Skip to content

Methodology

Evidence snapshot 2026-10-26-expanded-v2 · Model 3.0.0 · September 2026 purchasing power

These are deterministic illustrations, not forecasts, guarantees of employment, credit, admission, or investment returns. All paths start at age 18. The annual ledger ends at the end of age retirementAge - 1, the start of retirement. The source and model snapshot are included in shared state and CSV. Raw national wage observations remain in the evidence dataset in May 2025 dollars. Only those wage observations are rebased once by 336.940 / 321.465 to September 2026 purchasing power. Other source prices from different academic years are held at their listed nominal amounts as a mixed-academic-year approximation, not a coherent inflation-adjusted price series.

Earnings and score

For each national route, annual qualified pay starts at P25, reaches P50 in qualified year 5, the midpoint of P50 and P75 in year 10, and P75 in year 20, then stays flat. Values between knots are linearly interpolated. Physicians begin attending work at 90% of P50 rather than P25. These are assumptions based on cross-sectional wages, not observed experience profiles. The pay multiplier scales the assumed path, including training, employer cash, and the high-pay reference; independently entered public vested shares retain their entered amount. It is not observed pay.

Family medicine has three residency years at ages 26-28; anesthesiology has four at 26-29. Both start at $66,986, the preliminary first-year stipend anchor, and assume 2% real annual growth. Electrician apprentices earn 40%, 50%, 60%, 70%, and 80% of the rebased national median at ages 18-22, an illustrative progression rather than measured national trainee wages. The PA patient-care year at age 22 pays an assumed $35,000; PA school is at ages 23-24. Security analyst preparation includes two paid IT years at 22-23, modeled at $50,000 per year. Plumber apprenticeship at 18-22 uses an illustrative 40%-80% of the rebased occupation median. These are editable assumptions, not observed IT or apprentice wages. The preparation-pay override changes the first-year amount and scales subsequent years proportionally; the pay multiplier then applies. There is no concurrent qualified wage during any preparation stage.

Public tech assumes $250,000 salary plus $50,000 cash bonus, and frontier AI assumes $500,000 salary. Both are conditional on entry at age 30 for five years by default, then return to the national software path, which also applies before entry. The experience clock continues through the employer period. Public vested shares are an editable liquid, taxable annual assumption only during public-tech tenure. Private or unvested equity has zero assumed liquid value, which does not mean its observed value is zero. Employer examples are not national salary distributions, and the public-tech posting is historical and its direct page returned Job not found.

Typical pay is the average of modeled qualified years 5-10 even when retirement is earlier. High pay is the rebased national P90 or the conditional employer posting's high cash bound, including the public posting's 20% target bonus. It is not a national ceiling. Zero employer tenure uses the national fallback high-pay reference. Time to meaningful earnings is the first modeled age with at least $40,000 gross, including applicable liquid vested shares, searched through age 100 regardless of retirement. If no such year exists, age 101 is a sentinel meaning not reached by age 100. The main Score uses fixed anchors and the four financial weights, normalized among those four, independent of ratings and selections. Its default numeric values are unchanged and its financial coverage is 100%. Fit uses the same anchors and all six weights, and is available for a path only when every personal factor with positive weight is rated. A zero-weight personal factor does not require a rating; when both personal weights are zero, Fit is off. Scores do not depend on other paths or shortlist membership. All four financial weights at zero yield no main Score; Fit can still use rated personal factors. Lifetime wealth is outside both scores.

Education and financing

Undergraduate tuition is $11,950 per year for four years. Medical school uses $43,648 for four years; PA school uses $55,327 for two years; and law school uses $48,513 for three years. These are repeated source examples, not estimates for all schools. Dental professional tuition is an explicitly editable assumption of $55,000 per year for four years, in addition to undergraduate costs. Electrician and plumber apprentice fees are explicitly editable assumptions of $1,000 per year for five years, paid from earnings. The pharmacist route uses four undergraduate years followed by four PharmD years using Iowa resident 2026-27 tuition, excluding fees and living costs. Dental hygiene uses three M State 2026-27 tuition-only undergraduate stages; sonography uses a two-year college price proxy, and HVAC uses a school example with an unspecified price year. A tuition total override scales the route's annual costs proportionally, preserving their timing. These costs exclude living expenses and may omit other fees.

