If you’re like me, the thought of playing any sort online multiplayer game sends you into a crippling spiral of frustration.
Your eyes gloss over, the magic of special effects show the horrors of dropped connections and rude messages reflected in them.
Okay, maybe that’s a bit overdramatic. I’ve had plenty of great experiences with online multiplayer, especially of the cooperative variety. Payday 2, Dungeon Defenders, and Tom Clancy’s Rainbow 6 all sit pretty high on my virtual gaming shelf.
But there are still those frustrations in any online venture, which is why I much prefer the experiences that can be had on a TV or two, sitting on a couch with friends uncomfortably and unnecessarily close to one another.
A couple weekends ago I picked up Ghost Town Games’ Overcooked. A friend and I began in the afternoon, fending off rats and running between moving trucks.
This bled into the evening when another friend joined in to take on the arctic ‘bergs and spooky mansion kitchens. As evening turned to night, friend number two (who we’ll call Chris, because that’s his name) had to leave due to “responsibilities” or some nonsense.
Friend number one and myself cooked on and completed the game that morning, around 2:30 in the am.
I was thoroughly impressed with what the game brought to the couch co-op genre, so I reached out to and talked with Phil Duncan, the art and design half of Ghost Town Games.
Something I really appreciate about the game is that it is a very compact, hand-crafted experience. What games did you get inspiration from in designing the UI, menus, and overworld?
We took inspiration from lots of different places, we’d constantly be pulling out games as reference as we went along: Mario, Pikmin, Splatoon, Micro Machines, even seemingly unconnected games like Fable and The Last of Us were referenced at some points for their UI or their interaction or movement.
The UI was very much function driven, we knew we had a lot of information to convey to the players and we wanted players to be able to understand it at a glance, so we iterated on that a lot, refining it from a text based interface to the more pictorial approach we have now and artistically we very much tried to echo the art style of the world and the characters: bright solid images which read well from a distance.
The overworld seemed to us like a great way of adding some cohesion to the game, rather than just having a long list of levels in some abstract menu somewhere, as soon as we added the concept of a world map the game felt much more solid. It also helped that multiple players could now control the selection of levels together, it’s a small feature but I like that we were able to make that aspect cooperative as well.
Random generation has been a pretty big craze in gaming lately. Was there ever discussion of random elements in the kitchens, or did you know from the start that each level should be handmade?
We discussed it a little bit, procedural generation is obviously a great way to create a lot of content much more quickly than hand-designing, but it just never seemed appropriate for the type of game we wanted to create.
When designing levels for Overcooked we tended to work backwards. We would look at the type of experience we wanted to give the player in that particular scenario then we would try to create a level which elicited that response.
For example “What if the players couldn’t see the map and had to remember the location of assets”, “What if only half the players had access to the chopping boards” etc. etc.
We would spend so long refining each level and tweaking them in various ways to make them more enjoyable we just didn’t think it would be possible to curate the content in the way we would have wanted with a more random process.
Each level sticks to the same formula: Four minutes to cook as much as possible.
How did you arrive at this being the best gameplay loop? Were there other modes that didn’t make the cut? For example: reach a point goal as fast as possible or survive an onslaught of orders for as long as possible.
We did a lot of user testing in the early stages, taking the game to various small conventions or to universities and getting students to play the game. The original loop for the early prototype involved players trying to survive as long as they could as the orders piled in.
You’d have three lives effectively and you’d lose one every time you failed an order. It was still fun but it just felt a little more stressful than the approach we have currently.
Players would complain that they felt like they were drowning the whole time and didn’t have a chance to just enjoy the process of cooking or of trying to coordinate with their teammates.
While it’s obviously not a direct simulation of a kitchen, we definitely wanted to have that thread running throughout the game and I think what we have now does a good job of mirroring how a restaurant kitchen works (serve as many customers as quickly as you can while the restaurant is open) while not beating the player down too much!
While many elements make regular appearances throughout the game (conveyor belts, shifting workspaces) others are used once and never seen again (darkness, rats). How did you determine what mechanics were better suited to just one level?
There [are] a lot of different factors which contributed to the selection of levels we eventually ended up with. I’m not going to lie and say it was all entirely by design, there are definitely elements of the game we would have liked to have explored more, or mechanics we just didn’t have time to implement more fully, but in general we created more levels with mechanics we felt we could draw more interesting player experiences from.
