lion-king-broadway-scrape.r

  # import packages
  library(tidyverse)
  library(httr)
  library(jsonlite)
  library(lubridate)
  library(scales)

  # get list of shows

  headers = c(
    `sec-ch-ua-platform` = '"macOS"',
    Referer = "https://tickets.broadwaydirect.com/tickets/series/546654/the-lion-king-new-york-ny-1032631?startDate=03-01-2025",
    `Accept-Language` = "en-US",
    `sec-ch-ua` = '"Not(A:Brand";v="99", "Google Chrome";v="133", "Chromium";v="133"',
    `sec-ch-ua-mobile` = "?0",
    `User-Agent` = "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/133.0.0.0 Safari/537.36",
    Accept = "application/json, text/plain, */*",
    DNT = "1"
  )

  months = c("2025/03/01", "2025/04/01", "2025/05/01")
  datalist = list()
  for (i in 1:length(months)) {
    print(i)  
    Sys.sleep(10)
    params = list(
      requestedTime = months[i],
      salesChannel = "Web",
      seriesCode = "546654"
    )
    res <- httr::GET(url = "https://tickets.broadwaydirect.com/api/events/getbymonth", httr::add_headers(.headers=headers), query = params)
    raw = fromJSON(rawToChar(res$content))
    datalist[[i]] = raw %>% tibble() %>% select(ID, EndDate, AvailabilityCount, AvailabilityColor)
  }
  raw = do.call(rbind, datalist)
  shows = raw %>% 
    select(id = ID, timestamp = EndDate, availability_count = AvailabilityCount, availability_color = AvailabilityColor) %>%
    mutate(timestamp = ymd_hms(timestamp))

