What would an acceptable rate of authenticity mistakes be?
 

1%...3%.... .01%
I asked who guys trust on here for authenticating a pre-war card. The early consensus seems to be
"It depends on the card"
Reasonable. I ageee, in part.
The issue is, it is completely impractical and unrealistic for a grading company to staff experts on every issue.
Is the solution to limit what they authenticate?
Or, perhaps, pay knowledgable collectors enough to be worthwhile...or even train the staff better?
Many of these kids grading our cards are novices by our standards. Honestly, some of them were in 1st grade
when I filled out my first PSA submission form and decades away from conception when I built my first sets.
I'm taking careful notes on each topic I post.
Frankly....we're the hobby, Lol. We're obsessive, sometimes compulsive knuckleheads that are mesmerized by numbers, art, stories, history and the chase. Our egos strive for the approval and even jealousy of our peers.
I honestly believe...honestly...that an authentication service that operates as a non-profit would hurt PSA...a lot.
Putting the bones in place isn't too hard.

I've spewed a bunch and apologize for length...
Netflix wasn't connecting tonight:o

If they are willing to offer a grade on a card from a given set, then the acceptable rate is zero. If they are grading cards that their guys don’t have adequate expertise in, then they are making a fraudulent representation to the public and should stop grading those cards immediately.

Acceptable is zero now a reasonable amount to expect is a different discussion and I am honestly not sure what the answer is.

I'm not following here -- non-profit entities do not tend to have lower costs. Usually the opposite. They are less efficient and take their profits out by calling it salary, benefits, and perquisites. Some advantage by avoiding taxation, but that is easily offset by their conflicted approach to compensation. I'm generalizing, and I am sure none of my assertions apply to any non-profit favored by any board members.

My problem is the TPG's want to point out what a tiny percentage of mistakes they make. But they use their entire population to generate that number. A much more valid statistic would be: Out of stuff that has been altered, forged or otherwise made ungradable what percentage has been graded?

If 95% of what is submitted is legit but you're getting 20% of the bogus crap wrong you can claim a 99% success rate.

I think this topic needs to be broken down a little bit:

First of all, the TPG must identify and prevent counterfeits. Cards that are not authentic because they are counterfeits do not appear to be making it into graded holders. That is very important.

Next, we have the Blowout phenomenon where previously-graded cards are obviously cracked out, altered, and re-graded at a higher grade. That has to stop. Once a card has been graded, it has to be impossible to surreptitiously crack, doctor, and re-submit. If the TPG's can't stop that, they are in trouble.

Beyond that, the TPG has to disclose and adhere to a set of procedures and criteria designed to mitigate the likelihood of "unacceptable presentation enhancement". I use that term because everybody seems to agree that some effort to improve the way the card presents is OK: soaking, flattening, maybe even a little scrubbing with soap and water. Everybody also seems to agree that adding color or paper is out as is using chemicals and trimming. (Now the shouting starts.) Trimming is a particular problem because it seems to be devilishly difficult to detect. Chemicals also, perhaps.

The procedures used by the TPG's need to be effective, but they also need to be economical. Enhancements to the presentation of a card that are undetectable through the diligent application of reasonable procedures to identify them, are going to get through. The market will eventually determine how this uncertainty affects prices.

The question misses the mark, IMO. Mistakes will happen. Grade 20,000,000 items and you will miss some. The question is how they are handled. "Go chase the seller" is not an acceptable response. "We made a mistake and we will be accountable" is.

According to PSA’s methodology, as I understand it, two people review each card and, if they disagree on the grade, a third person breaks the tie. If 2-3 people are looking at every card there should be virtually no mistakes, a lot less than 1%.

People make mistakes. It's just a fact of life. Zero % error rate is simply unreasonable when a TPG looks at millions of cards. Here's my take on this:

Declared value:

Under $100: 0.1%
$100-$500: 0.01%
$500-$100,000: 0.001%
Over $100K: 0.0001%

The idea here is that the two graders that are reviewing the card would be more experienced the greater the declared value

Very good points, all.
Adam, I like your angle of accountability.
I do not believe that PSA has 2 graders look at a card.
If they do, it's exceptional. I cannot fathom a 3rd grader assessing it, frankly, unless we're talking about a card north of 100k...which I own none and will never be a threat.
Gary,
your point is very interesting. I would have to agree with your assessment on that at this point.

Zero

I laugh at the comments who say "zero" is the acceptable tolerable error rate. I can appreciate the sentiment but because it is a human thing, there's no practical way it could ever be zero. I don't like mistakes in the grading obvious or otherwise but to say there is no room for a mistake ever is sort of comical to me.

I think the appropriate answer is as close to zero as possible with the understanding that their opinion is just that, an opinion. Some rely and some don't. No rule says you have to buy or not buy what they're selling....

Acceptable tolerance rate for making honest mistakes, due to time constraints = about a quarter of one percent

Acceptable tolerance rate for turning a blind eye to obvious alteration = 0 percent

Acceptable tolerance rate for knowingly bumping preferred submitters' cards by a grade or two = 0 percent

The stamp places do have experts in each field, and will occasionally issue one saying "we decline to render an opinion" some specialties have very few experts, and some items are very hard to be sure what exactly it is.

They take their time, and their error rate is very low. And on more common items, nearly 0. (I'm sure they mess up occasionally, but I've never seen a cert I knew was wrong. )

Coming from a machining and equipment repair background, an error rate of 0 is always the goal.
Some jobs didn't work out, but they generally didn't leave the shop.
some jobs, the downside of being wrong was that someone could die. Those were done with extra care.

I would have to agree completely. It sounds like all of us are more upset with the lack of accountability and dishonesty than the mistakes. You can fix mistakes.

What would an acceptable rate of authenticity mistakes be?

Less than they are making today.