p2p download statistics
p2p news / p2pnet:- The mainstream media continue to report inaccurate mis- and disinformation emanating from the entertainment and software cartels as though they’re reliable indications of what’s happening with online file sharing.
The music, movie and software industries present themselves as honest, but hard-pressed, entities, desperately striving to keep their heads and the heads of their contracted performers and employees above water while hundreds of millions of criminal men, women and child file sharers get up every morning, bent on robbery.
In fact, no theft occurs. Nothing is stolen from its original owners and then sold, or passed along in any other way, without their permission or knowledge; no money changes hands; and no one is deprived of anything. Nor has a sale been lost because of a digital file shared and the truth is, the labels, studios and software firms are all reporting record revenues.
This doesn’t, however, stop them from declaring that file sharing has become so bad that they’re hemorrhaging profits and in consequence, are being forced – very much against their wills – to sue their customers.
The Big Four record labels are leading the onslaught, with the movie and software companies repeating the carefully coordinated disaster messages, saying the various sue `em all marketing campaigns are working and that they’re ‘persuading’ file sharing ‘thieves’ to abandon the p2p networks.
However, the reverse is true. P2p research firm Big Champagne told us the average number of people simultaneously logged onto the p2p networks around the world at any given time in September, 2003, was 4,319,182. In the same month in 2004, it was 6,784,574. And in September this year, on average, 9,284,558 people were on the p2p nets at the same time at any given moment.
In 2003 in the US, 2,630,960 people were logged on at any moment, in 2004 the figure was 4,549,801, and in September, 2005, it was 6,747,591.
Big Champagne says the percentage of total downloads by type were as they’re shown below and it seems only porn videos seem to have significantly fallen off, at least between May and July, 2005.
Every other month from November 2004 |
|||||
| FORMAT |
November, 2004
|
January, 2005 |
March, 2005 |
May, 2005 |
July, 2005 |
| AUDIO |
73.85%
|
74.77%
|
69.62%
|
74.30%
|
70.70%
|
| ALL VIDEO |
13.07%
|
13.10%
|
14.20%
|
10.94%
|
12.59%
|
| Pornographic |
8.17%
|
7.72%
|
7.94% |
5.44%
|
5.71%
|
| Feature Film |
1.68%
|
1.65%
|
2.04%
|
1.97%
|
2.77%
|
| Other Video |
3.22%
|
3.72%
|
4.22%
|
3.52%
|
4.11%
|
| ALL IMAGES |
1.79%
|
1.56%
|
3.21%
|
2.29%
|
2.27%
|
|
Pornographic Images |
1.31%
|
1.25%
|
1.96%
|
1.41%
|
1.28%
|
| Other Images |
0.49%
|
0.31%
|
1.25%
|
0.88%
|
0.98%
|
| Software * |
0.77%
|
0.68%
|
1.11%
|
1.00%
|
1.91%
|
| Other ** |
10.51%
|
0.68%
|
11.86%
|
11.48%
|
12.54%
|
(* Software defined as executable files and compressed archives
** Other defined as miscellaneous system files (.dll’s, etc.), documents, corrupted/unidentifiable files and so forth.)
What’s happening is this:
‘Consumers’ would willingly pay any reasonable amount for downloads, but the Big Four are determined to milk them for all they’re worth, and we use the word ‘milked’ advisedly.
For the first time in history, raw information is available to anyone with a Net account. This means the former cartel cash-cows are becoming customers, again – people with free choice who can, and do, communicate with each other and who can, and do, make informed decisions based on information gleaned from the blogs and news sites which are slowly replacing the traditional print and electronic media outlets.
It’s the beginning of the 21st century digital revolution.
Stay tuned.
October 10th, 2005 at 10:16 pm
Have you ever questioned how BigChampagne get these stats? Where do they get their user numbers from? How do they know how many edonkey users there are? How do they know whos a proper user and whos a fake? How for Kazaa? How many torrent users are there? No one trusts Big Champagne and if they arent gonna explain where theyre numbers come from why should they.
October 10th, 2005 at 10:20 pm
You say “no one trusts Big Champaigne” but you also note that you don’t know how they get their stats — so why do you say you (or anyone else) don’t trust them?
