FrontierCoffee in New York City

The 10 highest-rated coffee places in the committed New York City pack, ordered by each listing's own Google Places rating — listings with fewer than 20 ratings are not ranked, and equal ratings break toward the larger sample. This is pack-wide standing, not the map's drive-time order; open a place for its report, or jump straight to it on the live frontier.

A listing joins FrontierCoffee when its name mentions “coffee”, “café” or “espresso”, or its Google type is Coffee Shop, Cafe, Coffee Stand, Coffee Roastery or Espresso Bar. Matches come only from each listing's own name and Google Places type tags — cuisine alone is never inferred.

Counts are listings inside each committed pack — a coverage sample, not a popularity or quality ranking. The mapped regions overlap (the Bay Area pack includes San Francisco) and cover only crawled Google Places listings, so they are not comparable measures of a region's food scene.

Counts are Google Places (New) listings from each region's committed crawl data; a place joins a frontier when its listing name or Google type matches that category. The pack build date shown is when the routing pack was compiled — the original crawl collection dates are not recorded, so listing freshness is unknown. Drive times on the map are free-flow estimates — road-class speeds, no traffic, no parking, no queue at the counter — computed only out to a two-hour limit; beyond that a place is simply uncharted. Quality scores are ratings scaled to 100 — a rating-derived proxy, not a taste measurement. Unrated places cannot be charted.

Only places whose Google listing says “Halal Restaurant” or whose name contains “halal”. Cuisine alone is never treated as certification — many halal kitchens are missing here, and listed status is unverified.

The full pipeline — corpus, road graph, speed model, routing and ranking — is documented on the methodology page.

FrontierFood is an independent parody research project about the geospatial accessibility of lunch. Not affiliated with Cognition or Google.