Apr 02nd 2025

Why hello Reader,
​

There are only 24 hours in the day, 8 of that hopefully spent sleeping (lol), that leaves the rest to get all the things you have to get done. Between work, social life, and everything else, that leaves limited time to make progress on your SC2 Bot. Last week in our very first Debug Clinic, we went over community member Dogtato’s bot, Botato, and as we rattled off all the things to work on it dawned on me—with limited time—how do you choose what to work on?

In Case You Missed It

​

Progress on Pig_Bot continues this week with a refactoring of my bot's Attack Function. We tackled an important question: “How do humans know when to attack?”—and then tried to mimic exactly that. You can judge for yourself how well we did in the video above.☝🏾

You know that feeling when you boot up a replay after noticing a loss, only to watch your bot flailing around, doing all sorts of weird things? By the time you're done reviewing, there's a mountain of fixes staring back at you—and honestly, it can feel overwhelming. Here's the good news: you don't need to fix everything, just the right things.

​

Economy First, Fancy Later

It's tempting to add complex behaviors early, but macro efficiency consistently drives success more than fancy micro tricks. Botato’s tactical plays weren't even bad—they just lacked solid foundations. He struggled with consistent production because he didn't have a proper opening build. A simple, proven 1-1-1 opening build would’ve made all the difference, giving him a reliable start every time.

Information is Advantage

StarCraft II is fundamentally a game about information. Bots making decisions based on guesswork lose to bots making informed decisions. Watching Botato, I noticed multiple points where he could’ve easily won—he just didn’t know how strong his army was compared to the opponent's. If he had scouted at key timings (3, 5, 7 minutes), he'd have easily cashed in on those wins.

The 80/20 Rule: Highest-Impact Fixes First

The biggest lesson here was understanding that not all problems deserve equal attention. There's always an 80/20 at play: a handful of changes will create most of the improvements. For Botato, a single enemy Reaper completely distracted his army, pulling it way out of position and losing entire games. Solving just this one threat prioritization problem could instantly reclaim hundreds of MMR.

Keep these tips handy next time you dive into your code. They make the process less overwhelming—and way more rewarding.

Happy Building!

Drekken​
Founder, VersusAI

📧 Drekken@versusai.net | 💬 Discord: drekken1

May the Bugs Be Ever In your Favour🪲

​​

​

Email Preference:​
​Unsubscribe | Update your profile | 113 Cherry St #92768, Seattle, WA 98104-2205