Next layers I want to add: vision control, positioning, objective setup, the pressure that happens before the
fight.
Open to ideas on what "pressure" should include. What am I missing?
An idea inspired by the World Cup: a "match momentum" model for #MSI2026 LoL
The goal is measure who is generating pressure and taking initiative throughout the game
Still exploring how to capture every layer of pressure, but this is a starting point. I'll keep improving it
I ran unsupervised clustering on all 28 of #T1's 2026 losses. It found exactly 2 ways they lose:
1️⃣out-laned early — #Doran & #Faker get caved in
2️⃣even at 15, but lost the dragon/baron war anyway
They map perfectly onto how #BLG (lane) and #HLE (macro) actually beat T1.
What do the data📊 actually say about #T1?
I merged every T1 game from 2026 into a single model to answer one question before they face #G2 in this #MSI2026:
What actually decides a T1 game🏆?
So what do the teams that beat T1 have in common (BLG/GEN.G/HLE)?
• Pressure the solo lanes (Doran & Faker are both negative at 15 against top teams).
• Don't force fights into Peyz's lane (+550 GD@15).
• Give up Grubs if you have to.
• Keep T1 off Baron.