silky smooth. It's also when we started getting tons of traction from large influencers joining the game.

I got to play with some of the best Rust players, and we got raided

THE STATS

Playtest CCUs, players, conversions

Over 150k players signed up to the test
Over 80k players actually launched the game
- This is above our usual conversion rate into players
- 54% of all sign-ups converted into players
1,600 peak CCU
First weekend got botched because of the server lag
Averaging 700-1000 CCU during the playtest

The Thousand Hour Game

As soon as we announced the playtest was coming to an end, many players voiced their sadness. And then I saw this screenshot

Our discord has a few people that have played hundreds of hours during the span of 3 weeks. It's an amazing and wholesome feeling.

I won't share full play stats as they're a little bit skewed and sensitive (because we don't have an accurate way of separating players joining before the lag issues and after), but here's the gist: