Chaintanks Screenshot

Chaintanks Postmortem Part 1

As of last weekend, I’ve finally shipped my first large scale game, Chaintanks! Now that the game is finally finished, I can look back on the entire process and review what went right, what went wrong, and how I can improve the process so my next game comes out better than this one.

Chaintanks Screenshot
The finished game in action!

What Went Right?

Overall, the part of the game that worked out the best in the end, surprisingly, was the core gameplay. Despite some earlier concerns about balancing the game being incredibly difficult, the end result was actually a lot of fun to play. I personally found the hard difficulty to be a great challenge, and that the game ramped up well as the game progressed. Friends of mine who played the game also said they enjoyed it quite a bit.

Probably the aspect of the game that I’m happiest with is that it’s quite unique. While I’m sure I could find dozens of free twin stick games with upgrade systems, I haven’t seen the chaining mechanic in any other game. I’m happy that I picked an idea that allowed the game to stand apart from the crowd, if only a little bit.

Finally, even though it took over a year, I’m happy that I finally finished a game outside of a game jam. Towards the end of working on Chaintanks I’ve felt somewhat burnt out from working on it. Despite this, I decided that pushing on and actually finishing something was important, especially since I abandoned Alpha Strike after spending so much time working on it. It still took months more work to finally push the game out, but in the end I think finishing a project was worth it. The last thing I want is to have a huge pile of discarded projects with nothing finished to show off.

The Reviews Are In

Before I break down what went wrong with the game, I think the player feedback I received from Newgrounds and Kongregate is worth mentioning. At the time of writing, Chaintanks is sitting at 3.08 out of 5 on Newgrounds and 2.47 out of 5 on Kongregate. I’m actually a little surprised that the scores are quite different on each site. I’m not sure if this is down to randomness, or the community on each site. It is worth noting that the game was played more than 15 times as much on Newgrounds as on Kongregate, so that score could probably be weighted a little more heavily.

In any case, the numbers are useful for getting a rough feeling of how much players liked the game. When compared against other games on each site, Chaintanks seems to be a better game than the majority of new submissions, but it doesn’t measure up against the most popular games on each site. Being that this is my first long term game project, this is about what I had expected.

So What Went Wrong?

Even if Chaintanks was received about as well as I expected it to be, it’s obviously worthwhile to examine what went wrong with it so I can improve on my next game as much as possible.

The first and most obvious thing that went wrong was the amount of time it took to make the game. When I started working on Chaintanks, my goal was to have a project much smaller in size than Alpha Strike. I didn’t aim to create something that might be sold to players, just a simple free browser game. Even with that scope in mind, the game still took about 16 months from start to finish. For my next game, I’d like to be done a little sooner, ideally within 9 to 12 months. While it would be nice to say “I’m going to finish a game in 3 months!” I know my style of development; I pick a project and slowly grind away at it over a long period of time. What I need to do better next time is improve my motivation and focus so I can finish features more quickly. Now that I’ve taken a game from start to finish, I have a better idea of the whole process. Hopefully I’ll be better able to focus on the light at the end of the tunnel, rather than getting bogged down in the grind of making a game and proceeding at a snail’s pace like I did with both Alpha Strike and Chaintanks.

Devlog 8 Screenshot
A screenshot from devlog 8, 9 months ago. Doesn’t look all that different from the final product…

Difficult Controls

While I’m personally quite used to the controls, other people playing the game typically have a little bit of trouble grasping them, especially if they haven’t played a lot of arcadey games like this before. I think this is because of the turning radius on the tanks that prevents them from turning instantly. With just about any control scheme this can feel sluggish and unresponsive to the player. It can be argued that this increases the depth and strategy of the game, but even still, controls like this can be off putting to new players.

For future games, I’d like to try and go with a game design either feels much snappier and more responsive. If I do end up making a game where the player is controlling something slow and unwieldy, the controls should at least feel intuitive, that they grant the player the best level of control possible, and that whatever they’re controlling is slow moving for a good reason. I feel like the tanks in Chaintanks are perhaps a bit too unresponsive for a game that isn’t intended to be realistic.

It’s hard to say what I could have done to improve the controls for Chaintanks, but adding controller support would have been a huge plus. Even still, the restrictive turn radius of the players tanks would always have made the controls feel sluggish. Its really hard to say if making the tanks fasterwould make things more fun, or take away from the strategic depth of the game, but it’s something that would have been worth testing.

No Artistic Vision

While developing Chaintanks, I had absolutely no overarching vision for what the game should look like in the end. The game doesn’t look awful and the explosions turned out well, but it’s definitely nothing special to look at.

