Statistically, it is simply impossible for all members of a private BitTorrent tracker to maintain an upload/download ratio close to 1. However, even the less strict ratios of 0.4 or less are hard to maintain for newcomers with limited upload speeds, a notion that has now been confirmed scientifically.
Researchers from the Tribler P2P team at the Technical University of Delft, Netherlands, took an in depth look at the ‘ratio economy’ at various private BitTorrent trackers. Interestingly, they found some striking similarities between this virtual economy and some of the recent events in the global financial system.
In the paper, the researchers looked into the BitTorrent share ratios of members of a private TV-torrent tracker. They found that the current ratio system is heavily skewed towards a small group of people with a lot of upload capacity, often armed with seedboxes. These greedy ‘rich peers’ take away a disproportional of the available upload ‘credit’ so that new peers, or poorer ones have trouble keeping their ratio above the required figure.
In addition, several variations of swarm populations were simulated to see what their effect is on the users’ share ratios. Overall, the researchers found evidence of a credit crunch in the current ratio tracking schemes currently employed by most private BitTorrent trackers. Even when everyone has similar bandwidth limits this credit shortage still occurs.
“We discovered a strange phenomenon in existing private BitTorrent communities,” Dr. Johan Pouwelse told TorrentFreak. “Lack of credit is hampering performance. We have been able to reproduce this effect in a simulator to further understand it. It has many similarities with the credit crunch where the greasing oil of the world economy is drying out.”
“Lack of credit makes for bankrupt states and poorly functioning BitTorrent communities,” Pouwelse added. While it is unlikely that the trackers will collapse from it, the paper indeed shows that the current emphasis on share ratios is not fair to all members of the community.
Previously, Bram Cohen, the inventor of the BitTorrent protocol has spoken out against share ratios. “A better approach would be to not count it against people when they download from heavily overseeded torrents. Or to just not use total upload/download ratios at all, or if you do only make them advisory and not a source of banning,” he said at the time.
There are indeed hundreds of alternatives to the current ratio calculation, and most of them will indeed guarantee a healthier sharing environment (less favorable towards ‘rich peers’). In the paper, the researchers show that a ‘seeding bonus’ where peers get more credits for their uploads may help and others have suggested to use the total seeding time instead of the upload/download ratio.
While the paper presents some interesting findings, it doesn’t say much about download speeds on private trackers, which tend to be much higher than on public trackers. The parallels with our collapsing financial system are interesting, and with the only difference that a ‘bailout’ or ‘bonus’ would actually help the poor peers on private BitTorrent trackers, we can’t be too sure if that’s also going to work in the real world.