“ACG MORE CONSISTAIN THEN PCGS”

melikecoins · November 10, 2003 7:42AM

Yes according to CoinWorld article by Paul Richards. (11-17-03)

Remember the test CoinWorld did to test third part grading services?
Will Paul Richards put the data in to statistical form for use.

And in he’s article, through his analysis says that

“Accugrade demonstrates the best consistency compared to the other services,” and that “Sovereign Entity Grading Service displays desirable qualities of conservative grading and consistent grading”

Just a couple of statements from an article you should read.

Glen

Russ · November 10, 2003 7:45AM

I think a legitimate argument can be made that ACG is consistent. They consistently overgrade.

Russ, NCNE

chuckfarthing · November 10, 2003 7:47AM

LOL - I read the title as “ACG MORE CONSISTANT THEN PIGS”. Hope DH don't come looking for me.

MadMarty · November 10, 2003 7:50AM

Yep, they always suck!

tjkillian · November 10, 2003 7:53AM

I'm very interested in consistancy. The problem with ACG is not overgrading, it is grading problem coins. If all they did was overgrade by four grades, that would be wonderful, especially if they were very consistant. One would merely adjust the price scale to fit the grading scale and life would be fine.

My wife bought me a nice 1852 PCGS silver Three-cent piece. It was in a MS-63 holder. I cracked it out of its holder and put it in my type book. Later, I resubmitted it and it came back AU-58. Another 18 months later I resubmitted it to PCGS in its holder and it came back MS-62

Now, I've not been playing with this coin and being very careful with it, so why does the grade change EACH TIME I send it to them?

Tom

Bikingnut · November 10, 2003 8:05AM

I agree, consistently overgraded.

XpipedreamR · November 10, 2003 8:21AM

Accuracy and precision are related but differing concepts. ACG may be very consistent(high precision), but usually overgraded(low accuracy). Obviously, you want both precision and accuracy.

Manorcourtman · November 10, 2003 8:24AM

I agree with there consistency........They consistency suck. They suck the big suck.

dollardude · November 10, 2003 8:30AM

ACG fills a market need. They consistently overgrade coins so that dealers can take advantage of uneducated buyers who think that they are purchasing professionally graded items.

Barry · November 10, 2003 8:36AM

Throughout that article, the author correctly stated the sample size is too small to draw any valid statistical conclusions. So why did he run those stats anyway, and why did they publish it? A good example of Garbage In Garbage Out (as I think back to the nightmare days of Fisher's Exact Test, Chi squared analysis, etc.)

fivecents · November 10, 2003 8:41AM

Why would Coinworld publish such an article? ACG more consistant than PCGS ......give me a break.

dollardude · November 10, 2003 8:46AM

The guy was either a math nerd with no common sense, or he was somehow biased by ACG.

Edited to Add: As an engineer, I deal in statistics on occasion. Statistics are good, but you have to use common sense too!

Deadhorse · November 10, 2003 9:41AM

Most, if not all of those coins in CW were problem coins, wonder where they came from????

I don't put any stock in such a small sample with such coins. Let's try it with MS62 through MS66 coins and see who looks the best.

Oh yeah, better throw in a couple of AU58 Frankies just make it realistic.........

fcloud · November 10, 2003 10:08AM

It looks to be an interesting article. I only looked at the charts, and didn't waste my time reading the thing.

rainbowroosie · November 10, 2003 11:02AM

BOTTOM LINE: Which coins sell for more money??? Money talks, everything else walks.

melikecoins · November 10, 2003 11:05AM

The main prank in this whole thing is that the coins were bought from CoinWorld advertisers.

They where all bought as UNC coins and submitted.

Just goes to show you what some dealers will call Uncirculated.

Glen

Mark · November 10, 2003 11:08AM

I thought the article was fairly useless. An assumption underlying the method is that each grading service's grade of a coin is a random draw from the "true" grade of the coin. But I think that's most unlikely to be true because I have the prior that ACG (as well as other grading companies) consistently over-grade and so their grades are biased upward. So the grades from ACG, et. al, are not pulled from the same distribution of PCG, NGC, and ANACs, which means that analysis of variance is invalid. Note that the analysis is wrong--it's not just statistically weak--if the distributions aren't the same. (I think

)

Here's another way to think about this issue: Suppose PCGS and NGC were tested along with other grading companies that graded EVERY coin VG-10. Then PCGS and NGC would be shown to be more liberal--that is, overgrade each coin--and also less consistent--that is, their grades would be more variable than the other companies. But in this case the other grading companies are clearly inferior because every coin gets the same grade! Essentially this situation is an extreme case of what the Coin World article did. This extreme case makes clear the immense flaw in the Coin World article.

Mark

roadrunner · November 10, 2003 11:26AM

The irony of this situation is that Coin World went looking for an expose in the grading arena. what they really proved is that many of their major advertisers who they got the coins from are horrendous. That should have been the headline of the article:
"Primary Coin World Advertisers Shown to Sell Cleaned and Damaged coins."

Let's face it. NGC and PCGS don't have as much experience in grading problem coins as ACG, SEGS, and the others. Coin World
missed the point entirely.

roadrunner

Robgetty · November 10, 2003 3:14PM

Oooooh, I like that proposed article title!

“ACG MORE CONSISTAIN THEN PCGS”

Comments

Leave a Comment