GPAnalysis - Is it worth it??

[gpanalysis]

Hi - just a quick qualification - we do not just track eBay sales - we track the following:

 

eBay

Heritage

AllStar Auctions

Vault Auctions

Mastronet

Yahoo

 

and although only one auction set, we also track JPTheMint when they auction stuff.

 

We do not capture ComicLink auctions as they do not list whether an auction ended with reserve met, so until they send us their results directly, we wont be recording their data.

 

Regards

Simon

[sterlingcomics]

Simon,

 

Thanks for clarifying this information. To be honest, I really couldn't find this information on the gpanalysis web site (and I tried before I posted; I was going by information that was available to me at the time I subscribed). Making this extremely obvious to your current / potential subscribers would definitely be a positive! (of course, please feel free to post a URL to show my lack of web browsing ability).

 

And, I did see this on the web site, but wanted to clarify: you do not post data related to RAW books; only CGC books...correct?

 

Is the idea for gpanalysis long term to set up "web services" for all of the top auction and seller sites to feed your main site (maybe via XML or other post method) with additional sales data? I can see other potential usefulness from this data as well.

[gpanalysis]

Thanks for clarifying this information. To be honest, I really couldn't find this information on the gpanalysis web site (and I tried before I posted; I was going by information that was available to me at the time I subscribed). Making this extremely obvious to your current / potential subscribers would definitely be a positive! (of course, please feel free to post a URL to show my lack of web browsing ability).

 

I'm one of the developers, so I'm not sure where exactly that information is conveyed. I do know we mention a number of auction sites on our home page. I guess our marketing staff can do a better job at making these auction sources more obvious! I'll let them know when they get back from holidays.

 

And, I did see this on the web site, but wanted to clarify: you do not post data related to RAW books; only CGC books...correct?

 

That is correct. CGC-graded books only.

 

Is the idea for gpanalysis long term to set up "web services" for all of the top auction and seller sites to feed your main site (maybe via XML or other post method) with additional sales data? I can see other potential usefulness from this data as well.

 

yes, that is the vision.

[Aces]

Do you count CGC books that have been de-slabbed?

 

I buy low-mid grade early marvels and break them out for reading copies.

 

I always am upgrading and when I sell the dupe I will include the blue label that I always save.

 

Does that count when sold on eBay?

[gpanalysis]

Do you count CGC books that have been de-slabbed?

 

No we do not. Unfortunately there is no way to know whether a de-slabbed book is indeed the original one that came out of the slab; that's what our categorization staff tell me. No reflection on yourself - just that we made a decision early on to record only in-tact slabbed sales.

[Aces]

I didn't think so because of the switch-er-ru problem you mentioned.

I was just wondering that's all...

[murph0]

A cool idea would be to track a few slabs. Perhaps if you could find some high resolution scans on eBay or snag a couple off of Heritage. Then follow it around and see how many times it changes hands over the year or two. It'd be really hard, if not impossible, to do but it'd be quite interesting as well.

 

Brian

[sterlingcomics]

if the system is designed correctly and could also track RAW sales (with definite controls to prevent sellers from "curving" the data) one could do all of the number-crunching they desire. I would think (maybe I'm crazy) this could effectively BE the new Overstreet...a continually living and breathing database.

 

Just watch out for Skynet...

[fantastic_four]

How could raw sales be tracked? Nobody knows what grade they are except to say 2/3 of them are almost certainly overgraded by a lot. At least with Overstreet you hope that the advisors aren't overgrading by 5 to 10 notches (I'm not naming any names, they've already been named here before...), but with E-Bay sellers, this kind of stuff is common.

[panhandlefool]

FF's absolutely right - if everyone graded correctly, there would be almost no need for CGC!! It's the overgraders out there that cause us to pay multiples for a "real" NM 9.4 comic. I would be hesitant to follow an online "raw" database based on the overgrading out there of raw books. If anything, it would cause sellers to intensify their overgrading to try to get the higher $$ shown on an online "raw" database...

[Aman619]

wait. If the data is to reflect what a certain comic sold for in a certain grade, it doesnt matter WHAT the true grade was for any raw book. Think about it.... the data tracks what a book in an advertsised grade SELLS for. When you buy an overgraded book for X, you dont KNOW its overgraded....yet. But you have determined (in your own mind) that that book in that "advertised" grade was worth paying X for.

 

So in terms of sales data used to determine what books have sold for in all markets in a particular grade, overgraded raw sales data should be included!!

 

 

e.g. YYour sale enters the record books as yet another piece of data to support the $1000 price level for that book in 9.4. It is meaningless to the database's raison d'etre (excuse moi) that the POS you got in the mail was really a restored VF-, right???

