People Played My Game for the First Time… It’s Bad

A week ago I went to Full Indie Summit and listened to some great talks, but that’s not what I’m writing about. They tried something new this year where they set up some space for people to show off their games after the talks. I brought along my laptop with Chaintanks on it, but I decided that I’d only bring it out if there was a enough space for everyone else first. I figured it would be best to let people who are actually trying to make a living from their games show them off before I did. Once people had started filtering out and a few tables were freed up, I set up and showed Chaintanks to random strangers for the first time.

It was obvious from the first time someone played it that the game is actually ridiculously hard for someone who hasn’t played it before. Most people died within seconds of starting the game and got fed up after a few tries. I wanted to see people actually get somewhere in the game, so after the first two or three people were finished, I quadrupled the health of the player’s vehicles. One player also commented that the turrets turning slowly looked nice but felt crappy from a gameplay standpoint. I agreed that Chaintanks should have a more arcadey feel to it and changed them to always point at the cursor.

These changes helped people actually get through a few missions, but there were still more problems. In a typical game, you’ll want to start by upgrading your vehicle with your starting scrap, then pick a mission. The problem is that the UI doesn’t support this flow at all. You get to the upgrade screen from a button in the corner of the mission screen. Most players instinctively clicked on a mission and started the game with a tank with no upgrades. The mission system is also not very intuitive – I had to explain what the cards meant in detail, and even then about half of the players just picked the one in the middle anyway (the really hard one).

Mission Select Screenshot
Bet you can’t guess what all of these numbers mean!

The controls on the other hand were actually less of a problem than I thought they would be. I thought one player was having trouble with the default controls, so I switched it to the alternate controls for him. As it turned out, the default controls were what he wanted, but he was just getting killed by the AI so quickly that he didn’t have a chance to even acclimate to the game at all.

What Can I Take Away?

There are a few things I can learn from finally getting people to play the game. The first is that I need to get it in front of people sooner next time. The game that I brought out was balanced for me specifically. Since I made the thing, that meant it was ridiculously hard. While in theory someone could get good enough at it to beat it, it’s not all that likely. Gamers have so much choice today that they’ll only spend a few moments with a game before discarding it if it’s insanely hard from the very beginning.

In order to make my game worth playing for anyone other than myself, I’d need to do some significant balance changes. There’s a huge problem with that, however… The game’s difficulty curve is based entirely on an algorithm. This means that adjusting it to make a certain part of the game easier or harder without affecting the rest of the game is very difficult. There are different levels of enemies that appear as the game goes on, so I can tweak those, but even that isn’t quite enough since what predicting what exactly spawns in a mission is very difficult.

I’ve also found that the difficulty swings so heavily depending on which mission you select that it almost always feels either too easy or too hard. For the people playing the game last week, it was always the latter!

The Prognosis…

After having people play my game for the first time, I think I can conclude that Chaintanks is a bad game. The algorithm based balancing and enemy generation make the game design effectively unsalvagable. The only ways I could see making this game work would be to do away with all of the procedural generation and use only hand tuned levels, or make it multiplayer. There might be other ways to make it work, but most solutions would amount to almost as much work as creating an entirely new game.

Despite all this, my plan is to finish and release it. I’d considered putting it on Greenlight and selling it for $1-$2 if it made it through, but I’ve decided to stick to my original plan of releasing it for free. I expect that it will probably be received poorly, but I think I can still learn something and continue to become a better game dev by finishing it rather than letting it become another abandoned project.

What’s Next?

Lately I’ve been working on the title screen art, and I’m going to finish it so my UI doesn’t look terrible. After that, I’ll move the upgrade screen in the UI flow so it always appears before the player selects a mission. I’ll also add in a simple tutorial to explain how upgrades and mission cards work. I’ll need to add tooltips in too since I’m sure players will forget what the mission card stats mean. Changing some of the stats from numbers to bars would also be a lot easier for the player to understand.

Title screen art
The title screen art, still in progress

I’ve considered redoing the upgrade screen entirely to make it more intuitive, but at this point I think it would be in my best interest to simply finish the game as quickly as possible, so I’ll probably leave it as it is.

I’m also going to add in difficulty settings that will apply roughly the same nerf to the game I hacked in at Full Indie. The hardest difficulty will be the game as it is now, with two other settings that multiply the player’s health to make things easier.

After that, I’m going to find free sounds and music, or spend a little money to buy some if I can’t find anything for free. One of the talks I listened to at Full Indie was on how audio can be used as a game design tool, and my approach here completely wastes the potential of audio to enhance the game. However, I think the game is too flawed to be worth worrying too much about this point.

What About The Next Game?

Finishing the game will probably still take at least another couple of months, or less if I can manage it. Once it’s finished, I’ll spend a few months drawing, trying out composing chiptunes, and other activities to build skills while I plan out what my next game will be.

Gameplay screenshot
I’d like to have more tilesets, but shipping it is more important

I watched a talk by Jonathan Blow where he describes some of his process for designing games. While I’d never claim to have anywhere near the level of skill he does, I want to try approaching the problem of finding my next game idea the way he does. During this process of building skills, I want to spend some time actively thinking about new game ideas, but I’ll also see if I can add activities to my day where I can get to a relaxed state where ideas might just pop up. Waiting for the right inspiration to strike me might just be the way to find a great new idea.

The only caveat is that I’m still doing this in my spare time. I’ll probably have to discard a lot of ideas that would simply take too much time in favor of building a much more compact game that I can actually finish with what time I have.

Maybe nothing will come of it, but spending more time building skills certainly wont do me any harm, even if means taking a long time to start my next game.

In the meantime, I’m going to continue grinding out the last bits of work on Chaintanks and documenting my progress. If you’d like to have a look at the game as it was at the time of writing, you can get it here. (If you need it for Mac or Linux, leave a comment or ping me on Twitter or Facebook!)

Leave a Reply