Explicit stage types and start ages determine timing and modeled loan category, without establishing eligibility. Undergraduate borrowing allows $5,500, $6,500, $7,500, and $7,500 in successive years, within a $31,000 aggregate, at 6.52% nominal. Professional draws are limited to annual professional tuition, $50,000 per year and $200,000 total, at 8.07% nominal. These are simplifying assumptions about net proceeds, not a determination of program eligibility, actual annual or lifetime limits, credit, origination fees, or future law. Actual draws are limited by the engine to current need. Apprentices receive no assumed college loan. All school loans repay over ten years starting at full qualification, including after residency, with real interest converted using assumed inflation. Annual school support is unrestricted living support only during undergraduate or unpaid professional school, never during the PA work year, paid residency, or apprenticeship.

The engine covers remaining shortfalls with hypothetical gap debt and flags required financing, cumulative financing and peak debt. A result that requires gap financing is conditional on obtaining it, not evidence that credit has been approved. Default annual assumptions are a $30,000 living floor, 20% target saving from after-tax wages, 4% real after-fee and after-investment-tax return, 5% illustrative state wage charge, 2.5% inflation for loan conversion, 6% real gap rate, and zero initial investments or school support. Federal and payroll taxes use the simplified 2026 single-filer wage model; future tax changes, credits, benefits, unemployment, household costs, and investment risk are not modeled. See wealth model for the ledger ordering and tax details, and evidence decisions for source limitations.

State and export

Share JSON uses version 1 and an exact evidence/model snapshot. Invalid or unknown fields are ignored with warnings. The exact evidence-2026-10-26/model-1.0.0 and evidence-2026-10-26-expanded-v2/model-2.0.0 snapshots migrate with a warning using their historical career and field membership; incompatible snapshots reset to defaults. The shortlist starts from the validated chart selection on migration. Current links retain up to six shortlist IDs and up to three chart IDs, with chart IDs repaired into the shortlist when space permits. Ranges bound ages, money, rates, weights and personal ratings before inputs reach the ledger; state tax is limited to 15% to keep after-tax wages valid for the supported scenarios. An input that changes the snapshot cannot silently reuse old assumptions. CSV exports the caller-supplied shortlist annual rows, both scores and their coverage, model and evidence versions, price basis, assumptions, settings, all six original weights, annual wages, loan draw terms, tax, debt, gap-financing and balance-sheet flows. CSV financing remains conditional as described above.

Score anchors

Each factor maps to 0 to 100 against fixed anchors, the same for every path. Weights are normalized by their sum. These anchors and weights are product assumptions, not statistical estimates.

FactorAnchorsDefault weight
Typical pay$40K to $250K, log scale25
High pay$70K to $500K, log scale20
Time to $40KAge 18 to 32, earlier is better20
Tuition and fees$0 to $350K, lower is better15
AccessYour rating 0 to 100; counts only in the Fit score10
LifestyleYour rating 0 to 100; counts only in the Fit score10

Sources

Every source was accessed on its listed date. Indexed and historical entries are labeled; a historical posting is a past compensation example, not a current vacancy.

O*NET / US Department of Labor, BLS OEWS

US Bureau of Labor Statistics

Minnesota State Community and Technical College

Miami Dade College

Tennessee College of Applied Technology Dickson

University of Iowa College of Pharmacy

Career Opportunity model

Internal Revenue Service

Social Security Administration

Federal Student Aid

O*NET / US Department of Labor

College Board

AAMC

Duke University School of Medicine

University of Washington School of Law

American Board of Family Medicine

Tri-City JATC / IBEW Local 236

Google Careers

Anthropic