Build any matchup and draft by hand — pre-game Elo, comp profile, projected stat line, and the gold/time simulator that drives the live win probability. All client-side, no live game needed.
Draft Sandbox
build any matchup + drafts and watch the model react — no live game needed
Blue side
Red side
Gen.G 58%pre-game (Elo)42% Hanwha Life Esports
4× hard CC3× engage2× frontline
2× hard CC2× engage2× poke
Projected 15-min (team strength + the draft's lane edges)
Top
Red +416
Jungle
Blue +92
Mid
Red +159
Bot—
Support
Blue +40
Projected net gold@15: Red +90 · 4/5 lanes profiled
If that lead holds at 15 min → Blue 49% / Red 51%
Team strength sets the baseline lead (the better team is up gold at 15); the draft's lane edges add or subtract. The % is what that 15-min gold state historically converts to — a game-state read, and the draft still does not beat Elo at picking the winner.
First tower: Gen.G 55%·Hanwha Life Esports 45%
Draft profile: average length · Expected pace average (+1 kills vs avg) · champion-driven, directional
Projected — per team (expected · wide realized spread)
| Team | Kills | Towers | Drakes | Barons |
|---|---|---|---|---|
16.3 9.1–23.6 | 6.6 3.1–10.2 | 2.6 1.3–4.0 | 0.8 0.1–1.5 | |
15.1 7.9–22.3 | 5.3 1.8–8.9 | 2.3 0.9–3.6 | 0.6 0.0–1.3 | |
| Game total | 31.5 | 12.0 | 4.9 | 1.4 |
Length 33m (27–38)Kill pace 1.02 ckpm
From each team’s recent output, tilted to the matchup favorite — expected values with a wide realized spread, not a precise call. Objective counts scale with who’s favored.
Simulate a game state
Minute0:00
Gold leadeven
Kill leadeven
CS leadeven
Gen.G 58%live = pre-game (even state)42% Hanwha Life Esports