Hero Tactics — Post-Mortem

For the month of October, I’ve been taking the the Devtober challenge: Work on your game every day, tweet about your game every day, and at the end of the month, write a post-mortem. After four harrowing weeks, I developed and released Hero Tactics, an oldschool tactics rpg!

This was an old old project, something I was working on back in college. It had the basics of strategy gameplay, including moving your units, being attacked by enemy ai units, and either winning or losing.

I totally just borrowed the sprites from Fire Emblem 7 as placeholder assets.

But the art was borrowed, and the gameplay had no progression. If I wanted to release this game, I’d need to iterate on the art and extend the gameplay to keep players wanting to play.

So how did I do? At the beginning of #devtober, I posted some goals for the month, around art, gameplay and social. In this post-mortem, I’ll review what went well and what could’ve gone better for each of these goals.

Replace all the Art

What Went Well

• Replaced all the art!! Woohoo!
• Worked with some awesome folks to produce the art! I’d’ve never been able to do this by myself. Vinícius Menézio drew the initial game sprites, establishing the retro aesthetic and muted color scheme. Leopold Langer mixed together the character portraits.
• Overhauled the ui! A lot of work went into improving the ui/ux. In my older prototypes, there was never enough conveyance about what the heck was happening. By adding features like on-unit health bars, play-by-play animations, and action arrows, the gameplay was finally grokable.

What Could’ve Gone Better

• Had a lot of trouble with the chunky pixel resolutions. The game sports very big pixels at a 96x128 resolution. It looks very retro, but it also introduced a lot of challenges. It is really hard to cram in any ui elements and text elements when there aren’t enough pixels to support it, and tactics games are almost entirely communicated via ui.

Experiment on Twitter

What Went Well

• Tweeted every day of the month!! Woohoo!
• If a picture tells a thousand words, then a video tells a million words. I learned it was a lot better to tweet video instead of images. You can express so much more gameplay in a video than an image. Also, images can get cropped, while videos are always shown at full resolution.
• No shame in bumping your content. I got comfortable with retweeting my own content. I would work late into the night, finally posting something just before midnight pacific time… for nobody to see. But there is no shame in retweeting your stuff the morning after. Also, by adding a tweet to a thread, you are effectively bumping the other tweets in that thread, so each retweet is also retweeting your entire thread.

What Could’ve Gone Better

• Not all work is shareable. It was really difficult to find something to post each and every day. Not all features are sexy, not all bugs are funny. Maybe I didn’t get as much work done, or maybe I had too much work to do that I couldn’t find time to make a post. I had a few days where I could really only post a funny gif of a monkey.
• Harder than I thought. It took a lot of time to make a solid post, maybe a full hour a day. It definitely wasn’t wasted effort, but it definitely was more work than I thought it’d be.

Publish a Game

What Went Well

• Published the game!! Woohoo!
• Finished a lot of those meta progression features. In my old prototype, there was no reason for a player to keep playing from level to level. But I addressed that by finally adding features like leveling up, learning new abilities, exploring an overworld map, and a very lightweight story.
• Got a lot of actionable feedback from playtest sessions. By widely sharing my game, I’ve gotten a lot of critical bug reports, quality-of-life ui requests, and backlog features. When I make my next tactics game, I’ll have a lot of tasks ready to triage.

What Could’ve Gone Better

• But what makes it different? I will be the first to admit the game isn’t anything revolutionary. The design is vanilla and generic, and isn’t doing anything adventurous or experimental. This definitely derisks the scope of the project, but it also makes it difficult to market the game. When I make my next tactics game, I want to challenge myself to do something different.
• The last 20% is more difficult than the preceding 80%. The last week of the project was really rough. During the playtests, I identified a lot of critical bugs that were launch blockers, that I had to rush to squash. I also had a bunch of buggy ai behaviors that I couldn’t reliably reproduce until I performed a deterministic refactor, which was stressful to do so late.
• Show me the money. At the start of the month, I had aspirations to make something I could sell. In the end, I decided not to sell the game, at least not yet.

What’s next?

I’m not quite done yet! I’ve learned a lot from developing and releasing Hero Tactics, so I’m going to build off that for my next tactics game. Maybe something more experimental and adventurous. For the Ludum Dare earlier in the month, I prototyped a monster collection game that I want to revisit.

If you’re interested in seeing what I make next, you should also follow me on Twitter, where I post all my development updates! Want to see more nitty-gritty details of my devtober adventure? You can review this megathread where I posted all my daily devtober updates!

And in the meantime, you can play Hero Tactics now on itchio!