Flash Game Postmortem: Robo Riot, by Urbansquall

Panayoti from Urbansquall rejoins us for a postmortem analysis on the release of the engaging action puzzler, Robo Riot. He walks us through an analysis of what his learnings, successes and regrets in retrospect after releasing the game. Read on for more! – Ada

Robo Riot is different from the usual sort of project we do at Urbansquall in that it originated outside of the company. It was the brainchild of Marshall Powers, who spearheaded its design and programming. He approached me towards the end of 2008 and showed me an early prototype and asked me if I’d help him turn it into something we could get sponsored. Despite its very rough presentation, the gameplay was addictive so I had to say “Yes!” and thus Robo Riot was born. Well, sort of. The robots didn’t actually show up for a few months, but we’ll get to that later.

robot riot flash game - down with robots
Down with Robots

Robo Riot is a dual-mode arcade puzzler where the robot inhabitants of earth have decided to rise up against their human masters. In Crowd Control mode, the player is charged with electrocuting the assembled masses of these robot rioters. In Assassination Mode, the player is seeking to dismantle the rebel robots by eliminating their most dangerous leaders. Play the game over at Kongregate.

What Went Wrong

1. Elusive Theme

Most game genres have a few themes to which they naturally gravitate. Unfortunately, the puzzle genre offers very little in the way of theme suggestions, and we struggled to come up with something that we felt was compelling. We went through a couple ideas, including even doing graphical work for a gem themed version. Each time we felt we had a theme that we liked we’d change our minds a week later. Each theme we thought we had settled on failed to create an impact. We wanted the theme to be something that you remembered vividly about the game.

Tricksy robots!

After weeks of indecisiveness, Marshall, perhaps jokingly, suggested “robots” as the theme. Robots quickly led to robot genocide which quickly resulted in one of our “What Went Rights” on the project.

2. Hard Time Negotiating with Sponsors

I’ve got an established relationship with a few portals, so when it comes time to negotiate a deal, I generally don’t have a hard time finding a sponsorship at a price I think is fair. Robo Riot was the first game where I really struggled to put together a deal at a price I was happy with. One of my portal contacts threw out a low ball offer, the lowest I’ve ever seen at about 1/3rd of what I was expecting for the game. The other pointed out another Flash game using the same basic gameplay concept that had failed to achieve critical mass. In general, the overall impression I got from portals was “meh.”

This surprised me as I felt my relative objectivity, based on the fact that I hadn’t been responsible for the day-to-day programming, had created a more realistic expectation of value. I thought this would give me more of an edge in bargaining, but more importantly, I enjoyed the game and thought it was well done, so I was surprised when the portal reception was lackluster.

Sponsorships seem to be a little harder to secure these days, but I don’t know if it was this or that I had just overvalued the game. Either way, it didn’t go as smooth as I’d like. I was happy with the deal we ended up finally putting together with Kongregate, particularly because they’re so easy to work with, but I still wish it had gone smoother regardless.

3. Dubious Programming Choices

Marshall made two particular missteps when programming Robo Riot that would later come back to haunt him.

Assassinate their leaders!

The first was that his early prototype relied heavily on the sequencing ability of the popular Tweener library. As he later found out, this can cause some major headaches when you get to the final stages of the project and you’re adding features like “pause” or tutorial functionality which interrupt animations and their sequencing.

The second misstep that he experienced is when he moved some of the logic of the timer into the UI component that draws the timer. This was a time saving measure when he first began implementing the UI component, but as it became more complicated, he found that he was inadvertently leveling up the player or ending the game when he would need to modify the time.

The solution is to separate the gameplay logic and display logic to ensure that the engine is more capable of integrating the expected and unexpected changes that occur during any game’s development. Marshall learned this the hard way, and I suspect he will avoid doing this again on future projects.

4. Help Screens

We really struggled to make concise information screens that explained the gameplay mechanics without being too didactic. Feedback in this regard has generally been that we did well enough, but I know that we could have done better. We tried several different ways of presenting this information, but ultimately I think we were crunching at the end of the project, and we were too burnt out to make the changes necessary to turn this part of the project around. We ended up losing more players than we should have as a result of this oversight.

