First, we must distinguish between old torrents and new torrents. An old torrent that looks popular is most likely going to be a megapack. Although the torrent appears to have leechers, after downloading such a torrent, the new seeder finds that none of these leechers are connecting and downloading from him. The explanation for this is that they are not, in fact, leechers , but partial seeds who have downloaded what
interests them and are seeding it back. Anyone who does not have 100% of the torrent content will show as a peer rather than a seed. The bottom line here is that if it is old, it is not popular.
“OK”, you tell yourself. “I’ll find a popular torrent that is new, and even though I am not interested in the content, I can download it and seed it back to improve my ratio.” To analyze the wisdom of that strategy, we must look at some possible scenarios.
For this example I have chosen a 1 Gig torrent with a popularity of 16 snatchers in the 1st hour and declining by 25% thereafter. All peers use a cable modem with 900 kBps download and 90 kBps upload (which is the max I have achieved with mine). For all examples, I will use kBps, even though advertised rates often appear in kbps (1 kB being the equivalent of 8 kb).
This chart shows the level of bandwidth supply and demand. Initially, the demand far exceeds supply as the upload rate of peers is only about 10% of the download rate. As more peers complete the download and the snatch rate declines, excess demand drops accordingly.
Seeders average upload rate is at a maximum until the supply outstrips the demand, at which point leechers only use a portion of the available seeders’ bandwidth.
This and the next chart shows how each seeder fares depending on when they joined the torrent.
After about 12 hours, the snatch rate has slowed to a point where the additional data won’t significantly affect the results. As you can see, the breakeven point comes early, after about 2 ½ hours. Anyone jumping on after that point will have lost his gamble.
So far, not a very rosy scenario for the prospective seeder. But, hang on. There is more to the story. Things can get much much worse.
This may seem a familiar sight, but no, I have not switched the topic to astronomy. This is an upload bandwidth bubble chart and that giant yellow object represents a seedbox. The tiny blue orb is the cable modem we’ve been discussing, along with its tinier budget DSL moon (90 kBps down, 16 kBps up).
Here, a black hole seedbox leecher has wandered into the solar system and is being serviced mainly by the seedbox sun.
Just as our sun comprises 99% of the mass of our solar system, so the seedbox represents a similar proportion of upload bandwidth in this matchup. The seedbox here is one commonly available and has a 100 mbps bandwidth. That translates into 12,500 kBps, but just so no-one thinks I’m being unfair, I have reduced it to 10,000 kBps for the purpose of these examples. It doesn’t make much difference, though. Even if we remove a few logs from the bonfire, as we shall see, it is still plenty hot.
Admittedly, the preceding chart represents the worst case scenario of having a seedbox leecher. This chart shows the relative seeding when the leecher has cable.
Although the bubble charts were created for dramatic effect, this animation better represents the players because it depicts download as well as upload capacity.
Below, we will examine some scenarios with our new player, the seedbox, taking on the role of the uploader and pitted against him, the cable peers. No more mention need be made of our budget DSL amigos. They cannot even play on this table because the ante’s too high.
There is no excess demand here as the seedbox is able to supply the initial leechers at their maximum download rate. Consequently, they complete their download sooner.
The example has been set up so that the cable peers seed at their maximum until demand drops below what they can supply. Even so, the uprate drops much sooner since the seedbox is still soaking up the excess demand, rather than pushing it farther down the timeline.
Now we can see the contrast between the seedbox and cable seeders.
Zooming in our chart, we see can the impact on the cable seeders.
The breakeven point comes much sooner now. Only those who jumped on quickly come out ahead.
So, what happens if we make the torrent bigger, say 3 Gigs instead of 1?
Now the seedbox can rise to its full potential, filling most of the initial demand.An even starker contrast results between the seedbox and cable seeders.
Zooming in, we see that the cable gamblers were doomed from the start. Now there is no need to show an aging chart, since there is no breakeven entry point.
Going back to our 1 Gig torrent, let’s look at a more optimistic scenario, where 1 in 3 leechers drop off after they finish their download. (Those of you coming from public trackers are probably now thinking you have entered an upside-down world, praying for more leechers and fewer seeders.)
Once again, we can’t really see much until we zoom in.
Finally some good news. The ratios of the early-ins have improved.
The scatter chart shows the breakeven point has been pushed back, allowing later entry. Those who dropped still share while they are downloading, so they retain a small piece of the pie.
Though this shows an improvement, don’t count too heavily on seeders dropping. Except for those who need their bandwidth for other purposes, everyone has an incentive to stay on and seed.
Now, let’s see what happens if we make the torrent even more popular. We will double the 1st hour snatchers from 16 to 32, keeping the slowdown rate and other variables the same as in the 2nd example. (1 gig torrent and no dropped seeds)
The timeline has lengthened to allow for the increased popularity. Although the snatch rate is higher, the increased demand is still met by the seedbox and never exceeds supply.
Once again, even allowing the cable peers to upload at their maximum with the seedbox taking the excess, the average upload rate still drops in a relatively short time.
Just as in the “larger torrent” example, the seedbox can rise closer to its potential with a higher torrent snatch rate.
Zooming in we see that, unlike with the larger torrent, the early entrants here have a chance at exceeding 1 to 1 because they will be seeding longer and sharing more snatchers.
Although the entry time for breakeven is about the same as in the first example which had no seedbox and a lower snatch rate, seeding is required for a longer time period.
Additionally, there is an inherent danger when you increase the number of followers required to return your investment. Just as with a ponzi scheme, the bottom can fall out sooner than you expect.
There are some scenarios where your ratio is not so much at risk. We could reduce the size of the torrent, with the seedbox having much less initial demand to fill and getting proportionately less. But, the smaller the torrent, the less impact it will have on your tracker ratio, so even though small bets are less risky, they are also less rewarding.
We could also decrease the initial snatch rate, making it a much longer duration torrent with a similar reduction in demand. But this is an article is about “popular torrents”, after all.
Maybe you already have a 100 mbit seedbox and are encouraged by what you’ve read here. Don’t get too content. There is an even bigger threat out there than you; the gigabit seedbox!
Please share your comments and experiences on the comment area.
Similar:
0 comments:
Post a Comment