Concerned Joe: A Post Mortem

Concerned Joe has so far been, not only my greatest game, but simply the greatest project I’ve completed in my life. I’ve really learned a lot of valuable info throughout the development and release, and I thought it might be good to share our experience and perhaps inspire other developers and/or prove that, despite what you may hear on a lot of forums, making a flash game that would be financially successful isn’t as hard as it seems.

Actual Game Link

I’ll be trying to cover everything starting from the very idea down to development, bidding, release and aftermath. Why cover the boring early parts? Mostly because I think this project was different from the very beginning and it might be interesting to see how it evolved into what it is today. So without further ado:

The Concerning Idea

Concerned Joe was first conjured when my good buddy Xelu and I were working on what was then a ridiculously ambitious game called TimeLed. In essence, it was a simple top down avoider/shooter, but we wanted to include tons of different kinds of tiles, intelligent bots, dozens of different types of enemies and obstacles, lasers, and most importantly, a time manipulation mechanism. Y’know, nothing different from the usual developer dreams of “MAKE A GAME WITH EVERYTHING IN IT!!!11″ we had been working on it for close to 16 months. The final push however, was looking too daunting of a task and I lacked the motivation.

While I was struggling to get a stable engine done, Xelu had some free time in which he managed to think up of a new game. I loved the idea, and thought that if we develop this really quick game and release it within 2 weeks, it would motivate us to continue working on and finishing TimeLed, and if we managed to sell it, then that would just be a bonus.

Very first iteration of "Joe"

His idea for a game was simply “A puzzle platformer featuring a guy called Joe, who’s concerned about his health, because if Joe stops moving, he dies. Oh, and as long as he’s standing still, he’d be shouting obscenities and swearing at the top of his lungs because y’know, he doesn’t want to die and stuff”.

It sounded hilarious, but didn’t sound like it had much sponsor value because of the swearing we wanted to include, but we hoped we’d just get a sponsorship from Newgrounds or a similar site that didn’t care about mature content.

Now Xelu wasn’t very happy with abandoning TimeLed and starting a new project. I convinced him that I could get Concerned Joe completely done in 2 days of work (the engine, at least) and we could release within a week. He reluctantly agreed.

Early Screenshot

So, wanting to finish as fast as possible, I fired up Flash and began programming on the timeline. I managed to get everything working within 4 hours, and this was the result:

Link

And that was the first Concerned Joe prototype. Playing around with it for a bit, it seemed pretty fun, so we decided to continue. And here we learn our first lesson…

Lesson #1: Prototyping is crucial
I think this is the most important part in the development cycle. For any given idea, the only way to know if it works is to prototype it. If it’s fun, keep it, if it’s not, trash it and start over. A lot of developers underestimate this step and simply trust that the game will get better once the “enemies, powerups and graphics are in”. It took me years to realize that this is the point where the fate of the game is more or less determined, and until recently, I thought that this was purely what could make or break the game.

So with that, we had our engine, we had a good idea that was somewhat new, it seemed that all that was left was making the levels and polishing. We’d be done in no time!

The Actual Development

In my eyes, we were already done with the “development” phase. The game was already in its “final push” stage just after 4 hours from the beginning. And this is another interesting thing. Looking at Concerned Joe now, it seems to be a (relatively) huge game with 20 large and unique levels, tons of hidden goodies everywhere, 3 different minigames and a hidden bonus mini-game at the very end. The larger the project gets, the more time we tend to allot ourselves for the development. But from my experience, it’s best to always try to reach a stable prototype with all the game features as fast as possible. Otherwise, you’re risking never finishing the game or it simply turning out not so good, despite the effort poured into it.

But since we thought we were almost done, we pushed on quickly.

It didn’t take long for us to finish all 20 levels and give the game some polish, and we had what we thought was a very impressive build for a quick game made purely to take a break from a larger game.