  # get price array for every show

  cookies = c(
    .AspNetCore.Culture = "c=en-US|uic=en-US",
    series = "546654",
    mcuid = "N1739756221112",
    `_evga_1a0f` = '{"uuid":"45903a9e7026fe47"}',
    lantern = "8ef5c960-5796-4ddd-a5c0-42210201f91d",
    `_pin_unauth` = "dWlkPVpqazVNakF5Wm1VdFl6STNZUzAwTlRCaUxXSXdaVEV0TWpaa1lqRTRaVEZrWkdReQ",
    CookieInformationConsent = '{"website_uuid":"c3c64d1a-0563-4694-9bcf-8c2259cc76cd","timestamp":"2025-02-17T01:37:22.615Z","consent_url":"https://tickets.broadwaydirect.com/tickets/series/546654/the-lion-king-new-york-ny-1212243/bestAvailable?startDate=04-26-2025","consent_website":"broadwaydirect.com","consent_domain":"tickets.broadwaydirect.com","user_uid":"5f7ef6f3-f712-47a6-96a1-15db198aefd4","consents_approved":["cookie_cat_necessary","cookie_cat_functional","cookie_cat_statistic","cookie_cat_marketing","cookie_cat_unclassified"],"consents_denied":[],"user_agent":"Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/133.0.0.0 Safari/537.36"}',
    `_tt_enable_cookie` = "1",
    `_ttp` = "CJMkMZeO5khkYhLmYIUj-RXcu9R.tt.1",
    `_sfid_c26b` = '{"anonymousId":"45903a9e7026fe47","consents":[]}',
    `__attentive_id` = "1a2cb11664e24e83af3840732cf499b7",
    `__attentive_cco` = "1739756221825",
    `__zlcmid` = "1QGnOwBCBZfUS23",
    `_gcl_au` = "1.1.273712821.1739756243",
    broadwaydirect_ga = "GA1.1.1086508586.1739756221",
    `__cf_bm` = "P8TI01C.PsouUeDk2WUCBOXVDt7WpKCHxlar.qUuRtM-1740877339-1.0.1.1-G6UeYdznF.5xYkdKmsNIXg5OJlzpoPU8SlSm3N.0Qt9C8R8FHwwKj9aABfrpSq3hzYMsXnpbLXfrNoEFw.K8EE_kczNHnqYi7r8csGZjNzQ",
    cf_clearance = "GJYAoEiOPWcpq2V_1S1IhJ.HD8etJIovzr5vu1BP9x4-1740877340-1.2.1.1-tDYDl6fyMUEeWc_YVCDVrNf.g50FcTcoTeMVmaHXytMhJPh5dSB8xABPz24WilYoUw7kjNbfmrBaTMfS4of51mOXr3gp7Ydsyj8iLLS3i4TIoAeTu_Z.XmNy.a1IX0gc.GVInaVWiihozlzDGplx675bVC4c2YeZ2ltTfppa59K9SUhiS4ygVdspm4jUZcjNiWxsSA4qHi7XOvxsBkw.WYAg_YWxgS2kDdRGfhsH1vt6.DmzlYlC5cRPXyZLpK2vNbDAJHf9Iglw82k8E8x57JgT.hP57.usv7hDFkvWHO2ZtLD19Fcz19OKVvPcyZNM6mlF1sRRvKIddAXpVAF.J5nNKcjjDkEUlXNHZGCCmMkopcURm3ksfXIENXLJmAut03DqR_AxzuptLv4pHoJ__9YqwlaRDxaGwPQ9ZTZgIPk",
    `_gid` = "GA1.2.1790488932.1740877340",
    mcsid = "TAKjvYHr",
    `__attentive_domain` = "broadwaydirect",
    `__attentive_ceid` = "2kG",
    `__attentive_session_id` = "28c347251f954a0097773cba2a3edea1",
    `__attentive_ss_referrer` = "https://www.google.com/",
    `_pin_unauth` = "dWlkPVpqazVNakF5Wm1VdFl6STNZUzAwTlRCaUxXSXdaVEV0TWpaa1lqRTRaVEZrWkdReQ",
    `__attentive_dv` = "1",
    `_cfuvid` = "iglmWmE7RP832t93QvQZ9.GsZLkFaEy4Smtac.PG2nU-1740877342949-0.0.1.1-604800000",
    `QueueITAccepted-SDFrts345E-V3_broadwaydirect` = "EventId=broadwaydirect&RedirectType=safetynet&IssueTime=1740877343&Hash=465a122a768c56e816e53f64e0b7fde3f9a30ebb305d817e79d2fac85d5a19c7",
    accountAllowsLogin = "true",
    cf_clearance = "0iW7KI_724w1mX3rbT839TgEGlZLC1xpIH9mcbvt.L8-1740877344-1.2.1.1-ldaFLI9go71F9xPkH7trYe02drjSQ8fEP4bmt6TXBPn7bed6RjUxqTbIu.kDQ6BRZ_qHhuBJ00C4ZImoaO1ooVsT_XMARRaHJlZHkYzy4s3.kasEoSd7.CrqG9xnQe_dioh3pZDblgpVhyQZIf1yiYtxE2pB8nF4qAu9Mf2vPgLvujxG05ht6cCHVXARtcEfnr.2EADPTMBrz0z9FARviWyXOi4A5bQpvLL7gVjoejzj3dCrgv_00_Qr4KxId7BBFnShMRw.UdlHqPO97bOupakiAO1EtT6COsS8KzYjG_AjUKBMxB9s79yGJ6NYIT4LE35DvQW5619ASlX6uXb7TK5ymQR1fZk73T6hyfg1.gNQIStfeSze5sF.Rw3X.8DnHU_Ie1fzVi31.vPQBhJstZbH57gsPSqiy8H3UObtPpI",
    consumerFeatures = '{"showNewBasketPage":false}',
    `__attn_exit_intent_triggered` = "true",
    gtmProductList = "",
    returnPage = "/tickets/series/546654/the-lion-king-new-york-ny-1212270",
    sellMoreSameEventPage = "/tickets/series/546654/the-lion-king-new-york-ny-1212270?startDate=03-16-2025",
    actualReferrer = "https://tickets.broadwaydirect.com/tickets/series/546654/the-lion-king-new-york-ny-1212270?startDate=03-16-2025",
    selectSeatsUrl = "/Tickets/series/546654/the-lion-king-new-york-ny-1212270",
    `_gat` = "1",
    nLivenClientNavLoginRedirect = "https://tickets.broadwaydirect.com/tickets/series/546654/the-lion-king-new-york-ny-1212242?startDate=04-19-2025",
    `_uetsid` = "fe290b00f70111ef900a93a7cb8c010b",
    `_uetvid` = "65a380204c7411efa66fed5678b32b4f",
    `_ga_CE4BESHVHE` = "GS1.2.1740877344.2.1.1740877863.60.0.0",
    `_attn_` = "eyJ1Ijoie1wiY29cIjoxNzM5NzU2MjIxODI0LFwidW9cIjoxNzM5NzU2MjIxODI0LFwibWFcIjoyMTkwMCxcImluXCI6ZmFsc2UsXCJ2YWxcIjpcIjFhMmNiMTE2NjRlMjRlODNhZjM4NDA3MzJjZjQ5OWI3XCJ9Iiwic2VzIjoie1widmFsXCI6XCIyOGMzNDcyNTFmOTU0YTAwOTc3NzNjYmEyYTNlZGVhMVwiLFwidW9cIjoxNzQwODc3ODYzODEwLFwiY29cIjoxNzQwODc3ODYzODEwLFwibWFcIjowLjAyMDgzMzMzMzMzMzMzMzMzMn0ifQ==",
    `__attentive_pv` = "9",
    `_ga_F9SL9L2PG0` = "GS1.1.1740877344.2.1.1740877864.0.0.0",
    `_gat_UA-26194249-1` = "1",
    `_gat_UA-43841211-3` = "1",
    `_ga_2BCYEF8XG6` = "GS1.1.1740877340.2.1.1740877864.58.0.0",
    `_ga` = "GA1.2.948673810.1739756220",
    broadwaydirect_ga_Y4Z8911974 = "GS1.1.1740877340.2.1.1740877864.58.0.240948745",
    .ASPXANONYMOUS = "0KgJzfiL2wEkAAAAMGI2MTMwNDktZGQzYi00ZTc5LWExNTUtNjAzNDAxYTY0ZGRj",
    `__cf_bm` = "tJFi0p1eGbkwtZ.hjR6L.wUqGMN_ZSqQSS4gPQM2lUw-1740877864-1.0.1.1-nkpLYZkRV7pGUj9qk855pc3LpvbTE7CnkJnqhT9MAziYHv7f3p8EkOEaV.AGHA16s9r.0Zm70oFgcMjidT4H01J9tL9DlgZAfvYNej013ug"
  )

