Alpha School — a franchise of private schools advertising their ability to teach students more than traditional schools in vastly less time — are cropping up all over the country. There’s one in my neighborhood, and there’s another across the bay in San Francisco that charges $75,000 a year. Austin has a bunch of them, including the Texas Sports Academy — whose pitch seems to be that you can get your education out of the way as fast as possible so you can spend more time on athletics — NextGen Academy for gamers, Founders School for entrepreneurs, and Montessorium, which blends the Alpha model with Montessori teachings.
The commonality among all Alpha schools is the company’s “2 Hour Learning” technology, called Timeback. The idea is that every day, each student gets assigned two hours of activities to complete through a variety of edtech apps. These can include passages with reading comprehension questions, math lessons and practice problems, and recording apps to help students practice public speaking.
The work is individualized to the student’s current learning level. They’re supposed to get through their work as quickly as possible to free up the rest of the day for programs and workshops. And the company claims that, as a result, the kids learn much more quickly than kids at normal schools.
One of Alpha’s innovations is that the kids themselves are highly successful marketers for the program — alumni write articles about Alpha’s merits, and I see children constantly in short-form videos, showing off their checklists of skills to master before they move on to the next grade.
Alpha is quite specific about just how much faster its students learn. If you go to the website of Alpha Anywhere, the company’s homeschool/remote learning platform, it will tell you it offers “mastery-based learning tailored to your child, built for 2x faster learning in 2 hours a day, from anywhere.” If you look for their gifted and talented program, that one advertises “3-7x faster learning.”
As an education reporter, one of the questions I’ve been asked the most often over the past year is, “Is that true?” If it were, the implications would be incredible. A student who masters twice as much material in a year in math, for example, could take calculus in sixth grade.
But the numbers don’t mean that. They actually don’t mean much at all.
The headline claims of “2x growth” come from numbers that have been massaged to the point of being useless for understanding what kids are learning. And insofar as there’s a secret sauce to Alpha School — and there’s definitely something — it is decidedly not the collection of edtech apps that they package as Timeback. I’ve tested those out and suspect whatever makes Alpha tick cannot be found in those apps.
The Alpha School franchise has been growing despite repeated scrutiny in the press. As I tried to understand their test scores and student outcomes, it became clear to me that a lot of parents understood perfectly well that Alpha’s marketing was overstating the company’s case — and they loved their kids’ Alpha schools all the same.
Alpha’s appeal seems to lie not so much in the “2x growth” on the webpage but instead in the particular location it has staked in the culture wars over the purpose of education — and the eternal appeal of selecting a rich, tech-savvy, ambitious peer group for your kids.
What 2x learning means
Alpha says that its students show “2x growth” if they meet certain criteria on an adaptive standardized test called MAP Growth. Alpha projects each student’s expected growth in the next testing period — for example, a third grader who is a 50th-percentile reader might be expected to grow 7 points from fall to winter. That number is a median; of all the third graders who are 50th percentile readers, half will grow more than 7 points, and half will grow less. If a student grows 14 points, Alpha calls this 2x growth.
The school offers students huge prizes, payouts, and incentives to perform well on those tests. That the resulting scores are higher than at other schools isn’t shocking and doesn’t say much.
But this can produce extremely odd results, particularly in high school, where term-to-term academic growth, as measured by the MAP tests, is far slower than in the early years; the expected gains for high schoolers on the MAP growth tests are close to nil as a result. For example, a ninth-grade student who is a 90th-percentile reader is expected to gain only 0.23 points on the reading MAP test from fall to spring. But this means that a ninth grader gaining two points, which is within normal random variation, would count as 9x growth!
Then, Alpha parent Peter Naimoli alleges, Alpha averages “across all students to announce a school-wide ‘2.6x Learning’ metric.” So one student growing at the expected pace and this student who gained two points would average 5x growth. Many of the reported numbers from Alpha actually have sample sizes nearly that small. The gifted and talented school, for example, was based on five kids with two test events, according to a 2025 article published in Astral Codex Ten.
If we’re admitting sample sizes that small (which we shouldn’t), I can claim that my own kids’ microschool, with four kids who showed up to both test dates and who put up some very impressive scores, also shows 3x learning. My best guess is that the gains from Alpha, or my kids’ microschool, over a conventional program are real if the program is a good fit1 — but Alpha students aren’t learning twice as much as normal students, let alone three to seven times as much.
Cold water in Arizona
One other obstacle to evaluating Alpha’s claims is the difficulty of disentangling them from selection effects: In general, the overwhelming majority of the differences between schools in student performance are driven by the self-selection of involved parents and motivated students. It stands to reason that a school serving the children of people who can afford to pay $75,000 a year is going to have better outcomes than any school that doesn’t, even if instructional quality is no better or worse.