Link

We had a pretty fun game, and we estimated it could probably get around $2k or $3k in sponsorship, and we dropped the “yelling obscenities” idea for that reason. At this point, Concerned Joe was by far much more fun than TimeLed. And comparing the development times (16 months to 3 weeks) we were much more confident in Concerned Joe’s success.

Xelu being bored

Now I was happy with what we had and was ready to move on, but Xelu saw much potential in the idea, and didn’t want to just let it go that easily. We did some user testing and we arrived at the fact that the game was fun, but not as fun as it could be. It didn’t seem to have much replay value either. The puzzles at the end were rather hard, and we wanted to reward the players who managed to complete it all. So we went back and decided to add more content. And *that*, I believe, was what made all the difference from it topping off at $3k to the price it sold for (I’ll mention it soon, hold your horses). And here we learn our second lesson…

Lesson #2: Content and replayability are just as important as a fun game mechanic
For a long while, I believed that the only factor involved in making a successful game was having a fun core game mechanic, and nothing else mattered. The only thing I focused on when making games was making sure the core was fun. Once that was taken care of, I deemed the game ready for release. However, Concerned Joe has proven that the fun factor is only half the battle. It can only serve to hook the player, you need to keep serving the player to keep them playing. And while you don’t gain much from long play time with flash games, this is very important if you at all care about your players enjoying your games.

Adding Content and Replayability

And so we added the ability to fly once you’ve finished the game. A small idea that took only a minute to implement, but it opened up the chance for many players to re-play the game with their new flying powers and try to reach areas that were previously inaccessible. I think the best kinds of gameplay ideas are those that take the player longer to play them than it takes the programmer to make them. It’s hard to come up with these, but if you can, always try to maximize play time/fun while minimizing effort. This isn’t implying we should all be lazy. This is just to remind you that effort can have little correlation to how good a feature is if you’re not careful.

In order to give players more of an incentive to replay the game, we added a lot of hidden secrets scattered around that could only be reached by flying. Some of those enabled you to reach even more locations while others served only vanity reasons. All in all, I believe these extra goodies were very important in giving the player a feeling of being rewarded for their time playing, and that aided in their satisfaction the end.

Another thing that I think really went right was expectation vs delivery. By that I mean, once the player reached the end and the credits rolled, players would assume the game was over, but then they’d be surprised that they’ve gained a new power and there’s still more to explore. Once they’ve finished exploring, they are again not expecting to see a further hidden area, only accessed by completing the game twice and getting all the secrets, which leads to another mini-game which could very well have been released as a small flash game on its own!

But I think the most important addition we made during that phase was the monitors and voice acting.

Monitor Example

Several “monitors” throughout the game containing witty jokes and taunting you kept the gameplay interesting and provided a laugh every now and then. The voice acting however, was I think one of the main reasons the game was a success. Provided by the very talented Johnny Utah of Newgrounds, it transformed the game from just a cool puzzle platformer to this game with a “soul”. Where the players had mixed feelings of hate towards the narrator and actually felt good when they were finally victorious.

Throughout all this period, we were doing user testing at every step. I believe this is another very important step that must not be neglected. We put our game up in front of many developers and test players on FGL, we get our friends and family to play test as well, but the ones that provided the most valuable feedback were the two ends of the spectrum: the most experienced dev’s, and the First Impression players on FGL.

First Impressions is a wonderful service provided by FGL where they provide you with random players who will play your game and give their thoughts. The aim isn’t to get a detailed analysis. On the contrary, FI’s are very useful for knowing what the average players in the wild will think of your game. If the FI player says the game sucks and quits at the tutorial level, then it probably means you need to change the tutorial. This is what caused us to completely revamp our tutorial level. I can’t stress enough how feedback is very important.