  headers = c(
    accept = "application/json, text/plain, */*",
    `accept-language` = "en-US",
    dnt = "1",
    priority = "u=1, i",
    referer = "https://tickets.broadwaydirect.com/tickets/series/546654/the-lion-king-new-york-ny-1212242?startDate=04-19-2025",
    `sec-ch-ua` = '"Not(A:Brand";v="99", "Google Chrome";v="133", "Chromium";v="133"',
    `sec-ch-ua-mobile` = "?0",
    `sec-ch-ua-platform` = '"macOS"',
    `sec-fetch-dest` = "empty",
    `sec-fetch-mode` = "cors",
    `sec-fetch-site` = "same-origin",
    `user-agent` = "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/133.0.0.0 Safari/537.36"
  )

  params = list(
    includeOpens = "true",
    salesChannel = "Web"
  )

  ids = shows %>% select(id) %>% pull(id)
  datalist = list()
  for (i in 1:length(ids)) {n 
    print(ids[i])
    Sys.sleep(10)
    res <- httr::GET(url = paste("https://tickets.broadwaydirect.com/api/eventinventory/", ids[i], sep = ''), 
                     httr::add_headers(.headers=headers), query = params, httr::set_cookies(.cookies = cookies))
    raw = fromJSON(rawToChar(res$content))
    prices = tibble(raw$PriceMaps)
    datalist[[i]] = prices %>% select(DisplayPrice, Class, PriceLevelID) %>% mutate(id = ids[i])
  }
  raw = do.call(rbind, datalist)
  prices = raw %>%
    select(id, class = Class, price_level_id = PriceLevelID, price = DisplayPrice)

  # join together
  all = left_join(x = prices, y = shows, by = c('id')) %>%
    select(id, timestamp, price, class, price_level_id, availability_count, availability_color)

  # roll-up
  summary = all %>%
    group_by(id, timestamp, availability_count, availability_color) %>%
    summarise(Min = min(price), Max = max(price), Median = median(price)) %>% ungroup() %>%
    arrange(timestamp)

  # plot
  chart = summary %>%
    pivot_longer(cols = c(Min, Max, Median)) %>%
    ggplot(., aes(x = timestamp, y = value, col = name)) +
    #geom_point() + 
    geom_line(size = .9) +
    scale_colour_manual(values=c(Min="#f8069a", Max="#ecb516", Median="#6d1301")) +
    labs(x = '', y = '', title = "'The Lion King' Broadway Ticket Prices: Min, Max, & Median", 
         subtitle = 'March 1 to May 31, 2025 | Prices scraped from broadwaydirect.com') + 
    scale_y_continuous( labels = label_dollar()) +
    theme_minimal() +
    theme(legend.position = "top") +
    theme(legend.title=element_blank()) +
    theme(plot.title=element_text(face="bold"))

  setwd("~/Downloads")
  png(filename="lion-king-broadway-price-chart.png", width = 12, height = 6, units = "in", res = 1000)
  chart
  dev.off()

  # stats
  median(summary$Min)
  median(summary$Median)
  median(summary$Max)