well psa has auth guarantee and tag doesnt so it doesnt really matter what premium tag does or doesnt get when the buyer can just inad and be like ādurrrr this isnt the right card u sent meā
Thats what scares me.
If you have a big card
Even if you get signed delivery with insurance, buyer can be like āwhat are you talking about? I opened the box and this spiderman card was in here, no charizard to be foundā
Or
They can hit the item not as described button and when they ship the item back its an empty box
I think with a big tag card you want to rely on consignment and have them take the risk
im sure taking the same path as these sheeple who blindly follow āinfluencersā will work out the best in the long run
isnt that the same guy who takes a vintage pack and is like ādurrrr do i open it or keep it sealed?ā sure you can go ahead and open it if you failed hs math or buy lotto tickets
if u mean that all the ig accounts that buy all the big cards in tag 10s are also investors in tag the company then yes
1 Like
From my perspective, TAG comps against random PSA 10s are a complete distraction. They will always get attention on social media and can cause people to think things that arenāt true. A handful of isolated sales do not establish a long-term trend.
TAG has relied heavily on influencer-driven marketing, including people cracking PSA 10s and BGS 10s to submit to TAG. I think it is completely fair to question how organic that is, and to question the message it is meant to send. I also do not buy the idea that TAGās grading is some kind of stricter, fully objective standard just because they wrap it in āAIā language. Weāve seen enough examples that prove itās a facade.
Human decision making is still part of the process, no matter how futuristic they try to make it sound, and that human factor seems to be more present for the more valuable items and less apparent with items that wonāt get as much attention if the āAI-gradeā doesnāt align with the actual condition of the card.
To me, their marketing has come off as more desperate than credible for a long time. The āself-grading podā pitch was where it really crossed into absurd territory for me. That sounded less like innovation and more like a company trying to sell itself as the future before it had actually earned that status in the market.
Sure, but BGS proves that point does not really hold up. BGS 10 gold labels often outsell PSA 10s, and yet nobody seriously argues that BGS will overtake PSA in overall hobby preference. What people consider āpremiumā can sell for more without signaling a broader shift in market sentiment.
It is very fair to say TAG is now being considered āpremiumā like a BGS 10 might be to some people, but how ārealā that sentiment is and how long it will hold up? That is another question entirely.
Occasional TAG-over-PSA sales do not move me at all. That is not evidence of a durable trend. It is just noise unless it becomes consistent across the wider market over time.
If Collectors announced next week that they will be rolling out similar āAI-assistedā grading and reports at PSA and BGS, this would turn into a very different discussion.
11 Likes
Great point wow.
1 Like
I donāt think any reasonable collector wants a PSA monopoly. That leads to exorbitant grading fees, lack of innovation, groupthink mentality, etc. Competition is a good thing for most markets.
The biggest issue with TAG is its hyper-reliance on machine learning / artificial intelligence. Grading is inherently subjective, and this subjectivity can lead to misclassification at a population level even in well-trained models.
Letās say that you want an algorithm to classify a balloon. Your binary outcomes are Yes, itās a balloon or No, itās not a balloon. Now you train an algorithm on thousands of different balloons, including different shapes, sizes, colors, finishes, some with writing and some without writing, etc. This algorithm would identify balloons pretty well with a binary classification when compared to a cat, a box, a tooth, etc.
But we arenāt talking about binary classification with TAG.
TAG produces 36 1,000 point estimates for each card + consideration of surface wear on the front and back, dimensions of the card, etc. Can you appreciate how incredibly difficult it would be for an algorithm to have perfect or near perfect reliability and validity with so many 1,000 point estimates?
There are nearly an infinite number of ways that a card can be damaged, and this damage will not look identical across every card size, TCG, finish, holo, etc. Even if TAG had a well-trained algorithm off of hundreds of millions or a billion+ observations, it would be foolish to think that reliability and validity of each scan will be perfect or near-perfect to a 1,000 point mark.
Like human grading, there will be some level of variance around each estimate. I am still not convinced that TAG can grade more accurately than humans at a population level. Both types of grading will eventually make mistakes.
