The typical 5th grader at Alpha School has math achievement at the 85th percentile of 5th graders. This is not bad, but it's not remarkable.

I love the approach Alpha School is taking. And I believe that they're really trying to iterate to something that works really well. But I think many people are misled by the way Alpha School words claims about their achievement.

Alpha School's web site has this bullet point:

"99th Percentile: The majority of students consistently outperform national averages."

If you just glance at this, you may assume it means the majority of students perform at the 99th percentile.

But that's not what it's saying.

Alpha School's mean achivement score (across all students in a particular grade) puts the 'district' (collection of schools) at the 99th percentile of districts.

But that's not an amazing feat, because there are 10,000+ school districts in the USA. Most of those don't have the positive selection bias Alpha School has (due to the price and ideology). Moreover, most districts have adverse selection, as many academically-inclined parents will choose to send their children to private schools.

You can judge the results for yourself. Here is the school score report from Spring 2025: https://go.alpha.school/hubfs/MAP%20Results%20-%2024%2025/20...

Here is some of the data from the Winter 2026 school score report: https://x.com/jliemandt/status/2023011075029922131?s=20

This leads to the takeaway at the start of this comment: the typical 5th grader at Alpha School has math achievement at the 85th percentile of 5th graders.

I can't add images here, so I'll link to the evidence here: https://x.com/RahimNathwani/status/2023111922636476899

Students Are Being Treated Like Guinea Pigs: Inside an AI-Powered Private School