Go Back   Steam Users' Forums > Steam Game Discussions > T - Z > Team Fortress 2

Reply
 
Thread Tools Display Modes
Old 04-03-2012, 05:15 PM   #76
HotShot888
 
 
 
Join Date: Feb 2012
Reputation: 677
Posts: 3,099
Quote:
Originally Posted by MattieInfo View Post
took samples over a very long time, possibly resulting in a great deal of item overcounting as items moved between backpacks and were counted twice.
Well in that case, we can as well assume some items weren't counted at all, because it moved from a backpack that wasn't checked yet to one that has been checked. Logically, if an item moves, the chances of it being counted double or not at all are about equal, so it all shakes out. It's obviously not 100% accurate, but the approximate numbers are close enough
HotShot888 is offline   Reply With Quote
Old 04-03-2012, 05:17 PM   #77
YogiBear93
 
 
 
Join Date: Apr 2011
Reputation: 36
Posts: 314
Quote:
Originally Posted by MattieInfo View Post
Very interesting data. I like the write-up, though I would rather you posted your raw numbers rather than extrapolated numbers. Or start from the raw and then show your projections based on that sample.

I don't think your evidence backs this up:

If there are only order-of-magnitude 10,000 buds in the Money Supply (note that not all "traded" buds will be in the supply-- that's going too far IMO) and someone duped 200-1000 buds, that could have a dramatic effect on the prices. Imagine someone counterfeiting 2-10% of all circulating U.S. dollars and injecting them at once into the system.

I'm also very curious as to how you harvested this-- the only way I know to do this while tracking OriginalIDs means you either (1) violated Valve's ToS dramatically or (2) took samples over a very long time, possibly resulting in a great deal of item overcounting as items moved between backpacks and were counted twice.

We won't really know how to trust your sample unless you clarify a bit more about your collection methods. If you don't want to disclose them publicly, you can add me to discuss ( steamcommunity.com/id/mattie ), but that's obviously up to you.

Again, interesting research. I may be able to get you some other data to do some other analysis of interest.
The intelligence of a post in a thread is proportional to the intelligence of the OP.

I have to agree with Mattie when I think about it. The duped buds claim seems a little sketchy. It doesn't seem to be based on concrete evidence and is a little more like conjecture.

The collection method of information is also extremely important, for this reason: http://dilbert.com/strips/comic/2008-05-08/
If the data cannot be gathered by others in a similar manner (or at least with someway of explaining how the data was gathered at all), then there's no way to truly know how accurate this information is. It could be spot on, but there's no way for an outsider to objectively look at this information without some doubts.

Edit: That being said, we are working with rather large numbers so the Law of large numbers and all it entails could be rather safely in play.
YogiBear93 is offline   Reply With Quote
Old 04-03-2012, 05:24 PM   #78
Kered13
 
Join Date: Apr 2008
Reputation: 2082
Posts: 5,103
Quote:
Originally Posted by invishiro View Post
Well, unless you are arguing that demand doesn't affect the price... then I'm drawing my conclusions about demand based on current prices for things...
Supply and demand involves three numbers*: Supply, demand, and price. From any two you can infer the third using some (usually unknown) equation. But in order to show that supply and demand aren't working, you need all three, and you must show that the three numbers do not satisfy the supply and demand equation.

You only have two numbers, supply and price. You can infer demand from the equation, but you cannot use your inferred number to prove that the equation is wrong. That would be a contradiction.


*Technically supply and demand are functions, their intersection is price, but we'll keep this simple.
Kered13 is offline   Reply With Quote
Old 04-03-2012, 05:32 PM   #79
invishiro
 
Join Date: Jul 2010
Reputation: 229
Posts: 1,053
Quote:
Originally Posted by Kered13 View Post
Supply and demand involves three numbers*: Supply, demand, and price. From any two you can infer the third using some (usually unknown) equation. But in order to show that supply and demand aren't working, you need all three, and you must show that the three numbers do not satisfy the supply and demand equation.

You only have two numbers, supply and price. You can infer demand from the equation, but you cannot use your inferred number to prove that the equation is wrong. That would be a contradiction.


*Technically supply and demand are functions, their intersection is price, but we'll keep this simple.
You are correct in a classical economics sense. I know the type of graph you are referring to, where the slopes of the lines reflect the change in supply or demand affecting the price. A shift in either category affects the price etc. But I'm not really trying to be precise here (although I technically had the supply number and price, so I could have attempted to guess the demand for it).
invishiro is offline   Reply With Quote
Old 04-03-2012, 05:51 PM   #80
crysomemoar!
 
 
 
Banned
Join Date: Mar 2012
Reputation: 97
Posts: 403
nice thread.
crysomemoar! is offline   Reply With Quote
Old 04-03-2012, 06:18 PM   #81
ForeverSoldier
 
 
 
Join Date: Dec 2009
Reputation: 185
Posts: 1,598
Sticky this in the Trading forum so the speculating s there can get shot down by pure facts. Please.
ForeverSoldier is offline   Reply With Quote
Old 04-03-2012, 06:42 PM   #82
MattieInfo
 
 
 
Join Date: Nov 2009
Reputation: 16
Posts: 17
Quote:
Originally Posted by HotShot888 View Post
Well in that case, we can as well assume some items weren't counted at all, because it moved from a backpack that wasn't checked yet to one that has been checked. Logically, if an item moves, the chances of it being counted double or not at all are about equal, so it all shakes out. It's obviously not 100% accurate, but the approximate numbers are close enough
Yeah, that's a fair point, though it depends how often items were traded to newer accounts from older accounts, and his method of scanning, etc.

