SAT distribution for top colleges--see how they hoover up the top scorers, leaving crumbs for the rest

[Anonymous]

The figures for the number of 1570+ scorers are low.

When the College Board last released data on the number of scorers at a given score in 2015, the SAT was scored out of 2400. Approximately 17,500 scored in the top 1% out of 1.7 million test takers. There are now around 2.4 million test takers.

A 1570 concordance on the 2400 scale is 2360-2370.

There were only 2,500 scorers who achieved a score of 2360 or higher in 2015.

Even with superscoring, there are probably fewer than 7,500 1570+ SAT scorers each year.

[Anonymous]

Anonymous wrote:The figures for the number of 1570+ scorers are low.

When the College Board last released data on the number of scorers at a given score in 2015, the SAT was scored out of 2400. Approximately 17,500 scored in the top 1% out of 1.7 million test takers. There are now around 2.4 million test takers.

A 1570 concordance on the 2400 scale is 2360-2370.

There were only 2,500 scorers who achieved a score of 2360 or higher in 2015.

Even with superscoring, there are probably fewer than 7,500 1570+ SAT scorers each year.

I just find this implausible because my kid has a single-sitting 1570 and really doesn’t seem like she’s all that unusual.

[Anonymous]

Anonymous wrote:

Anonymous wrote:The figures for the number of 1570+ scorers are low.

When the College Board last released data on the number of scorers at a given score in 2015, the SAT was scored out of 2400. Approximately 17,500 scored in the top 1% out of 1.7 million test takers. There are now around 2.4 million test takers.

A 1570 concordance on the 2400 scale is 2360-2370.

There were only 2,500 scorers who achieved a score of 2360 or higher in 2015.

Even with superscoring, there are probably fewer than 7,500 1570+ SAT scorers each year.

I just find this implausible because my kid has a single-sitting 1570 and really doesn’t seem like she’s all that unusual.

The College Board published the numbers. There were 9,203 test takers who scored 2300 or higher in 2015.

https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-composite-crit-reading-math-writing-2015.pdf

You have to use the wayback machine as the College Board now hides the 2015 score behind a firewall.


The concordance table published by the College Board indicates a 2300 old SAT is equivalent to a 1560 new (1600) SAT. https://research.collegeboard.org/reports/sat-suite/concordance/previous

https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-composite-crit-reading-math-writing-2015.pdf

You have to use the waybackmachine to access it.

[Anonymous]

According to the College Board, its annual summary for 2025 says that only 2,000,000 people took the test.

So a top 1% score means only 20,000 scored 1530 or above..

https://reports.collegeboard.org/media/pdf/2025-total-group-sat-suite-of-assessments-annual-report%20ADA-v0.2.pdf

[Anonymous]

Anonymous wrote:According to the College Board, its annual summary for 2025 says that only 2,000,000 people took the test.

So a top 1% score means only 20,000 scored 1530 or above..

https://reports.collegeboard.org/media/pdf/2025-total-group-sat-suite-of-assessments-annual-report%20ADA-v0.2.pdf

That is for single sitting. With kids taking multiple times and super scoring, scores above 1550 are now pretty common.

[Anonymous]

Anonymous wrote:According to the College Board, its annual summary for 2025 says that only 2,000,000 people took the test.

So a top 1% score means only 20,000 scored 1530 or above..

https://reports.collegeboard.org/media/pdf/2025-total-group-sat-suite-of-assessments-annual-report%20ADA-v0.2.pdf

You have to be careful. CH reports 2 sets of percentiles:

Nationally Representative (imputed for all students) and User Group (actual test takers)

https://research.collegeboard.org/reports/sat-suite/understanding-scores/sat

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Also each PSAT section maxes out at 760, not 800, so 224 for California out of a maximum of 228 is actually quite hard. Only three combinations are possible:

• Perfect reading + up to 40 points off in math
• 10 points off in reading + up to 20 points off in math
• 20 points off in reading + perfect math

That's it. At least California is not New Jersey, Massachusetts, or DC where the cutoff is 225 which means only these would make the cut:

• Perfect reading + up to 30 points off in math
• 10 points off in reading + up to 10 points off in math

Respect for those who made NMSF in these states 🫡

Is there a way to know how many kids from other states achieved 224 or higher even though the state cutoff was lower?

Not that I know of. Art Sawyer at Compass (https://www.compassprep.com/national-merit-semifinalist-cutoffs/) has this big table listing the number of NMSFs for each state (second column from the right) but that is based on the state cutoff, not on 224. But those numbers can be used as lower or upper bounds for states whose cutoffs differ from 224.

For example, at 224 cutoff, California has exactly 2172 and Maryland has exactly 348, achieving 224. At 225 cutoff, New Jersey has at least 511, Massachusetts has at least 282, and DC has at least 37, achieving 224.

As another example – for those who think the south is becoming the bastion of education with endless threads on DCUM – at 214 cutoff, Alabama has at most 228 achieving 224, with the actual number likely being MUCH lower since 214 is 10 index point below 224 and since Gaussian tail thins out quickly.

TBF, the South is popular for its colleges, and it's past 5 years of remedial elementary school reform, not for the brilliance of its tippy top current and recent high school students.