Because Arizona’s Unbound Academy is a free public charter school based around the Alpha online platform, that particular school’s performance promises a window into the method’s effectiveness that is less distorted by selection bias.
The first year of results from Unbound, which came out earlier this month, are not encouraging on that front. Alpha had told the state of Arizona that it hoped to get 65% of the school’s students proficient in English language arts and 60% proficient in math. (The statewide averages are 42% and 34%, respectively.)
Instead, after one year, just 28% were proficient in English language arts and 10% in math.
Now, that doesn’t mean that Unbound failed to teach its students anything. The student body seems to have arrived at Alpha performing well below average to begin with. If you run Alpha’s usual MAP Growth analysis, the school can report 1.8x growth.
I don’t really find it plausible that these students in any meaningful sense “learned at 1.8x” in their year at Unbound. If Alpha really had taken a bunch of incredibly low-performing students and were helping them improve, then this would be a success story — but it seems far more likely that it took a bunch of weak students who almost entirely remained weak students. They improved their standardized MAP test scores a bit more than expected because Alpha teaches to the MAP test so directly. Overall this is obviously a failure on Alpha’s own terms.
It’s not just the mitigation of selection effects that makes Unbound Academy students’ experience instructive. Unbound is a “virtual charter school” — that is, it is an online program kids log into from home. That makes it the equivalent of Alpha’s online offerings, Alpha Anywhere and GT Anywhere, not of Alpha’s on-campus programs.
Alpha Anywhere is a bundle of edtech. The promise and limitations of edtech are pretty well understood at this point (and illustrated in a just-released paper looking at Khan Academy’s AI tutor): the effect of most edtech is almost nothing; it would work OK if used as intended consistently, but almost no one uses it as intended consistently. Most students will barely use it at all if they can help it, which means it is generally not anywhere near as good as high-quality instruction from an excellent teacher or even average instruction from an average teacher.
Alpha, insofar as it has gotten results, has done so by hosting kids in an environment where they actually reliably use the edtech. That is the true magic.
It’s no surprise that Unbound Academy doesn’t appear to be getting strong results in a purely remote environment. While reporting this piece, I tested out the publicly available apps that Alpha uses as part of Timeback, and I’m broadly unimpressed with most of them. Alpha has made the baffling and ill-advised decision to use AI-generated text for most of its reading-comprehension and content-knowledge instruction.
TeachTales, one of its reading apps, churns out Harry Potter fanfiction for kids in bite-sized chunks, gated behind comprehension questions. It might increase MAP scores, but I found it deeply depressing.
Alpha’s app for arithmetic is fine but uninspiring. Getting kids to drill their arithmetic is genuinely important for higher-level math and is often a chore, so I was hopeful to hear a rumor that Alpha had something good on that front. But it’s pretty much the same as other arithmetic practice apps.2
But the regular Alpha School (the kind parents pay up to $75,000 for) isn’t just the app. It’s a campus full of activities and games and prizes — some of them extraordinarily expensive prizes (like a Nintendo Switch or an iPad) — which the students earn by finishing their work.
It makes more sense to think of Alpha schools as the incentive system and environment that gets the kids to use the Timeback package properly — meaning, if there is an effective intervention, it is not Timeback itself. Just by getting kids to actually do the practice that the apps offer, Alpha seems to have made some real achievements, if not the miraculous numbers they claim.
(Alpha did not respond to my request for comment in time for publication. This piece will be updated if and when I hear back. The company has been made aware of at least some of these criticisms before: Naimoli says Alpha rejected alternative metrics he proposed, and Luke Shepard, who writes about AI learning, says he alerted the company in July 2025 to the apparent error in its white paper. Alpha has also declined to provide underlying test data to Wired and The Boston Globe. It continues to advertise 2.6x growth.)
What is school for?
To sum up, if you were imagining that “2x learning” meant something like students learning twice as much, that’s definitely not the case. And in a bunch of important ways, the students probably learn less — as far as I can determine, for instance, Alpha middle schoolers are not typically assigned any books.
But one of the most interesting things to me, as I’ve reported on Alpha School, is the program’s enduring appeal among parents who are well aware that the headline numbers are kind of meaningless.
“I love the school,” Naimoli wrote in a review that I would otherwise be inclined to call a scathing debunking of Alpha. “Alpha kids are increasing their MAP scores faster than kids at traditional schools, but the number is not 2.6 times faster. 2.6 is an erroneously computed and meaningless number.”
These sorts of testimonials put some of Alpha School’s negative press in a different light. Wired, for example, breathlessly reported on students who had a bad experience and left several years behind, dismayed by a school that set unreasonable expectations and deployed frustrating edtech. All of those things are bad. But they’re also all unfortunately and utterly normal.