The question then becomes, can a human-graded service be improved by ML/AI? To my knowledge, PSA is already doing this with centering and sizing standards, and possibly for catching alterations as well. If PSA creates the same technology that TAG uses, what will be the purpose of using TAG? I have yet to see evidence that TAG uses something that could not be recreated from much larger parent companies (e.g., Collectors Universe - PSA, Blackstone - CGC).
These were my original thoughts about TAG before it hit the market. This thread now exists as a time capsule of prospective views of the company before they paid off IG influencers and were caught up in drama. Itās interesting to see how dramatically perceptions shifted as marketing became more aggressive.
6 Likes
Iām all for grading companies to give PSA a run for their money, but the TAG support on social media is so obviously forced lol
2 Likes
the best grading company at any given time is the one which gives me the highest grade
- them guys
4 Likes
I want to clarify something though regarding the 1000 point scale. Recently a submitter sent 2 different submissions in at the same time
The lower tier was was graded and returned more quickly despite being a cheaper/slower tier
Tag responded to the inquiry that the higher tier is delayed because the 1000 point scale requires human verification (and they are backlogged at that step)
Their systems literally trigger every possible issue and variance, and they have dedicated team of humans whoās job is to comb through all of it and make adjustments, identify false flags, ext
So my point is, its not all just ai algorithms spitting out grades there is more to it when it comes to the precision of 1000 point grades
Thank you for this note. I think your response further proves my claim, that their AI algorithm is unable to be valid and reliable at a 1000 point scale. I would be interested in seeing their statistics on accuracy and the margin of error associated with algorithm-alone grading.
4 Likes
i thought their end game is to put āgrading vending machinesā in every airport, coffee shop, school, and grocery store how they gonna do this when they got hoomans intervening 
To each their own and I actually love TAG, I used to to do everything with PSA back in the day. To be honest, whether TAG does well or not, I will still send some cards every now and then because I love the aesthetic of the slabs. I know some say they are flimsy but I never had any issues, I am not normally handling them in a way they would āeasilyā crack. Also the fact they donāt upcharge just because of the grade is a win for me. On top of all videos and evidence of the scummy things PSA has done ā¦.but I understand that all it takes is a few bad apples in anything to taint the entire reputation regardless if it is a company or occupation. I do love all the details they show you in the reports that you receive to explain why you received that grade. I know AI isnāt perfect either butā¦
2 Likes
dont ask me lol
maybe part of this process is taking all this human adjustment and having the ai evaluate all the data. For instance maybe after they have a sample size of a few million cards, they will use machine learning to evaluate all the adjustments humans made to the deliver the final score and why. Which would result in a more robust model, perhaps capable of giving a 1000 point score on its own without human involvementā¦ but thats pure speculation. Only they know
Sounds great if they can put it in practice. If they are doing this already then something is wrong since misgraded cards still are appearing.
1 Like
This is part of the issue with AI in general.
I work for a very large company that help looking at use-cases to have AI simplify and remove work humans need to do.
However, the current human resource required to check the quality of AIās work is more than for the human to do it in the first place in most instances.
A lot of companies are pinning big growth numbers and budget reductions because of AI benefits, and IMO places will get about 20% of the benefit they think they will.
5 Likes
correct thats a big problem
In TAGs case, according to GemRate they are doing ~45k subs per month. There is a small bump in subs, but if you go to the forums everyone is under the impression that they are slammed. Lower tiers being pushed back by 2 months and express tier is behind 2-3 weeks
compare to last year at a point when they were doing 40k subs and people were actually getting cards back early
The likely culprit > 1000 point score was added to their standard tier
last year there was a time when they pulled the 1000 score altogether and just gave a score out of 10 and only offered 1000 score in limited drops for more money. Everything was very quick as their systems did all the grading.
but, maybe thats why they are choosing now to do it. If it will cause a backup better to do it while the other companies are also heavily backed up so the delay doesnt looking abnormal compared to industry
If this means they can study the data and machines can learn from the human score adjustments then in the future it might possible to self automate the process but for now its a huge bottleneck and pain point
1 Like
Why does everything promotional from TAG feel slimy?
All those ground breaking auctions going the same or above PSA 10s that justā¦stopped happening months ago.
And now some crazy sale out of nowhere.
Hmh.
Personally my belief is they will try to sell TAG into this bull market to some big player (eg fanatics)
3 Likes