[Anonymous]

Anonymous wrote:

Anonymous wrote:using the # of high scorers(above 1530) as a % of total incoming freshman class,
The winners, Ranked by the highest % of 1530+ are:

MIT, 593/1106=53%
Caltech 140/263= 53%
Penn, 963/2421=40%
Hopkins 518/1418=36%
Harvard 580/1647=35%
Duke 535/1740 = 30.7%
Yale 475/1551=30.6%
Stanford 514/1693 = 30.3
Princeton 391/1411=27%
Northwestern 529/2100=25%
Brown 418/1768= 24%
Georgetown 369/1600=23%

All of the rest on the OP's list are under 20%, including Dartmouth at 17%, Cornell at 16%, UVA at 7%

Here’s the real ranking of peer group intellectual ability! Too many hooked kids at some of those ivies.

Are those comparing apples to apples? Were some back to test required?

[Anonymous]

Anonymous wrote:According to SFFA, a 1550 gave awhite applicant a 30% chance at Harvard. Slightly less if Asian.

Good odds considering overall acceptance rates.

Reminder that SFFA compared races within thst SAT range, not to slightly lower SAT.

It doesn't mean that 1550 is a big boost over 1500.

[Anonymous]

Anonymous wrote:

Anonymous wrote:According to SFFA, a 1550 gave awhite applicant a 30% chance at Harvard. Slightly less if Asian.

Good odds considering overall acceptance rates.

Reminder that SFFA compared races within thst SAT range, not to slightly lower SAT.

It doesn't mean that 1550 is a big boost over 1500.

White: 1500-1550: 10.2% admit rate
White: 1550+: 21% admit rate

Asian: 1500-1550 8.5% admit rate
Asian: 1550+: 18.5% admit rate

That is statistically significant. The data can be found in https://studentsforfairadmissions.org/wp-content/uploads/2018/06/Doc-415-1-Arcidiacono-Expert-Report.pdf

[Anonymous]

Anonymous wrote:

Anonymous wrote:According to SFFA, a 1550 gave awhite applicant a 30% chance at Harvard. Slightly less if Asian.

Good odds considering overall acceptance rates.

Reminder that SFFA compared races within thst SAT range, not to slightly lower SAT.

It doesn't mean that 1550 is a big boost over 1500.

It would be fair to state that a 1580 scorer has a tremendous advantage over a 1500.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:After today's SAT, you get the familar complaints of how the real SAT was harder than the practice test, this module was s hard, not enough time, etc.

I think the realization for many is that getting a 1550 is just hard and they just complain.

In California, only about 2000 students get to be NMSF. Now of course this is based on the PSAT and selection index, but as a point of reference a 1550 is harder to get than a selection index score of 224.

Berkeley and UCLA enroll almost 13,000 freshman. Safe to say the great majority couldn't score a 1550.

Perceptions get skewed because a magnet school that already self selected might have 50% scoring above 1500.

Still rare and in actual number, quite few.

Why do you say a 1550 is harder than a 224?

A selection index of 224 can be achieved with a "1530" SAT. In quotes because obviously they are different tests. However, the College Board allows for an alternate entry if one misses a PSAT because of illness, etc. A 1530 will qualify you.

In fact, even a 760M+740V 1500 PSAT or a 720M+760V 1480 is enough for a 224 index.

But 800M+730V is only 222 index, and this

A ceiling 1520 PSAT is a 228 index.

However, the SAT is also offered up to a year after PSAT, providing a score boost from that extra growth and study time.

That's why the PSAT is limited to 1520. It is expected to show 80pts of growth if you take the final SAT opportunity.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:According to SFFA, a 1550 gave awhite applicant a 30% chance at Harvard. Slightly less if Asian.

Good odds considering overall acceptance rates.

Reminder that SFFA compared races within thst SAT range, not to slightly lower SAT.

It doesn't mean that 1550 is a big boost over 1500.

It would be fair to state that a 1580 scorer has a tremendous advantage over a 1500.

If by "fair" you mean "mere speculation", sure.

It's certainly true that high achievers with stronger applications tend to have higher SAT scores, but that's different from the SAT score difference being the controlling variable.


You'd have to compare "stronger application + lower score" vs "weaker application + higher score".

[Anonymous]

Anonymous wrote:

Anonymous wrote:According to SFFA, a 1550 gave awhite applicant a 30% chance at Harvard. Slightly less if Asian.

Good odds considering overall acceptance rates.

Reminder that SFFA compared races within thst SAT range, not to slightly lower SAT.

It doesn't mean that 1550 is a big boost over 1500.

The higher SAT rate of admission was correlated with a higher SAT across all races. No matter the race of the applicant, the applicant had a statistically significant advantage in admissions the higher the SAT. The chances of admission if you scored out of the band of 1500-1550 doubled your odds of admission.

Admission rate for <1300, 1300-1400, 1400-1500; 1510-1550; 1560-1600
Asian: .5%, 1.1%, 3.2%, 8.5%, 18.7%
White: .8%, 1.5%, 4.1%, 10.2%, 21.4%
Black: 2.1%, 5%, 12.8%, 24.1%, 41.2%
Hispanic: 1.2%, 2.8%, 7.5%, 16.3%, 29.3%

[Anonymous]

The data at Dartmouth:
https://home.dartmouth.edu/sites/home/files/2024-02/sat-undergrad-admissions.pdf

Less advantaged students admission rate:
Bin
1450-1499: 5%-7%
1500-1549: 9%-11%
1550-1600: 15%-21%

More advantaged:
Bin
1450-1499:7.5%-8.5%
1500-1549: 11%-13%
1550-1600: 18%-24%