Jump to content

Archived

This topic is now archived and is closed to further replies.

Jester

why is tf2finance used as evidence for key prices?

Recommended Posts

Jester

Isn't it a public beta? Does anyone have a concise explanation of how it collects and reports it's data?

 

It seems like a fair bit of the community refers to it for the average price, but i wonder how reliable the data is.

Share this post


Link to post
Share on other sites
MC22

Isn't it a public beta? Does anyone have a concise explanation of how it collects and reports it's data?

 

It seems like a fair bit of the community refers to it for the average price, but i wonder how reliable the data is.

Thats like asking why so much people use bp.tf. It's data is "outdated" too.

Share this post


Link to post
Share on other sites
Jester

I didn't say that, but lots of ppl trust it and refer to it as proof or counter proof to outpost.

 

I'm simply asking if anyone knows it's validity since it is a beta.

Share this post


Link to post
Share on other sites
Uranium235

He has a point though.

 

I know from first hand experience how hard it can be to properly extract data that is not well defined.

 

We trust tf2finance because it has shown in the past that the values seem correct. But there is a good deal of faith that you need to have.

 

I too wonder how it is possible to extract the data e.g. from OP correctly. You need to know the exact price and quantity.

Where does the data come from?

Is it the trades (posts) themselves or by comparing backpack deltas after a trade? How is that triggered?

How do you distinguish between "done" trades because something was traded or because the trade owner decided to re-open with a different price?

How do you know the correct price, with so many misleading trades?

How do you know the correct quantity?

 

I am sure a lot of trades are discarded because the data is not plausible, but on what grounds are they discarded and might that lead to a bias?

How hard would it be to manipulate?

 

Those are all unanswered questions, afaik.

 

I can understand that some answers may be perceived as trade secrets by some. It might also be that some answers would also show weaknesses in the system someone may want to exploit.

 

But I think tf2finance would do good by being a little more transparent in the data collection method. After all there is no such thing as security by obscurity.

Share this post


Link to post
Share on other sites
Jester

Thats like asking why so much people use bp.tf. It's data is "outdated" too.

Tf2finance is supposed to be"real time" though. Bp is not.

Share this post


Link to post
Share on other sites
MC22

I think only base64 can answer this fully.

Base64 's awesome list

Frequently used comments

 

Section 1

  • You have X Keys in your backpack
  • Nobody buys Keys at X refined
  • Outpost is full of highballers
  • The trades have no takers
  • Halloween is already over
  • Stop the damn inflation
  • I put up a WTB trade at X refined and people added me within minutes
  • By the month of X, we would have Y keys in existence.
  • The number of Keys in existence is actually growing
  • Trades from bots should not be counted as proof
  • Trades from quickbuyers should not be counted as proof
  • Price in Mann Co Store has always been $2.49
  • This has gone too far
  • The market will crash
  • Backpack.tf influences the market
  • When backpack.tf raises the price, sellers charge at backpack.tf price + 1 scrap, and their trades are used as proof again.
  • The vicious cycle repeats!
  • We need to stop this madness
  • 1 Key will be 10 refined next month
  • 1 Key for your 2 Keys
  • They need to go down
  • I remember that Keys at X refined 200 days ago
  • You can still get them at lower price from trade servers
  • You can buy Keys from Warehouse
  • Look beyond Outpost
  • Prices are rising without a reason
  • You are ruining trading

 

Section 2

  • You don’t have Keys in your backpack
  • People are buying at X refined
  • Outpost is full of lowballers
  • Keys have always been in demand
  • Steam Market makes Keys desirable
  • I put up a WTS trade at X refined and people added me within minutes
  • People get Refined Metal for free every week from idling accounts
  • The sellers at X refined are out of stock
  • You can’t get them at current price from outpost anymore
  • Keys from Warehouse are always reserved
  • If people stop buying Keys at higher price, they will go back down.
  • The majority of the trading community uses Outpost
  • X Refined reflects the majority of actual trades today
  • If it’s going to crash, let it crash.
  • Find me a seller at X refined
  • Add me if you are selling Keys at X refined
  • X Keys is a quicksell
  • I’m not buying for resell
  • I sold X hundreds of Keys this week at Y refined
  • More crates means more unboxing

 

