CWL roster planning: picking fifteen without starting a fight

Clan War League is seven wars in seven days with a roster you commit to up front. In a clan of thirty, that means fifteen people play and the rest watch — and the difficulty of CWL has never really been the attacks. It is telling the other fifteen why.

Publish the rule before the season, not after the picks

There is one decision that removes almost all of the friction, and it costs nothing: say how the roster is picked before anyone knows whether they made it. A rule announced in advance is policy. The same rule announced alongside the lineup is a justification, and everyone can tell the difference.

It does not have to be our rule. It has to be a rule. Something like: fifteen picked on war reliability over the last three months, with town hall used to break ties, and a rotation slot for whoever was benched last season.

The maths that should drive the picks

Seven rounds, one attack per member per round. On a fifteen-man roster that is 105 attacks available across the season, and every member who sits out a round removes seven of them.

• One member who misses two rounds out of seven costs you roughly 2% of the season's total offence — survivable.
• Three members at that rate costs 6%, which in a tight league is the promotion.
• One member who is present for all seven but averages a star less per attack costs about the same as a member who misses two rounds entirely.

That last line is the one people get wrong. Reliability and effectiveness cost you the same thing in the end, which is why the roster should be ranked on both — attack rate and average stars — rather than on the town hall number over someone's head.

The two ways a lineup goes wrong

Too top-heavy

Fifteen of your biggest accounts looks strong and produces a bracket you cannot three-star. CWL matchmaking reads the roster you submit, so stacking the top of it buys you harder opponents for seven days straight while your mid-table sits on the bench with nothing to hit.

Sentimental picks

The founder who has not attacked since March. The friend who is definitely going to be active this time. Everyone can see it, nobody says it, and it is the single fastest way to lose the clan's trust in the roster. If you make the pick anyway, say out loud that you are making it and why.

The CWL planner flags both: it warns when the fifteen skew top-heavy, and it marks anyone in the lineup whose recent record says they will not attack. It does not stop you — it just refuses to let you do it by accident.

A week-by-week plan that works

That last row does more work than the rest combined. A clan that publishes its CWL attendance at the end of the season gets a materially different roster conversation the following month, because the record is public and everyone has already seen it.

Locking the roster

Once the lineup is settled, lock it. Not for the tool's benefit — for yours. An unlocked lineup invites a week of renegotiation in DMs, and every swap you make after publication confirms to somebody that lobbying works.

In the manager, the locked lineup sits next to a season table that tracks each round as it plays out, so the plan and the outcome stay side by side. Post the read-only share link in clan chat and the bench can see the roster without an account, which removes the other half of the DMs.

What to do with the people you benched

• Give them the regular wars. A clan running CWL and normal wars in parallel has room for everyone, and it keeps the bench sharp.
• Tell them the one thing that would get them in — a rate, a hero level, a raid weekend. Vague encouragement changes nothing; a target does.
• Rotate deliberately, and say so when you do it. The rotation is the promise that makes the rule survive contact.