Do they though? That’s all I want. I don’t need a PhD dissertation for each of my 1,500 graded cards.
7 Likes
Yup psa is good because they’ve been consistent. If sub grades are preferred than go to bgs
Good question. If TAG fails, that’ll be social proof that people DON’T want more than just a label and number. Either outcome is fine, I’m just glad something interesting happening outside of the PSA/BGS duopoly.
1 Like
They have a basic grading option too, that just gives a flat grade and a basic report of the grade.
1 Like
Your it? (someone had to say it)
1 Like
CGC owns a larger share of the market than BGS.
Grading is a weird market because it’s not ever as simple as “company x is better than company y”. It operates more like an institution that establishes standards like the ISO. In some ways competition in grading is good because it nets out benefits for the consumer but in other ways it’s not good for the same reason it’s annoying that different countries use differently shaped outlets.
It’s near impossible to knock PSA from their position simply because it’s the standard that was first established in the hobby. If TAG is successful, it indicates it really is tapping into something people want. But if it can’t penetrate the market, I don’t know if it indicates anything. For the same reason you could invent an objectively better outlet but people won’t use it because it deviates from the current standard.
Either way I don’t think the value proposition is as great as people think it is. There is a difference between what people say they want and how they act. People might say they want a more detailed breakdown of a given grade but what the market shows is that people actually want as simple as a system as possible. That’s why PSA is successful, why CGC simplified their grading scale, why only successful BGS grade is the conceptually simplest concept.
6 Likes
I’ll say the slab design looks pretty well done. I can see this type of detailed, tech-assisted grading gaining traction over time in niche cases. To state the obvious, at the end of the day, people just want a third party assessment of the quality and condition of their cards. This looks to fulfill that and seems to add a further level of objectivity and precision to the process. The obvious obstacle they face is the inertia of everyone already having established grading companies’ slabs within their collections.
Another question is: is this level of detailed report of every flaw on their cards something people actually want?
With a report like this, people can more objectively assess whether a card meets their condition preferences (e.g. max number of scratches, location of scratches, alignment etc).
On the other hand, I’m not sure that’s a decision people really need or want to make objectively. Practically speaking, I don’t think I’d actually want a report precisely detailing every flaw on my card other than in the exceptionally rare case that there were no flaws. A lot of people see grading as “10 or bust” and this just increases the “bust” potential exponentially. CGC understood this and changed their standards accordingly. If this company’s version of the black label is designed well I could see people chasing the “1000 labels” or whatever they end up calling them.
Outside of those niche cases, I think people are generally not going to even want this type of service. I think most people prefer a simpler process of matching a condition preference or a label preference to a specific card and making a value assessment. It remains to be seen if this type of report will facilitate that decision making process or obfuscate it.
13 Likes
I think you’d get 1000/1000 at PFM grading for that response.
9 Likes
The mental gymnastics people go through to not just grade with PSA is insane.
8 Likes
I would argue their detailed report actually isn’t the main value they provide. It’s simply a “nice to have” because it’s there. People don’t have to look at the report if they only want to see a 1-10 number, but should anyone be curious why their card received a certain grade, they can spend time to figure it out. It beats having to spend an equally amount of time guessing why a card received a certain grade by tilting the card in the slab at every angle, or squinting at scans that can distort or conceal defects from the limited lighting it provides.
The main value would be the repeatable, objective consistency of each grade (supposedly, if you crack any card and regrade, TAG’s imaging can identify whether or not they had imaged the card prior, resulting in the same grade and cert#, assuming the resubmitted card was not altered or damaged before resubmission - aka a true POP report). I really don’t have a problem with human grading, but if their numbered grading scale is based on relative adjectives, their grading scale is purely an opinion, by definition. Professional opinion is valuable, but if inconsistencies are inevitable at scale, the value of this opinion also loses its value. Receiving different experienced graders based on the tier of your submission is a clear signal of this inconsistency.
If time permits, I will put together a more thoughtful post on this topic later this week.
Whether it’s TAG, or even if it’s PSA being able to reduce human biases and inconsistencies, I look forward to retiring the phrase, “this card can receive a 10 on a good day.”
cpbog1
3 Likes
I think you make some assumptions here. The first one is that the card will receive the same grade if you were to replicate the grading process. Surely they could reidentify a previously graded card but we don’t know how consistent the grade is if you were to repeat the full grading procedure. At least they could produce objective metrics of the variability of their measurements. They would have to demonstrate their accuracy is to at least 4 significant digits for the 1000 point scale to make sense
The second assumption that the grading algorithm will never change. I looked at the Illustrator page and found it identified some “damage” that was spurious. It’s hard to believe they would never update their grading code. And every time they do push an update it would change the grades of existing graded cards. So instead you get a “this is a 9 under TAG v0.98 but upgraded to a 10 in v1.14”
1 Like
I think this is my biggest concern with a company that relies heavily on technology. TAG supposedly has graded 100,000 cards. That number may seem like a lot to an average consumer, but it’s too small to train and validate an algorithm across numerous TCGs, card sizes, shapes, textures, foilings, types of damage, etc.
Algorithms developed for classification require a lot of training so that they can make accurate decisions on novel inputs. Because one card can have between 0 and ∞ instances of damage and potentially thousands of types of damage that will look different by card game, card type, printing era, etc., the data frame to train and validate the algorithm would have to be exceptionally large and representative to be reliable and valid.
I’m not saying that it can’t be done; perhaps their algorithms are superior to the image classification algorithms currently used in other fields. I am saying that I would prefer some clarity on how they developed this algorithm, how they found it to be reliable and valid, and what their plans are for fine-tuning in the future. If a new version of the algorithm would alter previous grades, then it is no better than human grading.
4 Likes
The other noteworthy thing is that if the product is simply an algorithm, it can be plugged into any grading service. The day that TAG is a genuine threat to PSA is the day that PSA (or CGC) buys out the company and includes an optional “TAG-powered” report in their own grading pipeline
6 Likes
I would imagine that a buyout is their end-goal. Their board is comprised of wealthy businesspeople who build and sell companies.
8 Likes
Bingo. This is a company dying for a PE exit because there is no long term winning here.
More information is never a bad thing. It creates more accountability which in my mind is something every grading company needs. They just need to figure out how to present it only to people that really want to see it.
The slab itself looks nice too.
I actually like it. Their report points out what the flaws are and where they are on the card. I think that’s useful. More information isn’t a bad thing.
As a person who was a absolute fan of machine learning and neural networks, 100,000 data points is very VERY little and has a lot of room for bias. I’m guessing that most of these cards are in top condition, so it’ll be hard for it to grade those middle level cards. There’s also obscure sets, newer sets, cards that weren’t even RELEASED that they still haven’t graded yet.
This is a exit at it’s finest. I’m guessing that they’re trying to use the grading system almost like a clinical trial, then looking for a buyout using that inforamtion along with the other data ports they have. I don’t really see any evidence that they’re looking to grow out this company, but I’ll try to see If I can spot anything.
1 Like
They are very experienced, as people said: 134 patents is astonishing, seeing the nightmare just trying to get even 1 for a business.
Interesting to see how “miss grade” described in the video will impact the value of the cards.
You might have a 10 but it could well be a 7 if it was graded prior to our fix
At least with manual grading you can blame the grader for a one off issue but here all cards graded so far had the same oversight.
9 Likes