Project 1999

Go Back   Project 1999 > Blue Community > Blue Server Chat

Closed Thread
 
Thread Tools Display Modes
  #1  
Old 10-05-2010, 04:40 AM
jetviper21 jetviper21 is offline
Kobold


Join Date: Apr 2010
Posts: 143
Default Dear, People of the interwebs using the auction tracker auction.goonquest.com

http://auction.goonquest.com

I was looking though my server logs the other day and noticed a lot of people are actually using the tracker. I've made some minor updates and added some feedback on when the last log was uploaded at the bottom of the page etc. Some data is a bit stale since i havent been uploading logs so if your an active user if you wouldn't mind uploading a log every once and awhile to keep the data from getting too out of date.

But this is more of a what kind of things would you like to see it do kinda post. so use this post for suggestions etc:

For turbo nerds that know ruby or general regular expressions this is the code for the parser ... if you have any improvements let me know: http://gist.github.com/611214 go easy on it i wrote this at about 2 am on a monday while pounding back jack and coke .. im surprised it even works at all (it catches about 90% of the auctions).

Your friend from the internet
- Stairs
Last edited by jetviper21; 10-05-2010 at 11:04 AM..
  #2  
Old 10-05-2010, 07:35 AM
Frice Frice is offline
Skeleton


Join Date: Oct 2010
Posts: 15
Default

Wow, thanks for that Stairs. I'll be sure to try and keep that updated. I've been spending a lot of time in EC, so should have a lot of auction prices.
  #3  
Old 10-05-2010, 07:59 AM
Dominick Dominick is offline
Aviak

Dominick's Avatar

Join Date: Aug 2010
Location: Tampa, FL
Posts: 92
Default

I have been using it. Its pretty good.

We should be able to remove data points, we cant right now. Sometimes I notice the parser gives strange values for some items.

Lets say there were four logs uploaded last week. Three parsed with "Fishing pole of calamity" having a value of 3000pp, 2500pp, and 5000pp. One parsed with a value of 92830pp. The average should be 3500pp, but the average is going to incorrectly be 25832pp
__________________
Phew, Sorry for skinning you amiable bear!<Vesica Dei>
"If someone points a gun at you, you call the police. If a bunch of guys are pointing guns at you, you call SWAT. If they're in spandex and pointing a super laser at you, you call OSI. And if they're dressed regular and pointing a super laser at your daughter, that's when you call SPHINX. (SPHINX!)"
Formerly of Morell Thule and Test Server
  #4  
Old 10-05-2010, 09:33 AM
FatMagic FatMagic is offline
Sarnak

FatMagic's Avatar

Join Date: Oct 2009
Location: Buffalo, NY
Posts: 410
Default

This is absolutely fantastic - thanks for this!!
__________________
  #5  
Old 10-05-2010, 11:05 AM
jetviper21 jetviper21 is offline
Kobold


Join Date: Apr 2010
Posts: 143
Default

Quote:
Originally Posted by Dominick [You must be logged in to view images. Log in or Register.]
I have been using it. Its pretty good.

We should be able to remove data points, we cant right now. Sometimes I notice the parser gives strange values for some items.

Lets say there were four logs uploaded last week. Three parsed with "Fishing pole of calamity" having a value of 3000pp, 2500pp, and 5000pp. One parsed with a value of 92830pp. The average should be 3500pp, but the average is going to incorrectly be 25832pp
ill add this later tonight

Also average is probably not the best way to track the price if if i hadn't have slept through stats in college i might remember how to do standard deviation =(
Last edited by jetviper21; 10-05-2010 at 11:08 AM..
  #6  
Old 10-05-2010, 11:20 AM
Tork Tork is offline
Kobold

Tork's Avatar

Join Date: Jun 2010
Posts: 127
Default

Mean = Expected Value = μ
Std Dev = Sqrt(Average((X - μ)^2)) = σ,
so it's the square root of the average of the quantity squared of each data point less the mean

(But I'd be surprised if your programming language of choice didn't have something already premade)
  #7  
Old 10-05-2010, 11:25 AM
Tork Tork is offline
Kobold

Tork's Avatar

Join Date: Jun 2010
Posts: 127
Default

And there's a strong tendency to assume in theory prices will trend towards normal (LLN), but in practice valuations are only normal on small time scales... when you have fewer numbers, and hence LLN's tendancy is lessened.

So it's a catch-22.

edit:
Just looked at what you have - it's actually pretty good. I'd change Average to Mean, and then let people set up their own triangular distributions - no need for Std Dev.
Last edited by Tork; 10-05-2010 at 11:28 AM.. Reason: additional comments
  #8  
Old 10-05-2010, 11:55 AM
jetviper21 jetviper21 is offline
Kobold


Join Date: Apr 2010
Posts: 143
Default

Quote:
Originally Posted by Tork [You must be logged in to view images. Log in or Register.]
Mean = Expected Value = μ
Std Dev = Sqrt(Average((X - μ)^2)) = σ,
so it's the square root of the average of the quantity squared of each data point less the mean

(But I'd be surprised if your programming language of choice didn't have something already premade)
Sweet thanks, ill get this added

This is pretty much how i did math in college
[You must be logged in to view images. Log in or Register.]
  #9  
Old 10-05-2010, 12:15 PM
jetviper21 jetviper21 is offline
Kobold


Join Date: Apr 2010
Posts: 143
Default

Ok thats all added so once people start cleaning up prices things will stable out somewhat

i also removed all the auctions with prices that where 9999999999 since someone was being a douche bag .. im logging this now so ill just ip block them if it happens again

but back to math shouldn't most prices be within 2 or 3 std deviations? so if a price is outside of that i should flag it to be checked? that would weed out most of the retards trying to break the system
Last edited by jetviper21; 10-05-2010 at 12:24 PM..
  #10  
Old 10-05-2010, 12:45 PM
Wizerud Wizerud is offline
Kobold


Join Date: Aug 2010
Location: Reno, NV
Posts: 149
Default

Just wanted to say thank you for this tool. I will be helping out with the removal of bogus entries while medding [You must be logged in to view images. Log in or Register.]
Closed Thread


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 On
HTML code is Off

Forum Jump


All times are GMT -4. The time now is 10:46 AM.


Everquest is a registered trademark of Daybreak Game Company LLC.
Project 1999 is not associated or affiliated in any way with Daybreak Game Company LLC.
Powered by vBulletin®
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.