Methodology
Nothing on this site is a proprietary algorithm. Every number is either taken straight from the source data or calculated with the arithmetic set out below, and every fit score on the site opens up to show its own inputs.
Where the data comes from
407 accredited US Family Medicine programs, compiled from four sources and cross-checked against each other.
- Published program-level application statistics
- The applicant-facing statistics that programs report each ERAS cycle. Almost every number on this site comes from here: ERAS applicant counts, interview rates, the composition of applicant and interview pools, resident composition, published Step 2 CK ranges and position counts. Where nothing is published for a program, this site shows "not reported" rather than filling the gap.
- Published national match results for 2026
- Match rates by specialty and state, used for match outcomes. In that reporting, and everywhere on this site, non-US IMG means a non-US citizen who graduated from an international medical school. US citizens who graduated from international medical schools are counted separately and are not included in any figure labelled non-US IMG.
- A national residency program directory
- Used to cross-check program details such as tracks, accreditation and program structure.
- The programs' own websites
- Consulted for every program that publishes one, and treated as correct wherever the sources disagree with each other.
Why the first three are not named. Each organisation has been asked whether it is happy to be credited here by name. Their names will be added as soon as that is confirmed. If you need to know exactly where a particular figure came from before then, ask and you will be told.
The figures describe application cycles that have already happened. They are not a forecast. This site is an independent project: it is not affiliated with, endorsed by, or published by any residency application, match, directory or accreditation organisation.
Corrections and removal
If you are a program director or coordinator and a figure about your program is wrong, or you would like it removed, get in touch and it will be handled. This site does not publish program contact details.
The fit score
Three components, each scored from 0 to 1, then weighted and added to give a number out of 100. It is calculated in your browser from the profile you enter. Nothing about you is sent anywhere.
| Component | Weight | How it is scored |
|---|---|---|
| Step 2 CK fit | 40% | Inside the program's published range, or above it, scores full marks. Up to 10 points below the bottom of the range scores half. More than 10 below scores zero. |
| Non-US IMG interview rate | 40% | The share of the program's interviews that went to non-US IMGs, scored against a ceiling of 40%. If you told the site you are a US graduate or a US citizen at an international school, the program's overall interview rate is used instead, because the non-US IMG figure would not describe you. |
| Signal lift | 20% | How much more likely an interview is if you signal, scored across a range of 1x to 4x. Counted only if you said you would signal. |
Weights rescale when data is missing
If a program does not report the numbers a component needs, that component is dropped and the remaining weights are scaled back up to 100%. A program with no signal data is scored on Step 2 fit and interview rate alone, at 50% each. It is never given a zero for a number nobody published.
Ties
Two of the three components are capped, so programs bunch up at the top: with a typical profile, thirteen programs score exactly 100. Ordering those alphabetically would imply a ranking that does not exist, so programs on the same score are ordered by the raw figures behind the capped components, non-US IMG interview share first, then signal lift, then overall interview rate, then size.
Visa is a hard filter
Visa sponsorship carries no weight in the score, because a program that cannot sponsor you is not a worse option, it is not an option. If you need H-1B, the 325 programs that do not sponsor it are removed from your results. Only 82 of 407 Family Medicine programs sponsor H-1B; 400 sponsor J-1.
How the derived numbers are calculated
- Signal lift
- If r is the overall interview rate, a the share of applicants who signalled and i the share of interviewees who had signalled, then signallers were interviewed at r × i ÷ a and non-signallers at r × (1 − i) ÷ (1 − a). The lift is the first divided by the second. A signaller rate above 100% is arithmetically possible from rounded inputs, so it is capped and labelled where that happens.
- IMG representation index
- The non-US IMG share of interviews divided by the non-US IMG share of applicants. Above 1.0 means the program interviews IMGs at a higher rate than their presence in the applicant pool.
- Estimated interview invites
- ERAS applicants multiplied by the reported interview rate. Programs publish the rate, not the count, so this is an estimate and is labelled as one wherever it appears.
- Percentiles
- Calculated only across programs that report the measure, using the midpoint method so ties share a rank. A program that does not report a measure has no percentile for it and is excluded from the denominator, rather than being ranked last.
The "limited data" badge
A percentage calculated from a small number is unstable. 51 of 407 programs carry a limited data badge because they had fewer than 150 ERAS applicants, or fewer than 15 estimated interview invites. At that size a single interview changing hands moves the reported share by several points. You can hide these programs from your results with a filter on the browse page.
What is deliberately not here
- Program coordinator email addresses. They are in the source data and are deliberately left out, because publishing them would get them scraped.
- Reviews, ratings or rankings of program quality. This site reports application statistics, nothing more.
- A geographic component in the fit score. The source data has no usable geographic figures, and in-state advantage cannot apply to a graduate of a non-US medical school. It will be added when there is real data behind it.
Data generated 2026-08-22. This site is an independent project and is not affiliated with any residency application, match or accreditation organisation. Nothing here is advice about your application.