October 10th, 2005 at 10:44 pm
interesting how porn has dropped and has very little overall market in Peer to Peer. Here’s an article with a video on why the industry really wants to use porn as an issue….
http://p2pnet.net/index.php?page=reply&story=1725
October 10th, 2005 at 10:45 pm
wrong link, this works better.
http://p2pnet.net/story/1725
October 10th, 2005 at 11:10 pm
Most of the reports come from spyware included in the p2p client. So many that are now free include this type of spyware. However those that use these applications are always on the lookout for those that don’t have spyware. The spyware also eats your bandwidth to make its report.
October 10th, 2005 at 11:22 pm
Ummm… i dont trust them BECAUSE i dont know where they get their stats…?
Here, drink this.
What is it?
Just drink it.
But what is it?
Trust me, its fine.
Who are you?
Im the person telling you to drink it.
etc etc.
October 10th, 2005 at 11:30 pm
you talk like bigchamagne is the enemy p2pnet often has their stats to show how fked up the cartels are so whats your problem?
October 10th, 2005 at 11:33 pm
the problem is how does anyone know how bc are getting their statistics? they could be mkaing the numbers up. why should i believe a thing they come out with. they want people to take notice of them.
October 11th, 2005 at 4:24 am
Why not believe BC IF and I mean IF they just make up stats then why not reflect what the RIAA is saying and get the Riaa to pay BC for bogus stat’s??????
October 11th, 2005 at 6:23 am
They run multiple supernodes/ultrapeers (which are basically users who allow searches to be routed througn them). So every search that passes through their computers are recorded and filtered. They can then extrapolate statistics from this, like for example the most popular artists/albums/singles. They then proceed to sell this data to big music and radio stations etc.
Of course this doesn’t take into account people blocking their ip ranges. On top of that, these stats only apply to gnutella/fasttrack, because of their supernode/ultrapeer protocol. Things like edonkey can not be traced the same way, that i know of atleast. And bittorrent certainly can’t be tracked using this method, so actually you should take any of their stats with a pinch of salt.
Likleyhood is, their stats are actually on the pessimistic side. You could probably add 50% to the overall number of sharers, or even more.
You could argue this is also an invasion of privacy. I’m not sure if data protection acts can be applied here.
October 11th, 2005 at 6:37 am
so:
1. how do they make sure they;re not counting fakes – because you can bet your hiney that the ‘most popular’ files are being flooded on networks with fakes….
2. they’re only counting *searches* not downloads…? how do they go from searches to a number for downloads?
3. they’re facilitating file sharing by acting as supernodes?
4. they have an interest in making sure the figures are as high as possible so they can sell good data to the record companies?
October 11th, 2005 at 10:49 am
Answers to your questions:-
(1) The have very complex filtering mechanisms in place.
From a programming perpective, i would assume, they use lookup databases of actual content they are monitoring/interested in. The search expressions will be broken down into tokens, and likely processed by some regular expression type algorithm, which will remove junk and give a uniform string. The content database will then be queried using SQL to find any matching content based things such as file sizes, dates, mp3 tag info and calculated hashes of the files.
These kind of automated filtering can usually weed out fakes, provided the system is sophisticated enough. But of course, no system is perfect.
(2) I don’t pretend to be an expert on the supernode/ultra peer protocols, but, i would think they can extract quite a bit of information from it.
When a client first connect to the networks, they usually contact a cached ip address of a supernode (atleast on fasttrack). The client then proceeds to upload a list of their shared files to that supernode. The supernode then contacts other supernodes passing indexed information along, and they pass those details to other supernodes, amost in a spider web kind of fashion.
Clearly just from this information, a supernode can determine statistics such as, how widespread certain files are. They can then combine this information with search hits, to extrapolate trends etc.
(3) Hosting supernodes, is not directly copyright infringement. Although they may be propagating p2p to a certain extent, the value of trend information far greater.
Companies spend millions of dollars each year in an effort to predict trends, monitor buying habits and such things. This information has an enormous value to the music cartels. With such information, they can determine which songs should receive greater airplay priority through radio stations. They can increase production of products, or decrease depending on demand. This can also be localized to specific areas/countries, depending on demand.
All this kind of information can be extracted from p2p networks, and it’s very valuable.
(4) Their mandate is to gather data and extrapolate focused trend information. It’s in their interest to provide accurate information, so, artifically inflating figures would probably not be beneficial to them.
But, i could be wrong, in the end they are just another company trying to make money, so they will do anything they can to achieve that goal.
Finally, other companies such as mediasentry use these same methods to root out potential litigation targets.