This is the problem with grading...

[Peter_Spaeth]

Dan are you sure those are the same card, I think I see some different markings.

[WhenItWasAHobby]

Originally Posted by Peter_Spaeth

Dan are you sure those are the same card, I think I see some different markings.

Peter,

I can't say positively either way. This was originally posted on Post#130 by "vintagetoppsguy". It would help a lot if we could see the backs of both cards.

[Peter_Spaeth]

Corners look different to me, and the 9 has a prominent black speck just in from the border on the top left that does not appear to be there on the 8.

[vintagetoppsguy]

Originally Posted by WhenItWasAHobby

I can't say positively either way.

I can.

They are definitely the same card.

Here are the bottom right corners blown up.

PSA 8


PSA 9


Notice two things.

1.) How the bottom corners hook down on both cards.
2.) The two indentions on the right border in the same spots. You can only see so far up the right border because it's hidden by the slab on the 9, but you can see enough of it to know they are the same card.

Also look at the same corner ding on the bottom left border in the scans Dan posted above.

[Peter_Spaeth]

The top one looks like it has a slight rounding at the very tip of the corner that the bottom one doesn't have?

[glynparson]

there was a computer that graded cards, don't remember the company. I think they had more fakes in their holders than anyone. Percentage wise anyways. The computer would only grade as accurately as it was programmed. PSA needs to work on QC in my opinion. Also the individual that thinks Joe O should look at all the 10s? That is completely irrational and is not his job. You want to say Reza should, Id kind of agree on high dollar or vintage 10s, but this still is probably not feasible due to their high volume, the isn't SGC's submission totals PSA gets waaaay more cards to grade than SGC.

Originally Posted by glynparson

there was a computer that graded cards

does anyone remember who and how long ago? I have been working on this since 2005. Never heard of this before. It's gonna cost about 500k for the machine and software.

You can't ever replace the human eye though. The machine can't tell if there is good trimming or authenticity of a card. A 25x scan of the card for graphing like a finger print for identifying the card and any alterations. You never know if the card night get stolen and cracked then submitted later. This will detect if the card has been reported stolen or not. A 10x scan for grading reasons as well as a dimension scan with the factory/know variances for the particular cards. A backlight scan for alteration detection. Once the grade is given the slab will be barcoded for information log. If the card is determined altered it will also have a 2-3 digit code below the graded area for use to know what was done and where it was found. For crossover cards there will be and area on the back of the flip stating company and previous grade.

Someday I will work in the hobby but until then............

[cobblove]

it was a company called CTA.
Computer grading

[WhenItWasAHobby]

Originally Posted by vintagetoppsguy

I can.

They are definitely the same card.

Here are the bottom right corners blown up.

PSA 8


PSA 9


Notice two things.

1.) How the bottom corners hook down on both cards.
2.) The two indentions on the right border in the same spots. You can only see so far up the right border because it's hidden by the slab on the 9, but you can see enough of it to know they are the same card.

Also look at the same corner ding on the bottom left border in the scans Dan posted above.

After studying the two scans, it's my opinion they are the same card. There is a blemish on the 9 on the top edge near the top left corner that isn't on the 8, but that could either be on the slab or even on the card that could have been acquired after the crack-out.

But as David pointed out, there are a lot of identical edge imperfections. Also, the words "Willie Mays - CF" are dead matches when observing the white snow on those letters on both scans.

What most disturbing is that, in my opinion, the corners have been improved. The bottom two corners on the 9 appear to be sharper than the 8. So with that said, subjectivity of assigning accurate grades isn't the only problem with TPG.

[Peter_Spaeth]

Right edge looks like it has a rougher cut in the top card than the bottom card.

[SetBuilder]

I'm sure the optical algorithims can be modified to higher accuracy. As far as detecting if a card is authentic or not, I think that's what a grader's job should be. When a card is sent to PSA, an authenticator first determines if the card is indeed real, then scans the card into the grading computer for final analysis and numerical grading. In essence, the professional grader's job is now limited to determining if a card is real or not before a computer does the actual grading.

I also think computer grading can put a stop to resubmissions over and over for grade bumps. If a computer can scan a card and determine that it has been submitted before, it can reject that submission, or at the very least, not double count the pop report.

Somehow the conspiracy theorist in me thinks that grading companies want people to submit cards over and over, and may actually prefer inaccurate grades.

[Peter_Spaeth]

http://www.ebay.com/itm/1961-Topps-W...item20ca96c5ec

Originally Posted by SetBuilder

I'm sure the optical algorithims can be modified to higher accuracy. As far as detecting if a card is authentic or not, I think that's what a grader's job should be. When a card is sent to PSA, an authenticator first determines if the card is indeed real, then scans the card into the grading computer for final analysis and numerical grading. In essence, the professional grader's job is now limited to determining if a card is real or not before a computer does the actual grading.

I also think computer grading can put a stop to resubmissions over and over for grade bumps. If a computer can scan a card and determine that it has been submitted before, it can reject that submission, or at the very least, not double count the pop report.

Somehow the conspiracy theorist in me thinks that grading companies want people to submit cards over and over, and may actually prefer inaccurate grades.

I think I stated all that already in what I am working on.