Featured Blog

Designing for Local Hordes - Problems and Solutions for 9-Player Design in Runbow

Thomas McCall, Game Director and Level Designer for 13AM Games' Runbow writes about the problems the team has faced in designing a game that supports 9 players locally.

Good Day Readers!


Thomas here, Game Director and one of the Level Designers at 13AM Games. I thought I’d take some time to talk to you guys about some of the design challenges we’ve run into trying to make a local 9-player game; not only because we love sharing our experiences with interested parties, but also because if I stare at Runbow any longer I will start seeing a multi-coloured version of The Matrix. My goal is to simply explain the problems we encountered and their solutions while letting you take away what you will.

Due to the nature of my roles here at the studio, I sometimes find myself creating just as many issues as I help solve. This gives me a unique opportunity to discuss just how much needs to be taken into account when you plan for 9 simultaneous players and a mechanic heavily focused on colour.

Problem: Rubber-banding

For those who are unaware, rubber-banding is the design codename for “make sure people still have a chance even if they’re not doing so great.” One great example of this can be seen in a small, relatively unknown series called Mario Kart.


Players get a variety of power-ups in Mario Kart that allow them to combat each other during the race. If you are at the back of the pack, these items and power-ups have a very high chance to be the really powerful ones, like 3 red shells, star power, or the dreaded blue shell. Conversely, if you are winning the race, the power-ups and items you get tend to be lackluster, like a single banana or coin. By doing this, the designers are effectively handicapping those at the front while giving an advantage to those at the back, leading to tighter more exciting races. Rubber-banding at its finest.

We knew rubber-banding would be an extremely important part Runbow if we wanted it to remain fun and exciting for all 9 players. Taking inspiration from Mario Kart, we set up a system that would allow us to tier out our power-ups in order to equalize the playing field. Because we are operating in a 2D space, and because our power-ups tend to have global effects, we had to use carefully spaced out single power-ups instead of intermittent rows of power-ups like Mario Kart. We had to think of a different way to ensure that power-ups picked up by the leader do not just simply extend his or her lead.


The solution was to make double-edged power-ups, or power-downs that appear more frequently if you are in the lead. An example of this is the Swapper power-up. It works by taking the person who got the power-up and switching them with any random player still active. Therefore, if you are in the back, you have the potential to jump ahead several positions. Conversely if you are in the front you are going to be sent back at best one position, if not all the way to the back.

Now, if we didn’t offer any incentive to the person in first to get the power-up box they would always ignore it. Not the most fun. In order to stop this from happening we allow them to get certain power-ups that may not directly extend their lead, but still give them some sort of advantage. An example of this is seen with the Super Punch. If you get the Super Punch in first there are two choices you can make: continue on your merry way and hope you can maintain your lead with the threat of whomping anyone who tries to take it from you, or become a merciless hunter and turn around to put those behind you out for good.

Ultimately, we want players in the lead to have a decision to make every time they come across a power-up. Do they risk taking it, to further push their lead? Or do they pass it by and potentially give their adversaries a chance at an epic comeback? Risk and reward.

Again, the aim of rubber-banding is to even the playing field to a degree, and we believe that making the winning player hesitant to grab that shiny power-up box may just give those behind him the chance to snatch victory from the jaws of defeat.

Problem: Level Flow and the Balance of Chaos

Another major issue that came with the addition of more and more players was how to keep the levels flowing nicely without creating unintentional areas of complete mayhem.


The biggest culprit for this actually came from the starting grid. Again, because we are restricted to 2D, having all players with an equal start requires us to stack them in their own stall. At the start of each round, players would rush out of their stalls and a chaotic spam fest of punches would begin. In order to address this issue we started thinking of different ways to start players, either by staggering or different grid configurations. Unfortunately, none of these really worked for us.


Eventually we came up with the idea of dividing players at the beginning of the match into a 4-5 split by using terrain. By doing so, we not only effectively cut the mayhem in half, but also introduced the idea of having multiple lanes with varying difficulties.


Seeing how well the starting division worked, we began experimenting more with multiple paths and found that players enjoyed being able to select-a-quest and avoid each other through level navigation. As level designers, it also gave us the opportunity to give more risk-reward choices to players by varying difficulty levels or placing power-ups in hard to reach paths. Naturally, because the colour mechanic typically causes players to fall, we designed to make the uppermost paths the quicker and more rewarding but also dangerous and hard to stay on, while making the lower paths easier but slower.

Two lanes split where a log carries slow players to the bottom.


The pathing, in combination with the starting grid split, gave us another opportunity to add a subtle rubber-banding method: By simply making the winning player spawn in the lowest stall we can handicap them by forcing them to climb their way back to the more rewarding paths that others can access more easily. We noticed pretty quickly that this made races tighter.

Overall, the idea of branching out paths and keeping players separate had a really positive effect on making the game more about out-platforming and out-racing, and less about how many times you got punched at the beginning of the round. That being said, we haven’t done this to all levels, so those of you who somehow thrive in chaos will still get their chance to shine.


So, if you’ve stayed with me this far I hope that I’ve been able to shed some light on how we’ve gotten around some of the bigger design issues we’ve encountered while making Runbow. I’ve also been told to inform you that our other Game Director, Alex Rushdy, will be writing soon about our game’s combat mechanics and the overhaul they’ve recently received.

Cheers from the jungle!


PS, let us know how your multiplayer designs are going in the comments!


Latest Jobs


Hybrid, Cambridge, MA or Chicago, IL
Quality Assurance Lead

Bladework games

Remote (United States)
Senior Gameplay Engineer

High Fidelity, Inc.

Game Interaction Designer

Fred Rogers Productions

Hybrid (424 South 27th Street, Pittsburgh, PA, USA
Producer - Games & Websites
More Jobs   


Explore the
Advertise with
Follow us

Game Developer Job Board

Game Developer


Explore the

Game Developer Job Board

Browse open positions across the game industry or recruit new talent for your studio

Advertise with

Game Developer

Engage game professionals and drive sales using an array of Game Developer media solutions to meet your objectives.

Learn More
Follow us


Follow us @gamedevdotcom to stay up-to-date with the latest news & insider information about events & more