Teamfight Manager 2 opponent scouting: read the pre-match report and build a matchup plan

What the report actually tells you

The pre-match report surfaces a handful of fields. Read each one as a question, not a stat — every answer maps to a lever you control through the in-match tactical dial.

The percentages are descriptive, not promises — they're the opponent's historical tendencies, and the Auto Patch System plus their own roster changes move them. Read the shape (early vs late, snowball vs contest), not the exact decimal. The worked examples below are drawn from creator footage and illustrate the shape, not canonical balance numbers.

Read tempo first: early team vs late team

Before anything else, compare two numbers: their <10 min vs >10 min win rate against your own team's split. This single comparison decides whether you want a short game or a long one — and almost every other decision flows from it.

Worked reads (shape, not canon): an opponent at >10 min ≈ 18% is a near-pure early team with no late game — survive the early window and it's nearly free. The mirror, <10 min ≈ 100%, will win any sprint, so your only path is to drag the game long. A team weak at both ends is simply behind on roster — play standard, don't overthink it.

Tie this to the in-match dial: bias toward “defend” (late orientation — hold waves, concede trades, scale) or “trade” (early orientation — take the fight, swap objectives) to match the tempo read, not your mood.

Objective policy: contest vs concede Dragon

The biggest unforced error in TFM2 is contesting every Dragon on reflex. Whether you fight for an objective depends on what that objective does for this specific opponent — which the report's objective control rate and Dragon win rate tell you.

The first-Dragon buff compounds (a permanent, team-wide stat increase that stacks across the game), which is exactly why the snowball read matters: against a team that converts leads, one conceded Dragon can be the game; against a team that throws leads (low “win rate when ahead”), it often isn't.

Baron is different. It's a closing tool, not a tempo tool — once a team that can use it has it, group and end. Against a known thrower you can take Baron and methodically siege; against a team that closes, Baron buff means end the game now.

Attack the structural weakness

Most opponents have a single load-bearing assumption. The report's secondary fields expose it; your whole game plan can be “break that one thing.”

• “Team collapses when top is losing.” Their top is the structural column. Camp it: jungle pathing top-side, draft a lane bully, deny recall windows. When their top is behind, the rest follows.
• “Top roams to Dragon every spawn.” The empty lane is the exploit, not the Dragon — push it and take the cross-map trade.
• A noted weak lane (e.g. bot or support). Direct ganks and lane priority there. Sometimes the entire plan is “go there, repeatedly.”
• Low “win rate when ahead.” They throw. Don't panic when behind; play patient, set up one clean fight, and let them give it back.
• High control rate but low conversion. They take objectives without snowballing them — let them spend time on Dragons while you take structures.

The method: read the report, write one sentence — “We win by doing X to their weakness” — and let it drive both your draft and your tactical dial.

Turn the report into a ban list

This is the handoff to draft. The report's preferred-champion pool is your primary ban input, but order matters.

• Ban what they're best at, not the global best. Cut the champions their players have the highest individual fit on, not the top of your own tier list.
• In Fearless / series formats, ban their pocket picks first. A champion banned in game 1 stays available, but a champion picked stays used up — so ban the picks they're deep on across one or two players, the hardest to replace over five games.
• Don't reveal your own answer in scrims the day before. Going full-tryhard in a scouting scrim hands the opponent your report. Sandbag it.
• One ban should serve the structural-weakness plan. Camping their top? Ban the top laner's safest scaling or escape pick so the camp lands.

The full first-ban / last-ban / counter-pick taxonomy lives in the BP strategy guide; this section is only about which targets the report hands you.

Worked gameplans

Three end-to-end examples, each framed as “report shape → one sentence → dial + ban.” All shapes are drawn from creator footage; numbers illustrate the shape only.

A — The early-snowball team (high <10 min, high Dragon win rate)

Plan: “We survive to late.” Dial → defend. Refuse early fights; deny first Dragon if you can do it safely, otherwise give it and out-scale. Ban their early-tempo enabler.

B — The early-fader (very low >10 min win rate)

Plan: “We give them the early game and grow past them.” Dial → defend/scale. Concede Dragon, farm, hit your spike, and don't get baited into the fights they're statistically built to win.

C — The one-lane team (collapses when top is losing)

Plan: “We break their top.” Dial → trade, top-side. Jungle camps top, draft a lane bully, ban the top's safest pick. The report says the team folds once that lane is behind.

What doesn't go stale is the method — read tempo, set objective policy, attack the one weakness, then draft. The report changes every match and the Auto Patch System moves the meta under all of it.

• Contesting every Dragon on reflex instead of checking what it does for this opponent.
• Playing your default tempo regardless of whether they out-scale you or you out-scale them.
• Banning the globally strongest champion instead of the one their players are best on.
• Going full-tryhard in a scrim the day before an official match and handing them your comp.

FAQ