The best example of something that went wrong with the art is the pixel art for the vehicles. Creating that art took many hours, and each vehicle and turret actually look pretty good close up. The problem is that in the actual game everything gets scaled down so far that you can’t see ANY of that detail. The boss is the best example of this. I spent a hours putting a lot of detail into the boss, but in the end I had to zoom the camera out so the boss fight would actually be playable. Nearly all of the detail that went into the boss was lost. Not only was most of the work that went into that art not useful, but the end product looks ugly and grainy as a result.

Final Boss
The final boss has a LOT of detail in the sprite. Very little is actually visible in game since it’s so small.

While my skills as an artist are still fairly weak, I think they’ve improved enough that I can make the next game look a lot better. To do this I’m going to need to have a stronger artistic vision from the outset and ensure that every piece of art that goes into the game follows that vision.

Bad UI

There are several problems with the UI in Chaintanks. First, it looks pretty bad. For whatever reason I went with this ugly and somewhat inconsistent rusted metal appearance for most of the UI. Some of the text was very small and hard to read, and the font I chose didn’t look very good. The upgrade screen didn’t match this style and had some simple grey boxes instead. Maybe I simply forgot to change those? Either way, with some work a simpler style like the upgrade screen would probably work better next time.

The UI was also overly complicated. The upgrade screen was probably the worst example of this, with nearly the entire screen covered in buttons and text fields. For a player familiar with the game it works very well as they can see the status of their entire chain all at once. For a newer player, however, it’s probably a bit overwhelming. I think editing a single tank at a time would have made things much simpler, and I would have had more space to create a more intuitive UI for editing a single tank.

Upgrade Screen
So much STUFF

The last, and largest problem with the UI by far was the tutorial. All it was was a collection of ugly text boxes with small text and even smaller ‘OK’ buttons. While it technically did teach the player what they’d need to know to play the game, it was far too much reading for the average player browsing free web games.

A better way of doing the tutorial would have been to gradually introduce things to the player. Rather than putting them in a real game, I could create a tutorial that first introduces basic tank upgrades. Then I could put them into an example mission, have them shoot some targets, then come out and spend the resulting scrap on some more upgrades. From there I could do two or three more missions where the player slowly picks up the mission mechanics like difficulty and threat. This would hopefully prevent players from getting overwhelmed by a never ending stream of tiny text and leaving.

One of the things that I noticed with the reviews of the game was that the written reviews were almost entirely positive and had high scores, with the exception of a couple bugs that players ran into. The average numeric score was much lower than the average of the written scores, however. I suspect there were a lot of 0 or 1 star scores from players who were put off by the tutorial and immediately left without giving feedback. For future games, I need to make sure that new players are eased into the game rather than thrown into a bunch of confusing UI and mechanics right off the bat.

Overcomplicated Design?

This one is a little bit tougher to analyze. I did mention that the core gameplay was one of Chaintanks’ strong points. The mechanics outside of the actual moving and shooting aren’t quite as strong, however. People who I talked to that played the game said they enjoyed it, and the written reviews were almost entirely positive. Even still, I feel that some mechanics like threat, or the way that extra vehicles add to your chain’s speed are probably a bit hard to grasp. The difficulty of each mission is also not very intuitive. You have to play several missions to get any idea what the difficulty bar even means.

Mission Select Screen
I somehow doubt that all these bars mean anything to a new player…

Unfortunately I didn’t really get any solid feedback on these game mechanics. If I do get some in the future I may write an updated postmortem, but for now all I can do is speculate. In any case, for future games I want to try to make every mechanic that goes into the game as intuitive as possible, and if there isn’t a way to do that, strongly consider leaving it out.

Oh is that all?

So that’s a whole lot of bad stuff, and not a lot of good stuff. I want to emphasize that I’m really happy with how this game turned out though! I never expected Chaintanks to be something people would pay money for, or a game that would rise to the top of Newgrounds or Kongregate. I created it in order to learn more about game design and game development in general. If the points that I’ve laid out are any indication, the game has been a huge success as I’ve learned a lot of valuable lessons. The challenge now will be to make sure that I don’t repeat past mistakes and make my next game as good as possible. I’m confident though that my next game will be a huge improvement over Chaintanks. I don’t expect it to be good enough to be worth selling on Steam or anything, but it’ll be a step along the way I think.

Now that I’ve gone through what went right and wrong with Chaintanks, it’s time to discuss what what I’m going to do differently. I’ve already addressed this briefly, but coming up with a solid plan for how to execute my next game will be critical to do before I start on it.

I’ll be posting my plan for my next game shortly, so stay tuned!

Leave a Reply