This is a course in game theory, and the game is how it is taught. Each of the fourteen lessons takes one idea (what a resource really is, what information is worth, when a risk pays, how a bluff is priced) and shows what it changes in your games, with real numbers and something to practise. It follows Game Theory with Application to Starcraft, the class UC Berkeley taught for credit in spring 2009, run by players from the team that won the first Collegiate Starleague that same semester.
What you will be able to do
- Spend with a plan. Say what every bank of minerals is about to become, and how long that takes.
- Win fights you used to lose. Count shots to kill, focus fire, and pick fights where more of your army shoots.
- Move an army well. Use chokes, arcs and surrounds on purpose, and know when a flank is worth it.
- Macro for a reason. Build the units you can actually use, not the biggest number you can reach.
- Scout to decide. Rule out what your opponent cannot have, and act on what is left.
- Harass and defend. Know which harass pays, and what you should spend to stop it.
- Use your hands better. Treat your actions per minute as a budget and spend them where they return the most.
- Time your moves. Hit when your opponent is weakest, and see their timing coming.
- Bluff and read bluffs. Know what a fake costs, and when the real thing is cheaper.
- Improve on your own. Review a replay, plan a series, and find the one thing that decided a game.
The fourteen lessons
- Lesson 1 Everything you spend is a resource Everything you own is a resource. Learn to name what you spent and what you got.
- Lesson 2 Units, damage types and upgrade breakpoints What each unit is for, how damage types and sizes work, and how to count shots to kill.
- Lesson 3 Micro: win the fight you are already in Micro as a set of rules you can practise: focus fire, pull back the hurt unit, never waste a shot.
- Lesson 4 Positioning: fight where your whole army can shoot Why a bigger army wins by more than you think, and how chokes, arcs and surrounds decide fights.
- Lesson 5 Macro: when to expand, and what it pays back Macro in practice: workers, expansions, production, and the numbers that tell you it is working.
- Lesson 6 An army you can actually use Why a huge bank loses games, and how to build the army you can actually use.
- Lesson 7 Scouting: work out what they have without seeing it Scouting as ruling things out: what to look for, when, and what to do with the answer.
- Lesson 8 How to scout, and keep them guessing Scout well: what to look for and when, and how to keep them guessing about your own base.
- Lesson 9 Countering: answer what they have, not what you fear Overloading your opponent: make them answer two things at once.
- Lesson 10 APM: spend your clicks where they count Your actions per minute are a budget. Spend them where they return the most.
- Lesson 11 Deception: make them pay for what they think they saw Deception: what a bluff costs, when it pays, and how to read one.
- Lesson 12 Timing windows: when to hit and when to wait Timing windows: when to hit, when you are exposed, and how to see an attack coming.
- Lesson 13 Containment: keep them in their base Containment: keep your opponent in their base and turn map control into a win.
- Lesson 14 Plan a game, then review it Your final project: plan a series, review your own games, and show what you learned.
How to take it
-
One evening
Read lesson 1 and lesson 4, run one fight in the Battle Lab, then play a game. About an hour. You leave knowing how to spend and how to fight.
-
A weekend
All fourteen lessons. Watch each lecture on its short cut: the five come to about four and a half hours. Each written lesson reads in about fifteen minutes. Do every checkpoint before you move on.
-
The full course
One lesson a week for fourteen weeks. Do the homework, review two of your own replays every week, and finish with the final project in lesson 14. This is the way that makes you better.
The weekly habit
The class asked every student for the same thing each week, and it is still the best practice routine there is:
- Play your games.
- Pick two replays, one win and one loss.
- For each, write three lines: how the game went, one good move, one bad move.
- Name the single thing that decided it.
Keep the notes. After four weeks, read them back. The same mistake will be there three times, and that is what you practise next.
Homework
Four problem sets, with answers, on the things that decide fights: damage and shots to kill, where to put a medic, how healing changes a fight, and how many units fit in a surround.
Set 1: read a game
Read a replay and say what the unit mix means. Write up one branch of a tech tree. Learn the players worth watching.
Set 2: units and damage
Medics against sunken colonies, what a probe is good for, how many tank shots kill a hydralisk.
Set 3: micro
Damage breakpoints, medic positioning, leapfrogging, and why mutalisks want a one volley kill.
Set 4: the surround
How many units can hit one target, and what a bad estimate costs you in a real fight.
The tools
You do not have to take the numbers on trust. Each lab runs the real game rules in your browser.
Damage Lab
Pick an attacker and a target and see shots to kill, with armour and upgrades.
Battle Lab
Run a fight between two armies and see who wins, by how much, and why.
Flux Lab
See how many of your units can shoot through a choke, in an arc, or in a surround.
Economy Lab
Workers, bases and income, and how long a bank takes to become an army.
Scouting Lab
Given what you saw, what can your opponent still have?
APM budget
Where your actions go, and where they would return more.
Credits and sources
Built on Game Theory with Application to Starcraft, taught at UC Berkeley in spring 2009 by Alan Feng, Anthony Chen and Sherwin Mahbod. The teaching on this page is written in our own words. Not affiliated with UC Berkeley or anyone named here; if something should not be here, tell us.