What to ban when you play duo: criteria that actually work

Banning well is one of the clearest advantages of playing duo: you can think together about which champions break your plan and remove them before the enemy picks them. But it isn't about banning the strongest champion, it's about banning what affects you the most.

The three ban questions

Before the game, ask yourselves three questions:

1. Which champion stops me doing my combo? A champion that always escapes, that immobilizes before you do or that kills you at level 2.
2. What is my weak point? If you're a fragile pair, ban what kills quickly. If you're a slow pair, ban what stops you reaching the late game.
3. What punishes my lane more than any other? The champions that cover exactly what you need.

Bans by type of pair

• Engage (fast entry and kill). Ban champions with shields, easy escapes or crowd control that slows you before you go in.
• Enchanter (protect the marksman). Ban very fast engages and high-damage champions that ignore your protection.
• Poke (wear-down). Ban champions that chase you without fear of taking damage or that heal what you take from them.
• Top + jungle or mid + jungle. Ban champions that react to ganks, with many ways to escape or counterattack.

When a ban is pointless

• If the enemy can pick another just as annoying one. Banning one champion doesn't help if three others do the same.
• If the ban is only about reputation. A champion being famous doesn't mean it affects you.
• If you force yourselves to ban something you already struggle with. Better to learn to play against it.

Agree on it before the game

The conversation is short: "I ban X, you ban Y" or "if Z shows up, we do this". Keep each decision as one sentence in your head. And remember: a ban doesn't win the game, it just removes one difficulty. The rest is done by your plan.

And the data

Bans based on win rate need a large sample of games and data for each champion against each pair. That's why, in our duo guides, for now we list bans by mechanic (what the enemy does that bothers you) and mark percentage data as pending until there's a large enough sample.

If you want examples of what to think about for each pair, see the guides for Samira + Leona and Lux + Caitlyn.