How the CRI Score Is Calculated

Last updated: June 2, 2026

The CAA Readiness Index (CRI) score is a 0–100 score that measures where your academic profile falls relative to the pool of accepted CAA applicants in our dataset. It is a percentile, not a probability. A CRI score of 75 means your composite GPA + test score profile ranks higher than 75% of accepted applicants we've seen, not that you have a 75% chance of acceptance.

1. The Formula

The CRI score calculator averages your performance across two equally-weighted dimensions: GPA and standardized test score.

1.1 Three-Component Formula (no Post-Bacc GPA)

gpa_z       = (cgpa_z + sgpa_z) / 2
composite_z = (gpa_z + test_z) / 2
CRI score         = empirical percentile of composite_z against the accepted pool

1.2 Four-Component Formula (with Post-Bacc GPA)

For applicants who submit a Post-Bacc GPA, we use a four-component variant that averages all three GPA signals together, then balances against test:

gpa_z       = (cgpa_z + sgpa_z + postbacc_z) / 3
composite_z = (gpa_z + test_z) / 2

Either way: GPA contributes 50%, test contributes 50%.

Each input is converted to a z-score (standard deviations from the mean) against the appropriate reference pool (cGPA against accepted cGPAs, MCAT against accepted MCAT-takers, GRE against accepted GRE-takers), and the composite is ranked empirically against every other applicant in our dataset who submitted comparable data.

2. Why 50 / 50

Two pieces of evidence drove this weighting.

2.1 AAMC Published Data on MD Admissions

The most comprehensive public dataset is AAMC Table A-23, which cross-tabulates GPA and MCAT brackets with acceptance rates for U.S. MD-granting schools across the 2023–2024 through 2025–2026 academic years (the latest aggregation in AAMC’s 2025 FACTS). The table shows that both GPA and MCAT matter, though the public MD data shows a steeper change across MCAT bands than GPA bands. Across all applicants, acceptance rates go from 0.6% to 79.9% across MCAT bands, compared with 2.3% to 63.7% across GPA bands. Because both GPA and test scores matter, and because neither our accepted-applicant data nor the public AAMC data supports GPA as the main filter, CRI uses a balanced 50/50 weighting between GPA and test performance.

2.2 Our Own Accepted-Applicant Data

We ran five inference techniques across our dataset of 374 accepted CAA applicants, using each applicant’s available GPA and test data (the GPA–test correlation work is necessarily limited to those who reported both):

Correlation analysis: weak positive correlation (r ≈ +0.07 to +0.17) between GPA and test scores among accepted applicants, meaning adcoms accept compensating profiles in both directions, not a strict diagonal.
Quadrant balance: “high-GPA / low-test” and “low-GPA / high-test” admits are roughly equally common, with no strong tolerance for one weakness over the other.
Compensation asymmetry: the redemption required in the other dimension is symmetric for sGPA pairings, with only a modest GPA lean for cGPA pairings.

The convergent signal across both datasets: adcoms appear to weight GPA and test scores roughly equally, not 2:1 in either direction.

3. What the CRI Score Is Not

Not an acceptance probability. We have no rejection data. A CRI score of 80 doesn't mean an 80% admit rate. It means your composite ranks above 80% of accepted applicants in our pool.
Not a school-specific score. Acceptance rates vary widely by program. A high CRI score doesn't guarantee admission anywhere, and a low CRI score doesn't preclude it.
Not a holistic evaluation. Adcoms weigh experience hours, recommendations, personal statements, interviews, and many other factors the CRI doesn't see.

4. When Your CRI Score May Be Misleading

A balanced composite score can hide a lopsided profile. If one of your inputs (cGPA, sGPA, or test score) falls in the lower range of accepted applicants while the others compensate, your CRI score may overstate your strength for that dimension. We flag this directly on your results page: when an individual metric falls below a z-score threshold among accepted applicants, you'll see a per-metric warning indicating that another part of your profile is boosting the composite. Treat those warnings seriously.

5. Data Caveats

The accepted-applicant pool is self-selected: these are accepted applicants who chose to share their stats
Sample sizes vary: cGPA = 374, sGPA = 318, MCAT = 112, GRE = 264, Post-Bacc = 61 (live from current dataset)
The Post-Bacc subset is small enough that we fall back to the three-component formula when fewer than 30 comparable applicants exist
All inferences about adcom weighting are based on outcomes we can observe (accepts) and public AAMC reference data (MD schools, not CAA-specific)

6. The Bottom Line

The CRI score is not a prediction. Use it to understand where your profile sits in a pool of accepted applicants, and to identify which metric (GPA or test) might benefit most from improvement before applying. Treat the absolute number as informative but not deterministic. The shape of your profile, and everything outside this score, matters more than any single percentile.

Questions or Feedback

If you have questions about the methodology or want to discuss the design choices in more depth, contact us at team@criscore.org.