Section 3

  • Stop being greedy
  • I don’t care about downvotes
  • I think this is optimal
  • 1 Key = X USD, 1 Ref = Y USD, therefore 1 Key = Z refined.
  • No proof downvoted
  • Where’s the proof?
  • You need more proof than that
  • This is not how it works
  • Stop whining
  • Not going to happen
  • Your links are from last week
  • Show me your inventory history
  • One trade does not represent the entire economy
  • Tons of links still doesn’t represent the entire economy
  • We shouldn’t touch the price of Keys
  • Range is too huge
  • Choo choo free rep train
  • The problem lies with Valve Corporation
  • Price is not based on opinions
  • We document prices here, we do not dictate them.
  • Last suggestion was passed X hours ago
  • I won’t vote
  • Prepare to be banned
  • Supply and Demand
  • (Discussions about the Great Depression, 1970s energy crisis, 2008 financial crisis, price of gasoline, milk, and other agricultural produce, gold, silver, and other precious commodities).

 

Section 4

(Insert valid proof here that helps this site to report true price so that users can make informed decisions.)

 

This summary covers the comments from the past 75 suggestions.

Share this post


Link to post
Share on other sites
Ryoudji

things

And how does this explain how the website collects data...?

Share this post


Link to post
Share on other sites
base64

why is tf2finance used as evidence for key prices?

 

 I don't know, only the submitter of the suggestion can tell you. 

 

I never submitted a single suggestion on Key prices. 

 

Isn't it a public beta? Does anyone have a concise explanation of how it collects and reports it's data?

 

It is a public "beta" because:

  • Its concept is from past experience of tracking Buds price. 
  • It's a rough sample, it doesn't track non-English servers. 
  • It doesn't track weapons, I don't have time to make a craft weapons list
  • The software is run on the same server as buds tracker. 
  • Steam API is getting slow nowadays

It seems like a fair bit of the community refers to it for the average price, but i wonder how reliable the data is.

 

Not really, only the pros.

The noobs say "I don't give a rat's ass about tf2finance"

 

proof or counter proof to outpost.

 

validity

 

It can prove that trades happened at certain price at certain time. 

It cannot prove that trades did not happen at certain price at certain time. 

It cannot prove that sample average of one hour is identical to population average. 

 

We trust tf2finance because it has shown in the past that the values seem correct. But there is a good deal of faith that you need to have.

 

I agree to this, the data has never been wrong, or has never been proved wrong. 

The site was originally a personal tool. Now it's information for people to decide what price to sell/buy based on 7 day trend.

 

Where does the data come from?

 

This has been answered many times, the combination of 5 sources.

 

Is it the trades (posts) themselves or by comparing backpack deltas after a trade? How is that triggered?

 

It's by comparing backpacks.

A user joins the watchlist under certain conditions, one of them being owning Earbuds.

The goal of selecting the sample is randomness. 

 

I am sure a lot of trades are discarded because the data is not plausible, but on what grounds are they discarded and might that lead to a bias?

 

Maybe 35% of trades for Keys.

They are discarded in the output process but not in the storage process.

They are discarded if they are impure, or exceeds +/- 4x Standard Deviation of the Sample.

 

How hard would it be to manipulate?

 

Very, it takes a lot of manpower. I don't think automated manipulation is possible.

 

Tf2finance is supposed to be"real time" though.

The published "average" it's not 100% accurate, but it's an approximation.

How good is it?

 

https://onlinecourses.science.psu.edu/stat414/node/167

 

Assume that observed standard deviation of sample is 0.1666 Refined (1 Scrap + 1 Weapon).

 

And assume that it converges to actual standard deviation of population as the number of data points go large.

 

Variance of Sample Mean = Variance of Population / Number of Samples

Share this post


Link to post
Share on other sites
Jester

This summary covers the comments from the past 75 suggestions.

What was the point of this post?

Share this post


Link to post
Share on other sites
Jester

@base64:

So in a nutshell, a trade takes place and you compare what the backpacks were before and after the trade for each user to determine the price?

 

How does the bud tracking work?

Share this post


Link to post
Share on other sites
Uranium235

How does the bud tracking work?

Why should it work any other way?

Share this post


Link to post
Share on other sites
Jester

Why should it work any other way?

Because it gives a cash amount. I understand the keys aspect. How does it come up with the USD value?

Share this post


Link to post
Share on other sites
Uranium235

Because it gives a cash amount. I understand the keys aspect. How does it come up with the USD value?

Right you are, forgot about that. I was referring to the value tracked in Keys.

Share this post


Link to post
Share on other sites
base64

Because it gives a cash amount. I understand the keys aspect. How does it come up with the USD value?

Rep thread difference. Only monitored when Earbuds to Key value crosses whole number.

Share this post


Link to post
Share on other sites

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...