[Timely]

wait. If the data is to reflect what a certain comic sold for in a certain grade, it doesnt matter WHAT the true grade was for any raw book. Think about it.... the data tracks what a book in an advertsised grade SELLS for. When you buy an overgraded book for X, you dont KNOW its overgraded....yet. But you have determined (in your own mind) that that book in that "advertised" grade was worth paying X for.

 

So in terms of sales data used to determine what books have sold for in all markets in a particular grade, overgraded raw sales data should be included!!

 

 

e.g. YYour sale enters the record books as yet another piece of data to support the $1000 price level for that book in 9.4. It is meaningless to the database's raison d'etre (excuse moi) that the POS you got in the mail was really a restored VF-, right???

 

Not true, sometimes you do know it's overgraded if you have dealt with that seller before, and you adjust your bids accordingly. Also many times you see a book graded NM but in the scan it obviously looks no better than VF, again bidders bid accordingly. Raw census is a BAD idea anyway you slice it!

 

Timely

[fantastic_four]

You should start a price guide just based upon stckwizard@aol.com's auction results. Mint prices on EVERYTHING would be only a couple of bucks! It'd be like the market got reset by 40 years!

[sterlingcomics]

In regards to RAW sales: I was thinking of getting the RAW sales data (that Overstreet gets periodically) in a more timely manner similar to how gpanalysis is tracking CGC sales today. And of course Overstreet already assumes in its recorded RAW sales that the sellers out there are following Overstreet "guidelines" for their sales. It didn't even dawn on me the over-graded comics on eBay from less knowledgeable sellers could taint the data (I've been trying to forget about it ya know).

 

Absolutely, I've said this on other threads too--CGC wouldn't be necessarily needed were it not due to the growth of eBay and the Internet--suddenly anybody with a computer and an Internet connection is competing with traditional comic sellers-and their grading skills not as keen.

 

The points made here are definitely valid. A curve at every junction. Maybe a weighted average? Just trying to tackle the problem logically, but its overloading my capacitors...zzzaapp! (static electricity sucks). I say let's reboot the comic world matrix! .10 a comic for everyone!

[Aces]

--suddenly anybody with a computer and an Internet connection is competing with traditional comic sellers-and their grading skills not as keen.

 

I don't know...

Some eBay collector/sellers do a better job grading than some of these traditional comic dealers. *cough* *cough*

[sterlingcomics]

Point well taken (been there, done that).

 

Probably a better way of saying that is: when we are gathering the data for RAW data, how do we handle these cases.

 

1. inaccurate grades

Well, CGC occassionally mis grades, but statistically speaking it averages ok

and overall the percentage is statistically "neglible".

Does the sell of RAW books consistute the same? Dunno.

 

2. outrageous prices

Well, CGC creates some multiples of guide, and if demand is there, wow!

But hey, the same could happen with RAW books.

 

I guess as I type this the CGC data gpanalysis collects is susceptible to the same problems as RAW books. Maybe invalid statistical data (outliers) should removed from a dataset, or keep them in and use a weighted average. I did read one of Chuck's articles in CBG--doesn't Overstreet do that already? I remember one way to control prices was to have one seller buy a RAW book from another seller--giving the illusion of the desired price of the book in the market.

 

Lots of ways to fudge this if someone really wanted to.

[Rickdogg]

Speaking of GPA, their site has been screwy for a while now

[Darthdiesel]

If the data is to reflect what a certain comic sold for in a certain grade, it doesnt matter WHAT the true grade was for any raw book. Think about it.... the data tracks what a book in an advertsised grade SELLS for. When you buy an overgraded book for X, you dont KNOW its overgraded....yet. But you have determined (in your own mind) that that book in that "advertised" grade was worth paying X for.

 

Blatant habitual overgraders would skew the data set constantly, and guys like me who still buy from them despite the overgrading, albeit at a highly adjusted lower price, would rendering that scenario moot.

 

for example - I know certain sellers hold to the "all grading is relative" motto, to a fault even and call VG books MINT or NEAR MINT. Buying a key issue for $5 in this seller's "NM" would wrongfuly skew the data in this hypothetical database (where true condition shouldn't apply), make it useless.

[joe_collector]

for example - I know certain sellers hold to the "all grading is relative" motto, to a fault even and call VG books MINT or NEAR MINT. Buying a key issue for $5 in this seller's "NM" would wrongfuly skew the data in this hypothetical database (where true condition shouldn't apply), make it useless.

 

So why is Chuck allowed to submit his data to Overstreet?

[Darthdiesel]

Wasn't even thinking about him but what books of his do you know that I can get for $5