SuperTracker - The perfect BT site

[bblogs]

WARNING: THE CONCEPTS IN THIS POST MAY REQUIRE SOME ACTUAL THINKING
Some of you might want to stop reading now ;) lol

Now,
I'm sure you've all considered the aspects of seeders, leechers, speed, content, success, and popularity of different private trackers.

Well, one of the most obvious features of most trackers that I'm aware of is:

Many seeders = difficulty in keeping a good ratio

Take Oink or SCT, or demonoid for example. Point proven :mellow:
So I was thinking- wouldn't it be great to have a tracker system that allowed really heavy seeding, but not way overseeded.......... or make it possible for the average user keep on downloading and using even when stuff starts getting overseeded.

Well to achieve that, you'd have to think outside the square...so how could you get the perfect balance?

Option A) Have a lot of free leeches - that's basically what a lot of high lvl sites do.
Free leech is great for a while! When everything gets overseeded, free leech clears it up, people are happy, everyones ratio goes up, and everything is good.
PROBLEM: At the start of a free leech, speed goes down, then soon after, speed goes up....Imagine the seeder/leecher ratio of the site on a graph. It makes a wave according to how many free leeches there are.
Basically, there's no controlling how long and high those waves are...that's not a good thing for site owners/admin trying to get the "perfect tracker".
My point is that option A isn't sufficient

Well this is my big thought........ATM the system is:
You have a ratio, if everyone uploads as much as they download, then the site goes well. Everyone gets what they want in time.
Then, for different reasons, when a site has a good speed or content, then stuff gets overseeded. So if you're the whinging type, then the current system that most sites use are ineffective.
So what if you think outside the square, like I said:

Option B) A designated base seeder/leecher ratio is set, based on how fast the owner wants his "perfect" site to be...lets say in this case it's decided to be 5. (ie. if there was 500 people leeching than there would have to be 2500 people seeding)

Let's say there was 100,000 users. Half of them seed a lot, and the other half don't seed much. The S/L (seeder/leecher) ratio averages out at the 1:5 mark so the admin are happy :)
Then, the users start getting greedy and they start downloading a bit too much. The S/L ratio drops below 1:5.
Then the system calculates the lowest 10% sharers and bans them. This gets rid of some of the leechers and then the S/L ratio goes back up to 1:5 or higher very soon.....if the S/L ratio becomes higher than 1:5 then no one gets banned. In theory, people don't actually have to seed, although if the S/L ratio drops then they risk being in the lowest 10% and getting banned.

When the S/L ratio gets dramatically higher then 1:5, then invites are distributed, and more users come in, until most likely the S/L ratio goes down (and the lowest seeding 10% gets banned soon)
This process is repeated.

Additionally, users with a ratio of over 1:1 (or whatever the owner decides) becomes immune to demotion of inactivity bans.

With this system, downloading becomes reliable, fast (as fast as the owner decides....the S/L ratio could be made higher or lower at any time).
When stuff becomes overseeded users don't have to get a seedbox just so they don't get banned :cry:
People should only be worried about getting banned if they're lazy, or if the S/L site requirement is unrealistically high (which is the site owners fault)

Also, if it worked to plan, then it would become the preferred method of private BitTorrent tracker, and content would be awesome due to uploading users and designated uploaders.

-
-
-
-
-
-
-
So....let's have some opinion on this idea of mine:
Would it work?
Why?
Would you like it?
Would you sign up?
Do you think we should try it?
Could you offer assistance?
Would/Could you help set it up?

Yes I do realize that it's a major step...but is FST really the ultimate filesharing site?
Well I thought so...so I figured if there's heaps of people that could help achieve this, then it would be FST members.

So if all you're going to do is post "I think it would work"...then do it! :D

[sp0ngeb0b]

tl:dr ?

[TheFoX]

Option C). A system that also rewards for staying connected when there are no leechers left in the swarm. Quite a few sites are now doing this, so people can still earn ratio even if they have no one to seed to.

[bblogs]

Option C). A system that also rewards for staying connected when there are no leechers left in the swarm. Quite a few sites are now doing this, so people can still earn ratio even if they have no one to seed to.

That's one of the bonus systems that can be added to a site that I didn't mention above. Simply using the "connection time" method as a primary system is not possible:
(selfish) users would limit their upload to 1kb/s or cheat the system etc, but I think it's one of the great ideas that have come into some trackers lately.

[Daniel]

There are literally dozens of ways to ensure a healthy tracker. What it comes down to and which option to choose best is simple: what's your goal for the tracker? Do you want very fast download speeds, do you want torrents that stay alive for long periods of time, do you want a large and growing tracker or are you happy with a smaller userbase? Based on these answers you will be able to decide on how to best regulate with free leech periods/torrents, ratio requirements, etc.

First off, I do believe that Free Leech - be it for single torrents or for the whole tracker and certain periods of a time - can be a good thing. Of course they invite activity in waves but they also help in building ratios for each and every single user which can only be a good thing in the long run.

