Who came, and from where

The two city hosted night markets are the only dated events on the map with a measured crowd. The Placer.ai visitor origin export for the event days, both days combined and a 2 pm to 10 pm window, records the home ZIP of each visit at ZIP level and nothing finer [SEPL]. December drew 4,292 visits from 70 ZIPs; May drew 4,146 from 68. Six months apart, the two markets drew almost the same volume and the same crowd: 28 ZIPs sent people to both, across Miami-Dade and into Broward. Between the two the catchment tightened toward local: the share of visits from within five miles of Mishcon Athletic Field rose from 48 to 57 percent and the share from within seventy five miles from 85 to 93 percent. The largest single origin both times is 33162, the market's own ZIP (893 visits in December, 953 in May), followed in December by 33179 (383 visits, 2 mi), 33161 (224 visits, 2 mi), 33138 (168 visits, 5 mi), 33181 (131 visits, 3 mi) and in May by 33179 (280 visits, 2 mi), 33161 (278 visits, 2 mi), 33160 (204 visits, 3 mi), 33023 (150 visits, 5 mi).

Both markets drew most on the Saturday and tapered on the Sunday: 2,300 then 2,000 in December, 2,400 then 1,700 in May. The evening moved later. In December the crowd built through the afternoon, crested at 6 pm and thinned fast toward 9 pm; in May the early afternoon was lighter, the crest came at 7 pm and the crowd held past 8 pm. In the project's reading the market matured between the two editions from a daytime into dusk errand into an evening market, on warmer nights. The hourly series is the provider's rounded index for both days combined, so it shows shape, not counts.

What this measures. Visits are device based estimates by the provider for the market site and window, both days combined, and home origins are assigned by the provider at ZIP level with locations obfuscated for privacy; the map publishes nothing finer than a ZIP and its visit count. The two markets are compared on the same site, window and method, so the change between them is the finding; the absolute counts describe the provider's estimate, not a turnstile. The full ZIP table is the Catchment sheet of the workbook and catchment.json in the site's downloads.