It’s hard to imagine a Wired exposé of a specific local elementary school where only 20% of children learn to read. Alpha failing, in other words, is news, while the default system failing is not newsworthy at all.
This critical coverage has tended to make Alpha’s defenders double down: The system is out to get us, and everything it has to say about us could just as easily be said about any other school. But Alpha’s critics, presumably, would answer that your local public school doesn’t charge $75,000 a year, nor does it claim that it has cracked the code to revolutionizing education. And they might point out that marketing misleading growth figures is an affront to any school that is genuinely offering above-average pedagogical results, which will find itself unable to compete with Alpha’s massaged numbers.
I think the critics are right, but I think they could also do a lot to make their case more persuasive. Delivering a better criticism of Alpha School requires grappling with it on its own terms, understanding what it’s trying to do and why there’s so much excitement — and ideally articulating what the platform could do to live up to its hype.
Toward that end, one of my favorite Alpha critiques is from the education writer Michael Pershan, who wrote: “I just watched like three dozen promotional videos from Alpha School. I saw young kids fencing, knitting, rock climbing, biking, jogging, petting a horse, talking to Dr. Phil, golfing, painting, training with Navy SEALs, playing piano, playing guitar, doing pullups, squatting, giving TED talks, kayaking, lighting stuff on fire, baking, running in a parking lot, and recording a podcast. But I struggled to find footage of a student reading a book.”
Alpha, he pointed out, has a very specific viewpoint on the purpose of education: They are trying to minimize time spent on core instruction while maximizing performance. This is simply not something that most schools — especially not most fancy private schools that serve the kinds of kids who can afford Alpha — are trying to do. Most schools will tell you that they are trying to build holistic thinkers and citizens or foster a love of learning. Alpha is trying to get school out of the way as fast as possible.
What to make of that depends a lot on what you expect to do with the time you save with the apps. “Get academics out of the way as fast as possible so you can focus on sports” strikes me as kind of depressing, because I don’t love youth sports, but it’s a perfectly consistent goal for those who do. Plus, we all know kids who had exactly this attitude while in a conventional school, so why not let them have it?
Much of the appeal of homeschool, too, is that you can finish work in less time and leave more time for life — for wandering barefoot in the forest, for learning to build real things with your hands, for befriending stray cats. I once saw a pitch for an Alpha School that used the freed-up time to teach kids how to build industrial civilization from scratch in the wild. I have to admit, I’d sign my kids up for that one.
But there are models I’m less interested in: The flagship Alpha campus and the San Francisco Alpha, for instance, use their post-edtech time to promote “leadership, grit, financial literacy & entrepreneurship.” If you hate this vision, then I think the right response is to articulate a competing vision.
We do have to admit that one thing that got us here was our attachment to standardized tests. I am a standardized test defender: Ultimately, standardized tests are the best tool we have to hold schools accountable for results. That is why, for all the complaints about No Child Left Behind, it got more kids proficient in math nationwide. Small improvements in test scores add up to millions of lives changed.
But while I’ll defend standardized tests, I will also acknowledge Alpha is the end stage of the testing-first mindset. “What if the only thing we care about is MAP Growth scores?” The answer is, apparently, a bunch of AI-generated comprehension questions and apps.
You can’t describe why there’s anything wrong with this unless you have a much more idealistic and comprehensive picture of what education should be.
And if you do offer a much more idealistic and comprehensive picture of what education should be — something as ambitious as rebuilding civilization from scratch, say — then you are confronted with the fact that typical schools are not achieving that either. You don’t even need as ambitious a vision as that to be disappointed all around: I condemn Alpha for not assigning real books, but that’s hardly unheard of at public middle schools either.
I hate the idea of assigning kids to read mostly AI-generated prose — but I also expect it to soon become the default, for kids and for everyone, as the whole internet drowns in a tide of AI-generated prose.
We can and should do better than Alpha. But are we doing better than Alpha?
Recommended reading:
Education research is weak and sloppy. Why?
Across the board, the state of education research is incredibly grim.
Four things I changed my mind on after a year of polling
Biased public opinion polls have polluted the information environment. We wanted to do something different.
Naimoli lamented in a post last year, “I’ve proposed alternatives - better metrics that would actually mean something. They were rejected. Apparently “1.4 Sigma Learning” doesn’t look good on a slide deck.” He’s not the only one to make the case for 1.4: Matt Bateman, principal of Montessorium (an offshoot of Alpha), also told me he believes that “the actual statistically precise effect of Alpha on student performance is 1.4 sigma.” To be very clear, if Alpha is actually achieving 1.4 sigma learning, that would be a big deal!! I’d far rather see a carefully articulated case for 1.4 sigma gains than a sloppy case for “2x/3x learning.”
