How the Fried Rice Index is calculated.

1. What the index measures

The Fried Rice Index is a food-based affordability index built from restaurant menu data. It compares fried rice prices across cities and studies what those prices reveal about everyday affordability, local restaurant markets, and urban economic conditions.

The index is not a replacement for CPI, PPP, rent data, wage data, or official cost-of-living statistics. It is a narrower signal based on one common restaurant dish.

2. Why fried rice?

Fried rice is widely available, easy to identify on menus, and common across many cities. It appears in both casual and higher-end restaurants, which makes it useful for comparing both everyday affordability and the range of a city’s restaurant market.

A fried rice price can reflect labour costs, rent, ingredients, tax, supply chains, consumer demand, and local pricing norms. The goal is not to claim that fried rice explains an entire economy. The goal is to use a familiar dish as a concrete comparison point.

3. Dish classification

The index collects all fried rice dishes, but it does not treat every dish as equivalent. Each entry is classified before analysis.

• Basic: plain fried rice, egg fried rice, soy sauce fried rice, or classic fried rice
• Vegetable: vegetable, mixed vegetable, or mushroom fried rice
• Meat-based: chicken, pork, beef, or BBQ pork fried rice
• Seafood: shrimp, prawn, crab, scallop, or seafood fried rice
• House special: special, combination, Yangzhou/Yang Chow, deluxe, or mixed fried rice
• Premium: lobster, truffle, wagyu, XO, or other luxury fried rice

4. Baseline price and market profile

The index separates baseline affordability from broader market analysis.

The baseline city price uses basic and vegetable fried rice entries. These are the closest comparison points for everyday restaurant affordability.

The full fried rice market profile includes all dish categories. Protein-heavy, seafood, house special, and premium dishes are not treated as identical to basic fried rice, but they are useful for studying variety, price spread, and premiumization.

5. Data collection

Restaurant prices may come from official restaurant menus, official online ordering pages, restaurant-uploaded menus, recent menu photos, third-party menu websites, delivery apps, scraper-assisted searches, and public submissions.

Every approved entry must include the restaurant name, city, country, dish name, original price, original currency, CAD price, source link, source type, date checked, confidence score, and admin approval.

Entries without a usable source link are not approved.

6. Source reliability

Not all sources are treated equally. Official restaurant sources receive higher confidence than delivery apps or unclear scraper results.

• Official restaurant menu: very high reliability
• Official ordering page: high reliability
• Recent restaurant menu photo: high reliability
• Third-party menu website: medium reliability
• Delivery app: lower reliability because of possible markup
• Unclear scraper result: low reliability until manually verified

7. Confidence scores

Each restaurant entry receives a confidence score based on source quality, dish clarity, price clarity, and location certainty.

• 95% — official restaurant website menu
• 90% — official restaurant ordering page
• 85% — recent menu photo
• 70% — third-party menu site
• 60% — delivery app
• 50% — unclear scraper result

Scores may be reduced if the dish is ambiguous, the source may be outdated, the price may include delivery markup, the portion size is unclear, or the restaurant location is uncertain.

8. Currency conversion

Prices are stored in Canadian dollars for comparison. The original local price and currency are also preserved so the dataset remains auditable.

Each entry records the local price, local currency, exchange rate used, CAD price, and date accessed. Future versions may use live or periodically updated exchange-rate data.

9. City-level calculations

The main city price is based on the median approved baseline fried rice price. The median is used because it is less affected by unusually cheap or unusually expensive restaurants.

The market average is a 5% trimmed mean of all approved entries for a city — the cheapest 5% and most expensive 5% of dishes are excluded before averaging. This reduces the influence of statistical outliers such as a $7 egg fried rice at a specialty spot or a $48 lobster fried rice at an upscale restaurant. The trim removes Math.round(n × 0.05) entries from each end, which means trimming begins once a city has at least 10 approved entries.

For each city, the index can calculate median baseline price, trimmed market average, minimum price, maximum price, standard deviation, restaurant count, average confidence, and data quality label.

10. Data quality labels

Cities are labelled based on sample size and source reliability.

• Preliminary: 1–2 approved restaurants
• Limited: 3–4 approved restaurants
• Moderate: 5–9 approved restaurants
• Strong: 10–14 approved restaurants
• High confidence: 15+ approved restaurants with strong source quality

11. Public submissions

Public submissions do not directly affect the index. Submitted restaurants enter a pending review queue. An admin checks the source, dish category, price, currency, and location before approving or denying the entry.

Only approved entries are added to the public dataset and used in city-level calculations.

12. Downloadable datasets

The Fried Rice Index provides a date-stamped full dataset download from the Cities page. The report is a single CSV file containing two sections:

• City summary: baseline median, 5% trimmed market average, standard deviation, price range, entry counts, data quality label, confidence score, and last-updated date for every indexed city.
• Restaurant entries: every approved restaurant entry with restaurant name, dish name, category, tier, local price, local currency, exchange rate, CAD price, source type, source URL, confidence score, and date accessed.

The filename includes the UTC timestamp of download. All prices are in Canadian dollars. Original local prices and exchange rates are preserved so the dataset is fully auditable.

13. Limitations

The index does not account for rent, wages, taxes, transportation, groceries, household expenses, portion size differences, service models, delivery markups, menu changes, or exchange-rate fluctuations.

It should be read as a transparent restaurant-price signal, not as a complete measure of cost of living.