What Went Right

1. Successful Prototyping

The game benefited from Marshall’s dedication to iterative prototyping. He started with a solid mechanic and kept grinding it till it shined with polish. The early gameplay prototype was fun and the game got better and better with each iteration.

Feedback from my usual group of testers (friends, families and associates) was more positive than any other game in Urbansquall’s past. A common complaint was that people had to force themselves to stop playing the game in order to send us feedback. I know that I secretly dreaded getting updated builds because I would inevitably find myself continuing to play long after I had achieved my objectives for the build review.

Weak and Helpless Robots

Many rounds of prototyping definitely helped lay the groundwork for a fun, enjoyable game.

2. Memorable Theme

We stumbled around quite a bit before landing on the Robo Riot theme. Hitting on the theme of putting down rioting robots was the catalyst that brought together many components of the game that had not yet integrated in a meaningful way. It gave the gameplay the context it was sorely lacking.

The theme was also memorable. I still chuckle when I think about dozens of robots marching up to a human compound only to be killed en masse while they riot in the name of equal rights.


Sometimes you can finish a game and never really find the glue that holds the whole project together. Luckily we found the perfect theme for Robo Riot and it really helped create a cohesive package.

3. Teamwork

This was our first official time working with Marshall and he integrated with the rest of the Urbansquall team like he’d been working with us for years. Projects go a lot smoother when everyone respects and trusts their team and Robo Riot was a textbook case of this. It was a pleasure to work with Marshall, and I believe he had fun working with us as well as we are already working on our next project together.

Reception, Ratings And Reflection

It appears the trouble I had negotiating with portals was the canary in the coal mine. The game launched on Kongregate to little fan fare. It has been out for almost two months and has not yet broken 30,000 plays. Its rating hovers at a forgettable 3.43. Ratings make or break a game, and 3.43 wasn’t enough to give the game the momentum it needed to spread.

The ratings, I believe, were really hurt by the fact that we launched with two gameplay modes. I think both game modes are fun, but I frequently see “I enjoyed Crowd Control but hated Assassination Mode” or vice versa in the comments people leave about the game. I think this necessarily had a downward pressure on the overall rating of the game as we were being scored not on people’s favorite mode, but on the lowest common denominator between the two. I don’t think we’ll do a dual mode game again.

Wrapping It Up

Robo Riot was a fun project for the Urbansquall team and we really had a great time working with Marshall to make his project come to life. As projects go, Robo Riot was a pleasure to work on and we look forward to building more games with him in the future. Despite its lackluster reception, we’re proud to include Robo Riot in the Urbansquall portfolio as an example of an engaging action puzzler.

  • http://www.elite-games.net Badim

    man, this is cool game. your problem not in bad rating on kongregate, your problem that you stumble on it! when you should move on to another portals! localization and distribution – that is true keys to monetize your game. do not relay only on few portals, and dont give up after poor ratings.

  • http://www.stormvideogames.com SomeNick

    Feel free to upload this game to my portal. And to several other dozen portals. In other words, I second what Badim said.

    I know the feeling, sometimes one works in projects that one consider will really break through, and then they barely get a decent score / acceptance rating out there. I have also experienced the opposite, I have created quick games just for fun, thinking they would not really impress people, and then I have seen said games earning a nice Mochi revenue and spreading around. Games development can be an unpredictable thing. Good luck!

  • http://www.urbansquall.com Panayoti Haritatos

    In my own experience, if a game isn’t spreading it isn’t because of a lack of distribution or exposure, it’s because of a problem within the game itself, whether it be design or technical depends on the game.

    As far as I know, the only game that has really avoided the typical Flash distribution curve is Platformer Racing. And that’s sort of a unique scenario. :)

    Flash games are inherently viral in nature. I’d say developers are better off focusing their efforts on making better games that distribute themselves, rather than spending hours trying to get a few more hits for a game that missed the mark.

  • http://www.elite-games.net Badim

    my point is, that Kongregate score is not score that you should look into. score can be lower just because it not right genre for Kongregate, or style. but that not mean that game is bad – it just got different focus.

  • Pingback: Your unique articles and blogs may be read by tons of readers, writers and webmasters