One notable exception to my general complaining about Alpha’s app selection is Math Academy, which I previously tried out for my oldest daughter. It is a third-party program that Alpha licenses. Math Academy is a very fast and effective program for learning the standard grade 4-12 math curriculum, but I really dislike how it teaches proofs and anything at all that’s open-ended or complicated. I think it would be pretty reasonable to use as half of a math curriculum, but I’d pair it with a lot of puzzles to teach the half of math skills it doesn’t focus on.
It feels like what a lot of people like about AlphaSchool is less that it’s proactively enacting some amazing set of outcomes, and more that it *lacks* a great deal of the self-imposed downsides of many other schools.
Alpha insider here. Going to confine my comments to the “2x learning is bogus” part of this critique.
As noted in footnote 1, my view is that there’s a more statistically precise way of saying this, a way that (a) virtually no one understands and (b) *is in fact very impressive*. A 1.4 to 1.5 sigma effect size, which is what I typically get when I run the numbers, is very large. So the worst-case scenario for 2x learning is that it’s just an incorrect description of what it is in fact an impressive effect.
But how imprecise is “2x learning” exactly? Putting aside the standard deviation for a moment (I’ll return to it), and using my 6yo as an example, the following things are true: she scored a 179 RIT on the fall 2025 math MAP. She scored a 222 RIT on the spring 2026 math MAP. The median student scoring a 179 on the fall test gains 15 points over the year. My daughter gained 43 points.
Moving beyond what’s indisputably true: I don’t think my daughter would have had 43 points of gains in a typical school, or even another good Montessori school. I think it would have been much closer to 15. (If you believe something like this, almost all the other discussion is irrelevant, and it answers the question as to why Alpha is beloved to many parents.)
Alpha markets that to me (in my capacity as a parent) as 2.9x learning. That seems... fine to me? It’s not what I would publish for data wonks, but it gives an impression that is correct: she outperformed, by a lot.
Does that mean that she learned 3 years worth of material? Not necessarily. I don’t think Alpha has ever said that. It’s presented to parents as a testing target. *Grade levels gains are reported separately and clearly parents.* (Incidentally, in fact, she did, almost. She went from mid-K material to 3rd grade material. But this doesn’t always line up with the RIT multiples.)
Returning to the shape of the data. So the median yearlong score improvement is 15 points for a 179 math MAP. That the median is 15 doesn’t tell you anything about how likely it is that a student scores a 43; if the curve is relatively flat, it could be about as common as any other result. We need to know the standard deviation to estimate the effect size (e.g. (gain - median_gain) / sd). I can basically look that up (inferring from some other data that MAP has published). It’s 8.5. (43 - 15) / 8.5 == 3.3. 3.3 is an insane effect.
There are all sorts of reasons one might want to tweak or dispute that, or use slightly different measures, but you’re going to end up in the same ballpark. It’s going to be a big effect, because it is one.
Does this happen every single time you compute the results of 2x (or 3x or...) learning? Not exactly in this way. The SD isn’t always the same across ages and subjects. But this *is* the basic shape of the effect. As far as I can tell, having run this exercise across many contexts: it never turns out that the range is such that a 2x effect is basically nothing to write home about.
What about outlier cases, where you get, like, 20x learning? Again, this happened to my 6yo. Her reading last year went from 201 to 222 fall to winter. MAP exepcted her to gain 1 point; she gained 21. Is that “21x learning”?
Well, I mean... what do we make of the fact that her score might have gone up by 1 in a typical school? She’s a gifted reader; gifted readers are typically not challenged in most Kindergartens, and the Alpha platform absolutely challenged her every single day. It is entirely plausible to me that if I enrolled her in either a decent public school or a decent Montessori elementary program (two of my benchmarks for Montessorium, the Alpha Montessori program she is enrolled in), she would not have been challenged.
To be clear, I wouldn’t put 21x learning on a billboard. But I don’t think it’s fair to say that the numbers are meaningless. (If you run the same math as I just did with math with derived SDs from public MAP data, you get a 2.3 sigma effect. Not incidentally, you are basically breaking the models at this point, which are not designed to be informative about this kind of gain, from 99 point something to a higher 99 point something.)
As far as probably overhyped and/or imprecise marketing claims go, Alpha’s seems minimally objectionable. And the minimum objection is not that it’s *misleading*, not that there’s some sort of false promise embedded in it, but that the effects are quite real but confusingly described for people who want to understand it precisely. And they are confusing because the actual statistical truth is objectively hard to communicate; our marketing department has wisely opted to not communicate in sigmas.