I had a long talk with base1024 (the OP) whereby I was convinced his numbers were based on solid data. He also asked me to review the data, methods, and even some of the calculating source code he used.

I have a lot of knowledge of the Valve API and data mining (compared to the average developer at least). All of his data and process looks perfectly legit to me-- and the data is even captured in a more academic way than most people who would try to capture this. The timing of his crawling makes it unlikely that items were duplicate counted (in any real quantity).

His calculations may be a little fuzzy in some areas due to the way the data has to be harvested, but I bet his projections are +/-20% correct for all areas he mentioned. Probably closer to +/-10%, but even with a wide range the data is very interesting/useful indeed.

I do feel the point about duplicated buds still stands after reviewing his analysis, it doesn't have the data it needs to really test the duped-promo impact. That being said, his numbers are the best data we have by far and should probably be the "official" numbers everyone uses going forward, IMHO.

Thanks to the OP for his detailed work here. It really helps put things in perspective.
-Mattie
MattieInfo is offline   Reply With Quote
Old 04-03-2012, 06:46 PM   #83
YogiBear93
 
 
 
Join Date: Apr 2011
Reputation: 36
Posts: 314
There's one statistic that's missing that I think should be useful: the percentage of Steam Users who've actually played TF2.
YogiBear93 is offline   Reply With Quote
Old 04-03-2012, 06:47 PM   #84
Snyphurr
 
 
 
Join Date: Jul 2010
Reputation: 564
Posts: 2,891
Quote:
Originally Posted by MattieInfo View Post
Yeah, that's a fair point, though it depends how often items were traded to newer accounts from older accounts, and his method of scanning, etc.

I had a long talk with base1024 (the OP) whereby I was convinced his numbers were based on solid data. He also asked me to review the data, methods, and even some of the calculating source code he used.

I have a lot of knowledge of the Valve API and data mining (compared to the average developer at least). All of his data and process looks perfectly legit to me-- and the data is even captured in a more academic way than most people who would try to capture this. The timing of his crawling makes it unlikely that items were duplicate counted (in any real quantity).

His calculations may be a little fuzzy in some areas due to the way the data has to be harvested, but I bet his projections are +/-20% correct for all areas he mentioned. Probably closer to +/-10%, but even with a wide range the data is very interesting/useful indeed.

I do feel the point about duplicated buds still stands after reviewing his analysis, it doesn't have the data it needs to really test the duped-promo impact. That being said, his numbers are the best data we have by far and should probably be the "official" numbers everyone uses going forward, IMHO.

Thanks to the OP for his detailed work here. It really helps put things in perspective.
-Mattie
Rep to you and the OP.
Snyphurr is offline   Reply With Quote
Old 04-03-2012, 08:28 PM   #85
base1024
 
 
 
Join Date: Oct 2011
Reputation: 178
Posts: 13
Quote:
Originally Posted by YogiBear93 View Post
Thanks, OP for taking the time to do this. If nothing happens because of this information, at the very least it was an extremely interesting read. Will the website auto-update itself as Items are traded, new refined is made, and new accounts are made?
The website is static.

Refer to the last line in the page:

Quote:
There will not be a complete scan or a second scan from this site.
base1024 is offline   Reply With Quote
Old 04-03-2012, 08:56 PM   #86
swade9wade
 
 
 
Join Date: Jan 2009
Reputation: 33
Posts: 968
Quote:
Originally Posted by ForeverSoldier View Post
Sticky this in the Trading forum so the speculating s there can get shot down by pure facts. Please.
Please let this happen please!
swade9wade is offline   Reply With Quote
Old 04-03-2012, 11:09 PM   #87
SpyPirates
 
 
 
Join Date: Oct 2011
Reputation: 29
Posts: 30
It doesn't really provide any useful information if it doesn't give us trends over time.

So we have a lot of keys, tons of ref, and a bunch of bills & buds. The economy will go on as always.

But if we knew how the supply of keys moved, and how quickly ref was added to the economy, it would impact how we view the stability of these currencies.
SpyPirates is offline   Reply With Quote
Old 04-04-2012, 12:35 AM   #88
amiorjs
 
 
 
Join Date: Feb 2012
Reputation: 286
Posts: 143
Just gonna post this here:
http://tf2stats.net/items/

You can choose which link you think is more reputable.
amiorjs is offline   Reply With Quote
Old 04-04-2012, 12:50 AM   #89
Vixori
 
 
 
Join Date: Nov 2009
Reputation: 552
Posts: 2,920
I'm blind?

I can't see any proof :/
Vixori is offline   Reply With Quote
Old 04-04-2012, 12:52 AM   #90
penguinm
 
 
 
Join Date: Apr 2009
Reputation: 53
Posts: 192
You referenced V.Bills stats in your post, but didn't follow up on it with your website.

I guess you'd be the person to know, so how many are there? More than the ~600 estimate? Just wondering because I got my hands on one the other day and want to know actual numbers
penguinm is offline   Reply With Quote
Reply

Go Back   Steam Users' Forums > Steam Game Discussions > T - Z > Team Fortress 2


Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is Off
HTML code is Off



All times are GMT -7. The time now is 01:25 AM.


Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2017, vBulletin Solutions, Inc.
Site Content Copyright Valve Corporation 1998-2015, All Rights Reserved.