Your whole seeder/leecher ratio idea isn't bad because the way I understand it, there's no concern about a minimum ratio requirement for any account as long as the tracker as a whole is working healthy. Do I understand it correctly that a seeder is someone who maintains a global ratio of 1.0 or higher while a leecher is everyone below 1.0? In this case, you have to be VERY careful about the overseeding issue, because users with fast connections and/or seedboxes tend to seed in the hundreds of GB's while users with slow connections have trouble keeping a good ratio with a total transfer of only 10GB.




I think that the need to keep a certain ratio is a good thing, because all users should give back a large part of what they've taken. This can be a low value (ie 0.500) with little help otherwise or the required ratio can be higher (up to 1.000) with more additional support. Some trackers successfully use bonus points, karma points or whatever they're called on different sites. These bonus points generally help to keep a larger number of torrents alive - which should raise the tracker's attractiveness.

Another few ideas that go in a similar direction are some I haven't seen on a tracker yet. If someone downloads an older torrent (with no or an insignificant amount of leechers) the system gives him a small bonus, i.e. only 75% of the traffic count, because seeding it back is often impossible. The same goes for torrents that are significantly overseeded, because it's hard to keep a certain ratio if other people keep supplying through fast connections.

One has to find a middle ground between the option to let some users overseed and to let those other users keep their account who have difficulty in keeping a ratio because of this. You're very right with one fact: as long as people with fast connections are willing to seed beyond their requirements, the tracker will be healthy and fast.

[Roark]

I think hdbits points' system is pretty innovative in trying to extend the length of torrents.

Sorry, but I don't care enough about users with low bandwidth to risk my speeds on schemes designed to help them. And it's not like there's any surfeit of 0day trackers they can use and maintain ratios on. In any case, I think your Option B allows for too many freeriders. Just the top 10 uploaders on ScT (which is pretty damn close to a perfect 0day tracker for me, linespeed 10/1) for example have collective upload of ~350TB. It's not that they can't support freeriders, but they won't.

[nadeem111as]

i think points for seeding is a great solution why because the torrent speed will still high and the people that has a mid connection will get a good ratio with the points

[TheFoX]

(selfish) users would limit their upload to 1kb/s or cheat the system etc, but I think it's one of the great ideas that have come into some trackers lately.

With the proper tools, it is possible to detect users who restrict their upload significantly while the swarm has a healthy leecher base. Besides, the concept of adding a small boost for staying connected should always take a secondary role to actual seeding, so someone using a restricted upload would need to remain connected for much longer than someone who unrestricts their line and allows a natural upload to take place.

[marksman]

FSC ratio system is the best imho

[sense]

I agree that there is something inherently wrong with the way most trackers are currently set up. We also have to realize that there will never be a perfect system. With that being said, I like the idea of option B with a few tweaks here and there. I really do not want to reward people with leeching as their #1 goal, so something needs to be done about that. Some places have really low ratio limits (.4, .3).

I think part of the main problem are those users who are insanely obsessed with keeping their ratio over 4+. Those users who think they are doing the community a favor by purchasing a seedbox and seeding until they send out 10 copies. I don't really see how this helps out the majority. Aside from the original uploader (who leaves anyways after a certain amount of seeders are present) I think there needs to be a seed cap with popular, new torrents. This is so those people with poorer connections can at least have the chance to seed to 1:1. What that cap should be, 2 or 3, I am not so sure.

Another problem are those users whose sole purpose is to build a huge buffer so he can trade the account for a better rarer tracker.

[nanajeebus]

Oh man, "filesharing" has become so deliciously convoluted. Pay for a better connection to get a better ratio, pay for a seedbox if that falls through. Sooner or later all trackers are going to be are a bunch guys with seedboxes uploading to each other at unnecessarily high speeds.

But yeah, I support this and stuff. Also, old article, I know, but further reading just in case you haven't seen it yet.

http://thoughtyard.com/twiki/bin/view/Main/TheProblemWithRatios

edit:

I think there needs to be a seed cap with popular, new torrents. This is so those people with poorer connections can at least have the chance to seed to 1:1. What that cap should be, 2 or 3, I am not so sure.

Another problem are those users whose sole purpose is to build a huge buffer so he can trade the account for a better rarer tracker.

The problem with this is that the accounts with those huge ratios are the ones with the fastest connections, which means faster downloading. While it would be good for the people with slow connections who want maintain their ratio/"share back", it would be but bad for the people they're uploading to. And would go against the sites who pride themselves on their fast download speeds.

[Daniel]

Download speeds are a critical topic, that's true. The lower the allowed overseed cap is, the faster people disappear from torrents and what's left is the standard group of people with smaller connections. If that is what you desire, you will most likely find a similar constellation on public trackers. And that's not really good advertising for a private tracker ;)

The problem is: people who are overseeding help the tracker with faster speeds but they also hurt a part of the users who have slower connections.

[bblogs]

That's the point of having a base seeder/leecher ratio, because it directly controls the speed....

A site owner could set the S/L ratio to 50, and the entire system would still work well...and with awesome speeds (it would just only be possible for high bandwidth users.)