A final note on content is…

Lesson #3: Never underestimate your players
I was surprised to see the lengths the players would go through to find secrets in the game. We wanted to make a game with depth, so that there’ll always be something they haven’t discovered, so we sort of layered the secrets. Some were obtained simply by finishing the game, others by doing some exploring, some by doing ridiculous tasks such as waiting for 2 minutes next to a specific monitor in a game, only for the secret to flash for 2 seconds and then disappear. And yet, a lot of people found these. There were even secrets that had no hints whatsoever in the game, that we just hinted at in forums, like typing the konami code in game for example. 

The Music

Rather than just picking a few pre-made loops, or just getting a composer to whip up a few songs, we went ahead and got the excellent musician Jesse Valentine to create multiple short loops for various points in the game. We had a menu loop, a mini-game loop to play during the mini-games and 6 different in-game loops that play randomly. I believe this helped to keep the music non-repetitive as it changed every level.

Concerned Joe Soundtrack

We also attempted to sell the soundtrack, but it was rather overpriced and out of the 80k people that clicked “Buy soundtrack” only about 6 or so bought it.

The Bidding

So we had a game packed with content, we had done a lot of user testing, we had done everything we could. And we polished the crap out of it. There was still a lot of polishing to be done, but we didn’t want to wait any further. So we put up the game for bidding while we were still working on it.

I figured that most sponsors would barely play past 3 or 4 levels, and what we still had to do at that time was finish up a few of the secrets that are obtained in the second playthrough, and actually think of and make that final secret mini-game. Bidding usually takes a month or two, so I thought we’d multi-task by finishing up the game while it was up for bidding. We did label the game as “100% complete” though.

The game emerged with an editor’s rating of 8.5. I had just read a post-mortem of how another game rated 8.5 sold for $13k exclusive, so I was pretty optimistic. My goal was $10k for this game.

We received a bid on the first day for $1,500. (The $500 bid in the chart was a non-exclusive) This was a very good sign, and I was eager to see where the bidding would go. The bidding however, went a bit slower than I had anticipated. The graph below is a rough representation of the increase in bid values over a period of 2.5 months

Bids over 2.5 months

Now I didn’t mind the long bidding time, mostly because the bids were increasing and we still wanted to keep adding more things. But as we approached the end of the third month, I feared the bidding might top off at $4k. I started PMing and Emailing sponsors and negotiating. They all seemed to be waiting for the other sponsors to counter bid. There were two main sponsors who were bidding, who at some point were increasing with $50 increments. It was getting ridiculous. I was disappointing at the outcome of the bidding. $4k was more than my highest sponsorship, but it was definitely not what had in mind for Joe. I wasn’t ready to sell just yet.

Several large sponsors had contacted me, but had ceased communications as soon as I told them we wanted $10k or just ignored that part and placed a lower bid.

I thought about pressing Last Call. That was a big risk though. If the bidding didn’t go much higher, then we were basically doomed. I had had bad experiences with Last Call not even generating a single new bid before, but I decided to just take the risk and press the big red button.

And here we see the graph for the bidding over a period of 10 days:

Bidding over final 10 days

And as you can see, the bidding erupted as soon as we pressed Last Call! We were having an average of 2 bids per day with huge increments. The oscillations you see in the graph are due to some bids being exclusive and others being primary. But in terms of overall value, the bids were growing immensely. At one point, the bidding was at $10k exclusive, and the primary was at $6k. It then reached a highest of $14k exclusive and $10k primary.

I was very pleased with that $10k primary and wanted to accept it, but decided to wait even more.

After a few more days, we had a $14k exclusive, and two $12.5k primaries. We could have probably gotten the bidding up to $17k exclusive or more, but, Xelu was getting impatient and it didn’t really matter to me getting a few extra thousand as much as finally releasing soon (Because at this point the game had been in bidding for 3 months). So Xelu PMed Armorgames and telling them that “if they bid $13k now, we’d accept their bid immediately”.  And that was that, the game was sold.

