Which communities are showing weekly growth? How are you ranking them?
We are using the z-score of the percent of growth relative to the size of the community. Think of a z-score as a “how unusual was this week?” meter for your percent growth.

Why use percent of size?
It makes comparisons fair. Adding 50 people to a 100-member server (+50%) is very different from adding 50 to a 100,000-member server (+0.05%).
Percent growth levels the playing field.

How are you scoring and ranking communities?
We are using the z-score of the percent of growth relative to the size of the community. Think of a z-score as a “how unusual was this week or month?” meter for your percent growth.

Why use percent of size?
It makes comparisons fair. Adding 50 people to a 100-member server (+50%) is very different from adding 50 to a 100,000-member server (+0.05%).
Percent growth levels the playing field.

How to calculate z-scores:

1. Your week's percent growth: Imagine your community membership grew 72% this week.
2. Typical week across everyone: We look at the average percent growth of all communities that same week (e.g. 1.6%).
3. Typical deviations: Community data bounces around week to week; we measure that usual bounce in steps aka standard deviations (e.g. 31.4%).
4. How many steps above/below average are you?
   • (Your % − Average %) ÷ Typical step = z-score
   • (72% − 1.6%) ÷ 31.4% = 2.24 z-score
   • What 2.24 means: your week was two "typical steps" above average—a standout growth week compared to other communities.

How to read z-scores:

• ~0 -> about average
• +1 -> stronger than usual
• +2 -> standout week
• +3 or more ->  rare accomplishment
• Negative -> slower than usual (or shrinking)