Three years on: Pixonic on the evolution of War Robots

COO Igor Klyukin on the game's life so far

Three years on: Pixonic on the evolution of War Robots

It's often said in the world of free-to-play development that launching a game is the beginning, not the end.

These aren't boxed products released onto shop shelves, never to be worked on again. These are games-as-a-service that require constant operation and updating, often over a period of several years. has long been investigating the Making Of notable games soon after their launch, but what happens long after a game is released?

In an attempt to find out, this regular feature will talk to the developers behind maturing live games about their experience so far. You can read all previous entries here.

In this entry we speak to Igor Klyukin, COO at Pixonic, about the 2014 PvP third-person shooter War Robots. With War Robots now nearly three years old, how do you reflect on its performance - from launch to the mature title it is now?

Igor Klyukin: We've come a long way since the launch. Our initial goal was to launch the product as soon as possible, so as to develop it based on user feedback and raw data.

Looking back after all this time, I'm fully confident that we did the right thing.

The first versions of the game were very different from the ones the players know now. Before the release, we scrapped some mechanics and concepts that we considered essential.

Who knows how far we would've taken messed-up concepts if we didn't get players' feedback.
Igor Klyukin

For example, robot repairs initially played a large role in the gameplay - a robot destroyed in battle became temporarily unavailable while he was being brought back to life.

This could last several hours, and the passing of that time could end up costing the player several million silver coins/units - a sum much higher than the average earnings of multiple battles. As such, the number of battles you could have per day was capped.

We made "warranty repairs" available within this framework - you could buy a warranty for a particular robot, which would allow you to use it an unlimited amount of times before the warranty ran out.

Who knows how far we would've taken this messed-up concept if we didn't get players' feedback after our soft launch.

And that's only one of the most memorable examples. I don't even think anyone kept track of the exact number of changes and U-turns we took while developing the game.

How big is the team currently handling live ops on War Robots?

We have about 50 developers - 12 game designers (incl. UI/UX specialists), about 20 client-side and server-side developers, 12 members of the art team and 13 quality assurance engineers.

And of course there are about 30 people in the community, customer support and marketing teams, all of whom also make a huge effort to keep the gaming community informed about War Robots.

How important do you consider customer support and updates to be? What has been your approach to this?

War Robots is positioned as a game as a service, and we put user experience quality first.

Good customer service, community management and constant updates are all key aspects of this approach.

Many of our players have been with War Robots for several years now - we wouldn't have been able to achieve that if we ignored their concerns.

As developers, we have to analyse the project's metrics and compare them to the reviews and complaints of different user groups.

Once we have that information, we can create updates that will make the game more interesting for a larger amount of players.

How have significant updates been responded to by players?

Each update causes a lot of debate in the community. Development blogs, polls and discussions make up most of the content on our official pages - players need to see their input.

Each update causes a lot of debate in the community.
Igor Klyukin

The launch of the clan system, our first special event on Halloween and the modified matchmaking system were our “loudest releases”.

Clans were hardly a new thing, and players were having group chats and adding prefixes to their nicknames long before we introduced clan functions to the game. The update really made life easier for clan members, allowing them to concentrate on managing their clan rather than fighting the game's shortfalls.

We held our first big event two years after the launch. It turned out to be an experiment that drew a mixed response. We’ve made an effort to highlight the good sides of each special event since.

One of our latest updates was the introduction of a league system in place of the usual selection method (matchmaking). This is still being widely discussed, and our audience is split down the middle - some people strongly support the move, while some are highly critical.

The main thing is that we'll continue to refine all our features until the result satisfies most of the players and us as developers.

How have you gone about ensuring that War Robots maintains a sizeable and active player base long after its launch?

To maintain our current user base, we're constantly introducing new improvements and changes to the game, adding new content and thinking about new mechanics and features.

In implementing these changes, we're guided by data analysis, as well as our community of players, as I mentioned earlier.

The second important aspect is the attraction of new users, as making an interesting game is only half the job.

To tackle this, we formed a marketing department three years ago, which has progressed in strides and has grown to oversee hundreds of ad campaigns.

We're just as reliant on metrics and data when it comes to marketing, constantly trying out new promotional strategies offered by our partners.

For instance, we actively use all the innovations offered by Facebook and Google AdWords. Depending on the results, we filter out the methods that don't work and actively develop those that show good potential.

This approach allowed us to increase our marketing budget from hundreds of thousands of dollars per month to several million, all while staying cost-effective.

What are the specific challenges involved in running a game that's built around multiplayer?

The top three issues are connection quality, finding the correct robot and weapons balance, as well as troublemakers who ruin the experience by abusing the game's loopholes or by constantly logging out of matches.

We set a DAU record this year, exceeding 1.5 million DAUs and 70,000 peak concurrent users.
Igor Klyukin

Our server team can name a much longer list of issues, but I'm sure their answers would be enough for a few specialised articles.

Any KPIs such as downloads, DAU or retention you’re willing to share?

We set a DAU record this year, exceeding 1.5 million DAUs and 70,000 peak concurrent users.

50 million installs are also just around the corner.

What lessons have you learned/are you still learning from War Robots? Is there anything about the game that, in hindsight, you'd now handle differently?

Our entire company has gained - and continues to gain - invaluable experience.

I'm sure that many team members, including myself, "upgrade" ourselves with each solved problem and every released update.

We're learning how to correctly read the data and understand players' demands, learning to prioritise our tasks and goals, and learning how to manage a large project like War Robots.

What could we have done differently? We could've introduced more social (and viral) interaction much earlier - we're only just preparing to release viral mechanics.

We committed another error following our soft launch, choosing to focus on creating additional content and features for our "hardcore" gamers, who finished the game within the first month of its release.

For several months, we were hell-bent on adding new things, instead of focusing on improving the game as a whole, and its initial stages in particular (retention).

What's next for War Robots in 2017 and beyond?

I’ve said this before - we have massive potential.

We're releasing a new mode soon, our first step in experimenting with the gameplay. New robots and weapons are also almost ready.

We're also preparing to launch new functions that will allow players to organise duels and tournaments on their own. Some users are already doing this, and we’re responding to our audience’s demands.

Features Editor

Matt is really bad at playing games, but hopefully a little better at writing about them. He's Features Editor for, and has also written for lesser publications such as IGN, VICE, and Paste Magazine.