machine learning

Dispatches

they think it’s just a game bot
ByDrekken June 11, 2026June 11, 2026

Why hello Reader, “No one takes bot making seriously.” Was the response I got on Discord from a member telling me about his struggles applying a new role, after I asked why his SC2 bot never makes it into his LinkedIn posts. I sat on that for a while. Then, with the World Cup coming up, Boston Dynamics drops this awesome breakdown of their Atlas robot playing football. They ask a singular question: “Can football teach a robot to move?” They’re basically asking if a game can actually train a…

Read More they think it’s just a game bot
Guides

How to train your bot’s micro from scratch
ByDrekken June 6, 2026June 26, 2026

Why Hello Reader, Training micro with reinforcement learning sounds like something only Alphastar could do. But what if you could start with just one unit and two actions? SharkNice showed how to easily train unit micro from scratch using Sharky RL Matrix Training (or Stable Baseline 3 for python) , starting with a single Stalker kiting two Roaches and scaling up from there. But the key take away though : scope your action space to preset functions your bot already knows. Instead of giving RL…

Read More How to train your bot’s micro from scratch
Dispatches

Three ways to win before you fight
ByDrekken March 25, 2026August 21, 2026

your bot’s about to fight. here’s how to know if it should

Read More Three ways to win before you fight
Dispatches

My solution to Zerg rushes wasn’t more code. It was data
ByDrekken March 18, 2026August 21, 2026

Took 150 games and one scikit-learn import to fix what weeks of rules couldn’t

Read More My solution to Zerg rushes wasn’t more code. It was data
Dispatches

You don’t need AlphaStar money to use ML in SC2
ByDrekken November 26, 2025May 5, 2026

Machine Learning with SC2 bots is actually doable, if you do it right.

Read More You don’t need AlphaStar money to use ML in SC2
Dispatches

Grok is just terrible at Chess
ByDrekken September 17, 2025May 5, 2026

Flashy plays look good, but ladders are won on fewer mistakes. o3 vs grok showed why

Read More Grok is just terrible at Chess
Dispatches

Bots don’t lie – unless you teach them regret
ByDrekken May 28, 2025August 21, 2026

K-2SO crushes in combat but folds at the table. Why? AI learns to bluff through regret—and your bot can too

Read More Bots don’t lie – unless you teach them regret
Dispatches

Pinch of Machine Learning For Your Bot
ByDrekken April 16, 2025May 5, 2026

During last year’s Roundtable, Timo mentioned that the key to making Reinforcement Learning work in SC2 was breaking the problem into smaller chunks. That idea stuck with me. And now that we’ve kicked off a new Terran bot for the MicroLadder this week, it feels like the right time to finally give RL a shot.

Read More Pinch of Machine Learning For Your Bot
Dispatches

ChatGPT is playing StarCraft like a strategist
ByDrekken March 5, 2025May 5, 2026

a language env called TextStarCraft2 allows LLMs to play the game with interesting results

Read More ChatGPT is playing StarCraft like a strategist