BitTorrent is a well-tuned game-theoretic system during download and a tragedy of the commons after. Private trackers turn the commons into a poorly tuned centrally planed economy (big improvement).
Let's reframe private trackers in economic terms and tune the system!
Thread 👇
Let's reframe private trackers in economic terms and tune the system!
Thread 👇
Private torrent trackers enforce upload/download ratio requirements. This is actually an "upload credit" centralized currency in disguise!
For example, a 0.25 minimum ratio requirement can be reframed as a credit system:
Credits = (4 * upload) - download
For example, a 0.25 minimum ratio requirement can be reframed as a credit system:
Credits = (4 * upload) - download
On private torrent trackers, you can view every file as having:
• File price = upload credits it costs to download
• Resale value = upload credits the user can expect to earn by seeding
Both variables strongly influence how users consume content on private trackers
• File price = upload credits it costs to download
• Resale value = upload credits the user can expect to earn by seeding
Both variables strongly influence how users consume content on private trackers
On private torrent trackers, users understand that older uploads have a low resale value. During "free leech" (a period where users can download anything for free) users download more old content that they wouldn't otherwise be able to afford!
Private torrent tracker users understand their "earning potential" (determined by factors like internet speed) and they work more or less within the tracker's economy, accordingly.
Users with slower internet seed for longer and download fewer files!
Users with slower internet seed for longer and download fewer files!
On private torrent trackers, status and altruism drive users to earn more than they need for subsistence (just like real life!)
DIME encourages users to maintain 1:1, 0.25 is just the minimum. Separate titles and exclusive perks distinguish the torrenting elite.
DIME encourages users to maintain 1:1, 0.25 is just the minimum. Separate titles and exclusive perks distinguish the torrenting elite.
Old torrents on private trackers have bad "resale value." Old file demand surges during free leech but supply doesn't change.
Bandwidth supply > content demand, but download prices can't change!
At the least, the price of downloading files should fall with age (like real life).
Bandwidth supply > content demand, but download prices can't change!
At the least, the price of downloading files should fall with age (like real life).
Imagine if every book was priced the same: 1¢/page. That's how private torrent tracker ratios work: a byte is a byte.
Trackers are markets. Each torrent is a distinct good. Instantaneous spot pricing should be negotiated based on content, rarity, age, and bandwidth supply
Trackers are markets. Each torrent is a distinct good. Instantaneous spot pricing should be negotiated based on content, rarity, age, and bandwidth supply
Private torrent trackers improve incentives, but central pricing is inefficient. Per-torrent spot pricing would help a lot.
What if we move the price negotiation and credit tracking to the protocol? Well, that's a token! Here's how that might work: medium.com/@jbackus/what-…
What if we move the price negotiation and credit tracking to the protocol? Well, that's a token! Here's how that might work: medium.com/@jbackus/what-…
☝ If you enjoyed this thread, my blog post goes deeper: medium.com/@jbackus/priva…
The charts+data are from this paper: users.eecs.northwestern.edu/~hq/papers/dim…
If you don't follow me yet, I tweet regularly about crypto, decentralization, and the history of file sharing. More to come, stay tuned!
The charts+data are from this paper: users.eecs.northwestern.edu/~hq/papers/dim…
If you don't follow me yet, I tweet regularly about crypto, decentralization, and the history of file sharing. More to come, stay tuned!
