Skip to content

Latest commit

 

History

History
85 lines (51 loc) · 4.15 KB

Exploratory.md

File metadata and controls

85 lines (51 loc) · 4.15 KB

Exploratory...

...domain writing, question refinement, plots, & data analysis

Agora and Anonymous Marketplaces

Agora was chosen as a market to analyze because of it's immense popularity and high usage. Additionally (and subjectively), the conditions of Agora's shutdown were unique to me in that the admins voluntarily shut it down after a paper was published in August 2015 that exposed vulnerabilities that could de-anonymize Tor users.

This is in contrast to other markets of similar scale. The largest examples are Silk Road's demise at the hands of law enforcement; darknet markets Evolution and Sheep turning out to be massive exit-scams.

It's a stretch to say (and impossible to prove) that Agora's administrators were completely altruisitic in their voluntary shutdown; but such protections of themselves and their clients might suggest that conducting business professionally was a priority above others.

What's in the data?

Exploratory Plots of Categories, going from high level to fine grain. As the list prices increase, there are less drugs on offer and less places they're coming from.

Why set a price cutoff at $1000?

$1000 can generally be thought of as an upper bound for the amount a person is willing to spend on drugs for personal and social use (Demant, Muksgaard, and Houbourg 2016). Much of the price listings from the market data fall within this range.

p4 <- subset(p14, p14$usd <= 1000) # 905805

p4p <- ggplot(p4, aes(subsubcat, from, fill = usd)) + geom_tile() +
  labs(title = "Agora 2014: Drug ~ Location / Price (USD < 1,000), n = 905805", 
       y = "", x = "", fill = "Price (USD)") +
  theme(plot.title = element_text(family= "FranklinGothicSSK", face = "bold", size = 14,
                                  margin = margin(0, 0, 20, 0))) + 
  theme(axis.text.x = element_text(family = "FranklinGothicSSK", size = 12.5,
                                   angle = 45, hjust = 1)) +
  theme(axis.text.y = element_text(family = "FranklinGothicSSK", size = 11)) +
  theme(plot.margin = unit(c(2, 2, 1, 2.5), "cm"))

p4p + scale_fill_gradient2(low = "deepskyblue4", mid = "bisque1",
                           high = "firebrick3", midpoint = 500)

If that's the case, then why set a price cutoff above $1000?

Above $1000, there's a mix of 'placeholder' listings vendors will put up while they are out of stock - to maintain a presence on the market. The prices are deliberately set higher than is reasonable to discourage someone actually buying. But in addition, above $1000 there are 'legitimate' listings for products in higher ('intent-to-distribute') quantities.

Variables of Interest

  1. date: date the page was created - as found in metadata.

  2. vendor: vendor's username - as pulled from html.

  3. product: product listing - usually containing:

  • name (Master Kush, Tramadol, Chanel Watch)
  • quantity (100mg, 1/4 oz)
  • 'selling points' e.g. 'Reagent tested', 'MOST POTENT', 'free shipping'

  1. price: price in Bitcoin

  2. usd: price in US dollars

  3. rate: exhange rate of BTC-USD, for that date

  4. cat: product category, as extracted from html

  5. subcat: product subcategory, as extracted from html

  6. subsubcat: product sub-subcategory, as extracted from html

  7. feedback client feedback left for vendor

  • e.g. "5/5, A+++ Vendor 2 days ago anon"
  • scale of 0-5 possible, as well as no feedback at all
  • as extracted from html

  1. from: location product ships from

  2. to: location product ships to