AW scoring is unfair to alliances with different numbers of members. The current matching sets a maximum number of players difference; last I read from the devs, this was capped at three. That helps, but the problem is still there, plus in addition, members of an alliance can leave after matchmaking, either before the war starts or partway through, making this problem worse.
It is actually a double whammy against the smaller alliance:
- A larger alliance gets more hits, (six per member, more members)
- The average team is worth 2000 / members when killed, so the kills by the larger alliance are also worth more.
My spreadsheet multiplied them out to show the maximum score possible for each side and then the difference between them. The farther apart in size the alliances are, the bigger the advantage is. (Yes, every member having six perfect kills is very unlikely, but the advantage is proportionally there with only one, two, or three kills on average)
I realized that this scoring problem is readily fixable. Make the average team worth 2000 / (# attacking members) instead of 2000 / (# defending members) Relative point differences between stronger and weaker teams in a given alliance need not change.
In the fair case where both alliances have the same number of members, the scoring is exactly the same, but there is no longer a structural scoring advantage to a larger alliance if they aren’t.
|A1||A2||A2-Kill||A1-Kill||Advantage||A2-Kill (fix)||A1-Kill (fix)||Advantage (fix)|
A few more supporting columns and the formulas are in the google sheet, or other member numbers can be tried out: https://docs.google.com/spreadsheets/d/1LpCfQh8gtpUSNPGZF-I4glF4rghDtUafM1NdM8Jw4_E/edit?usp=sharing