Methodology
How the MoneyRoom Card Score works.
Every card page shows one score out of 10. It comes from three pillars — usable value (40%), effort (30%), and earning (30%) — computed from verified card data by the same engine that powers the app. No editor opinion, no advertiser input, and no affiliate links anywhere on this site. This page publishes every parameter.
The formula
Three pillars, one weighted score.
Each pillar scores 0–10 on its own, rounds to one decimal, and the headline is their weighted sum. A card can be brilliant on paper and still score modestly — the score prices in whether a typical cardholder actually collects the value.
Pillar one · 40%
Usable value, not face value.
A $200 credit you must remember every month is not worth $200. Each credit's face value is discounted by realization factors — one for how often it expires, and one for each real restriction the card's terms impose. The factors only apply on evidence from the verified card data; a credit with no restrictions keeps its full value.
| Credit resets | Factor |
|---|---|
| Monthly | × 0.65 |
| Quarterly | × 0.80 |
| Twice a year | × 0.90 |
| Once a year | × 0.95 |
| Every 4 years | × 0.95 |
| Restriction in the card's terms | Factor |
|---|---|
| Spend must go through the issuer's portal or a named program | × 0.85 |
| Spend only counts at a named merchant list | × 0.80 |
| One-time enrollment or app linkage required | × 0.95 |
| Posts later as a statement reimbursement | × 0.95 |
| Unlocked by a moderate spend threshold | × 0.85 |
Only the strongest single “where” restriction applies — they never stack. A credit marked automatic in the card data skips the enrollment and delivery discounts entirely. The pillar then compares the summed usable value to the annual fee: a card with no fee scores 10, credits that exactly cover the fee score 7.0, and roughly 1.43× coverage reaches 10.
Pillar two · 30%
Effort: what the card asks of you.
This pillar is deliberately dollar-free. It counts the mental load a card imposes: how many expiry deadlines you face per year, how many separate enrollments you must complete, and how many credits are locked to specific merchants or portals. A monthly credit adds 12 deadlines a year; an annual one adds 1. The combined load feeds a curve where zero upkeep scores 10 and each additional obligation costs progressively more.
Credits the score already excludes (see below) add no load — a typical holder ignores them, so the score does too.
Pillar three · 30%
Earning: real rates on a real basket.
Every card's verified earn rates run against one published reference basket, so no card gets a flattering assumption. Earning caps are blended in at the $30,000-per-year basket scale, and any category the card doesn't bonus falls back to its base rate.
| Category | Basket share |
|---|---|
| Dining | 20% |
| Groceries | 20% |
| Gas | 10% |
| Flights booked direct | 8% |
| Hotels booked direct | 8% |
| Everything else (base rate) | 35% |
Points are priced at the app's midpoint cents-per-point — the same values the in-app tools use, not a best-case redemption. Cash back is always worth exactly one cent. An effective earn rate of 3.0% on the basket scores 10.
| Rewards currency | Cents per point |
|---|---|
| Chase Ultimate Rewards | 1.5¢ |
| Amex Membership Rewards | 1.3¢ |
| Capital One miles | 1.15¢ |
| Hilton Honors | 0.35¢ |
Rotating, choose-your-category, and other conditional bonus rates never count toward the score — they fall back to the base rate, because a rate you must opt into or that changes each quarter is not what a typical cardholder earns. Portal-only travel multipliers are treated as a channel choice, not an organic spending category, and stay out of the basket.
What the score refuses to count
The honesty rules.
- Welcome bonuses are never scored. The score measures what a card is worth in year two and beyond — steady state, immune to bonus-chasing distortion. Card pages report first-year extras separately.
- Interest rates are out of scope. The score measures rewards value, not borrowing cost. If you carry a balance, no rewards card is a good deal.
- Per-use credits count as zero. A per-stay or per-trip figure multiplied by unknown usage is fiction. They are listed on card pages, never scored.
- Perks without a defensible dollar value are listed, not scored. Free-night certificates, lounge access, elite status, and insurance coverage vary too much by person to price honestly. This understates some cards — each card page says so specifically.
- High-spend unlocks are fenced off. Credits that require more than $25,000 of annualized spending are excluded from typical-holder value and labeled as unlocks.
Live worked examples
The same engine, every published card.
These scores are computed by the rating engine as this page renders — nothing here is typed in by hand, so the table can never disagree with a card's own page.
Cards whose details are still pending final verification carry a caveat on their page; the caveat never moves a score. Calibration sanity checks drew on published industry research — Frequent Miler's per-credit valuation discounts and LendingTree's rewards-breakage findings — but every weight and factor here is our own, and no third-party score or prose was copied.
Run your own numbers
The score is the average. You aren't.
The realization factors model a typical cardholder. If you know you'll use every credit — or none of them — the Profit Calculator lets you weigh any card's credits against its fee with your own usage, free and without an account.