Draft to Game Plan

Given a completed 5v5 draft, identify each composition's strengths, likely win conditions and preferred game plan before watching the match, then compare the prediction with what happened.

Due: 3 May 2027 · Weight: 25%

The brief#

Given a completed 5v5 draft, identify each composition’s strengths, likely win conditions and preferred game plan before watching the match, then compare the prediction with what happened.

You will receive a completed draft for both teams, with no further information about the match. Before watching a single minute of it, analyse the draft using Week 9’s vocabulary. Only once your analysis is complete do you watch the match and compare.

What you submit#

Part 1 — before watching the match.

1. Each composition’s strengths and constraints: what its tools are good at, and what they need in order to matter.
2. Each side’s likely win condition, stated as a kind of game state (Week 9), not a single instruction.
3. The important resources and map areas each side is likely to prioritise, and the fight structure (Week 10) each side would prefer.
4. Where the two compositions’ plans conflict — the point in the game where one side’s win condition is most directly threatened by the other’s.

Part 2 — after watching the match.

5. Compare your prediction with what actually happened. Where the match matched your analysis, say what confirmed it. Where it didn’t, explain why — a wave state, a jungle read, a tempo decision, an information gap, or an objective call that changed which plan was actually winning.

You are not graded on predicting the result. A well-reasoned pre-game analysis that the match goes on to contradict is not a failure of this assessment — reality diverging from a defensible prediction is exactly what Part 2 asks you to explain.

What the criteria reward#

Composition analysis rewards identifying what each draft’s tools are actually good at, over a generic strength list.

Win condition identification rewards naming the kind of game state each side needs, over a single instruction like “get ahead early.”

Specificity of game plan rewards a plan concrete enough to be wrong — naming resources, map areas and fight structures a side is likely to prioritise — over restating the win condition in different words.

Conflict identification rewards locating where the two plans actually interfere with each other, over listing both plans without connecting them.

Post-match comparison and explanation of gaps is weighted heaviest because it’s the hardest part: explaining why reality diverged from a defensible prediction, using concepts from Weeks 8–10 and earlier, is worth more than noting that it did. A prediction that turned out wrong for reasons the analysis correctly anticipated as possible scores well here.