I think both Oli and I would have liked to have done more levels with rats in them and who knows it’s something we might get to explore more of in the future. Darkness is something we felt at the time was a little more one-shot, but we’ve had a few people asking for more dark levels. It would be great to incorporate more ghostly moving assets in a completely dark level so you can never fully rely on objects being where you last left them.
The point requirements for three stars can vary quite a lot from level to level, and from playing solo to playing with friends. How did you go about testing and setting each level’s goals? Did you reach a stage in development that you could estimate fairly well what the point requirement would be for a level?
As a two person studio this was definitely the hardest part for us to work on. We held lots of playtesting sessions where we would get groups of two, three and four players to go through each level and record their scores.
We signed up with Team 17 at the end of the project to help us market the game a little, but they also were able to help with their QA department to gather more data for level score boundaries.
When you’re creating a game like this, any small change you make to a level can drastically change the score players can achieve, there were lots of time where we might change the running speed of the characters, or the chopping speed or the length of time you have to complete an order and you’d then need to completely change the score boundaries for every single level and with every different combination of players.
It was definitely hard at the end of the project to just commit and say: this is the way the game is going to stay now so we could start finalising the score boundaries.
In the end we used a system which would record the highest score achievable on a level and the average score achieved by teams of varying size and capability, the further you get through the game the closer you as a team have to get towards the higher score.
It’s not perfect and, as with everything in the game, we wish we could have spent more time gathering more data to really refine it – but I think we’re pretty pleased with what we ended up with.
Regarding your signing up with Team17: you say this came about towards the end of the project; who approached who? Did you know from fairly early you’d want a publisher? If not, at what point did you decide on this?
We knew fairly early on that it was an option we would probably want to consider, mainly just because we knew it would be difficult time-wise to balance developing the game and giving it the proper marketing it would need.
We approached Team 17 fairly early on with the game but we didn’t actually sign a deal with them until a few months before the release (heads up to any indies: deals with publishers can take a lo-ong time to finalise 🙂 ).
The content for the game was more or less in and we were working on the console ports at the time but it worked out really well actually, Team 17 were able to offer a lot of great support, not just in terms of marketing but we got a lot of help testing the game too.
You’ve previously talked about enjoying fan-made co-op content in LittleBigPlanet. Could an Overcooked custom kitchen/mode creator be a potential release down the road?
It’s not something we’re currently planning, I can absolutely see the appeal, I myself really enjoy creating new levels for the game but it would take a lot of work to make this something approachable to outside users.
Finally, do you have any questions for me?
I’d love to know what your favourite co-op games are or what your favourite co-op experiences have been in games. Having spent the last two years thinking about nothing else it’s always interesting to try and dissect what it is people enjoy or hate about the experience of having to work together with their friends/family!
Oh boy, where to start?
I think the things I enjoy most in co-op games are having a common goal to strive for, like reaching three stars in every Overcooked kitchen. Some games that come to mind for me in this regard would be New Super Mario Bros. (the Wii and Wii U versions), and Mario Kart 8.
It’s important that the game have challenges beyond what sits at surface level. In New SMB this would be the star coins, Mario Kart has the illusive 40 point/3 star victory.
Next up would be the horde mode genre. Gears of War 2 and Iron Grip Warlord don’t see too much play anymore, but are excellent examples. Dungeon Defenders is a bit of a mix between wave, progression, and loot based gameplay, which we sunk dozens of hours into.
Above all these though is Call of Duty Black Ops Zombies. The many nights spent with system linked Xboxes and music blaring (a lot of Meat Loaf, Katy Perry, and Dragonforce) are among my favorite gaming memories. I think one of the driving forces of this game was how you had to stay on your toes. One slip up and it was back to the start. But we always knew we could do a bit better the next time.
I’ll briefly also touch on the joy of playing single player games cooperatively. Most recently Until Dawn and Life is Strange. I imagine the Telltale games would work well for collaborating on story decisions too.
Mario Golf, when not played competitively, has some good controller-passing sessions in an (often fruitless) attempt to unlock new courses and characters.
Really interesting reading about your favourite co-op experiences, definitely a lot of overlap with my own, loved playing Until Dawn as a group for example. I’ve always enjoyed playing games way more with other people which is almost certainly what lead to us developing Overcooked.
Shout out to a little indie game called Storage Inc. which was on Xbox 360 – a great little co-op game that was a big inspiration for Overcooked, if you can track it down I definitely recommend giving it a go with some friends, and the sequel is currently trying to get through Greenlight on Steam.
Thanks to Phil for taking the time for this interview. Overcooked is available now on PC via Steam, PlayStation 4 and Xbox One.