Release

It was finally time to let our baby out into the world. We had worked our asses off, we’d maneuvered the treacherous waters of bidding and sponsors, and our game was ready…except for the final secret mini-game which we hadn’t even started yet. I had put it off thinking I’d just do it while taking care of the sponsor branding and whatnot, and that was *just* enough time to finish it…or barely, we had to cut out a few features which we thought we’d just do “as an update after the game is released”. So basically, the game wasn’t done when we decided to get feedback, nor was it done when we put it up for bidding, nor was it even done when we released it!

Xelu being even more bored

The reception to the game was incredible though! It was released on Armorgames first, and for the first 100k people that played, about 5k made it to the very end of the game, and 4k finished the game a second time, and 2k found all the secrets and got to the final minigame, which I thought would take weeks before someone would find!

The average playtime was also a whopping 50 minutes. Which I think is amazing for a flash game.

A lot of the players expressed how amazing they found the ending to be, and I think there’s another lesson to be learned here…

Lesson #4: The Ending is important
A lot of developers know that the percentage of players that will play all the way through to the end is probably relatively low, and so they focus more on pleasing the majority of the casual players and simply throwing a “Good job bro!” at the end to make up for the lack of reward. And I’ve always done this myself, however wrapping up the game and leaving the player satisfied may very well be just as important as the rest of the game. If the player bothered to play through to the end, you at least want to reward them for doing so. (This is only my opinion so, you may take this specific lesson with a grain of salt)

Not all of the responses were positive, in fact, players were mostly split into two factions. Those who thought the game was amazing and loved it, and the other who hated our guts. That was mostly due to the lag…

The Lag

That, I believe, was the one reason that stopped the game from truly achieving much more. The game lagged horribly on slower computers, and was almost unplayable. Which caused a rift in responses. The one that could play the game at full speed were dumbfounded as to why some people said the game was really slow paced and had no idea that the game lagged, while others thought the game sucked because it was running at 10 fps or less for them. This was something I should have fixed, and I did attempt to do so, but as I mentioned above, the game was programmed on the timeline and was heavily reliant on movieclips, so optimizing would have meant redoing the entire game system. Which brings me to…

Lesson #5: Movieclips are evil…seriously.
Avoid them whenever you can. Movieclips are the root of all lag (at least they have always been for me) and are simply horribly inefficient. You’d have more control over everything and your game will fun much smoother if you rely on bitmaps and blitting techniques instead. In fact, it will even force you to code more properly.

Concerned Joe level design making of from xelubest on Vimeo.

Aftermath

So far the game has about 2 million views. It’s not a lot, but as I said, I think the main reason behind that shortcoming is the lag. It has gotten quite a couple of sitelock offers, mainly from large sponsors including Yahoo (not yet launched at the point of writing this article) and even Atari! (the Atari sitelock was pushed back until they actually make a web portal however)

The players that did manage to play the game at full speed and enjoy it seemed to be huge fans of it though. So far we have over 600 likes on our facebook page, tons of youtube videos and reviews and a lot of players have expressed how they wouldn’t mind paying to purchase a Concerned Joe 2. That has sparked the idea that the concept of Joe is a lot more fitting for the download market as it targets hardcore players more, and I really think that could be a success if we a develop a downloadable version of Joe with much more content and on a larger scale.

Overall I’m more than happy with the outcome of everything and I’m very excited about the future of Concerned Joe on future platforms. I hope this also proves that you shouldn’t listen to whatever people say about how the flash market is declining or how about there’s not as much money in it as there used to be. Just because you’ve had previous big sales but can’t now doesn’t mean the market is in decline, it just means the bar is rising.

(Note: Concerned Joe went up for bidding during the end of July, when a lot of people were saying this was the worst time to put a game about for bidding and blah blah. As long as you’re patient and have a good game, you should hopefully find a sponsor)

Oh, and there’s also a hidden tile in the game that plays the Nyan Cat song if you click it that I don’t think anyone found so far…

~Omar Shehata


3 responses to “Concerned Joe: A Post Mortem

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

%d bloggers like this: