I have a feeling those comments are a bot network, they are all over reddit as well.
8 Likes
TAG isn’t perfect, but it solves a key problem: grading becomes rule-based, explainable, and reproducible. PSA grading is typically opaque and can vary from grader to grader, which makes it feel inconsistent and sometiems super random.
Even if TAG’s AI improves/changes over time and standards evolve, you can still anchor outcomes to a specific model/version and understand why a card received a grade, something PSA does not provide. With PSA, any updates to grading standards are essentially invisible to collectors
2 Likes
You astroturf TAG without even grading with them, it’s like leaving a 5 star review without ever buying the product.
In short: avg youtube commentor on ShortpocketMonster
But TAG does not give same grade each time it grades a card, so far from perfect. How is it possible to get “wrong” grade from TAG, since there is no human grading?
I haven’t submitted to TAG yet, so this is theoretical, but I’d genuinely like to see data on how much TAG grades vary if you submit the exact same card multiple times.
Even with no human grading, TAG is still a measurement system, and every measurement system has tolerance/measurement error (like measuring your car length 10 times with a huge ruler and getting slightly different values). Small differences in setup, e.g. positioning/alignment, angle/distance, lighting/reflections (holo), and camera/focus noise can shift individual sub-scores, and because TAG combines multiple measurements into one grade, that small variance can propagate.
So I wouldn’t expect 100% identical results every time, but I would expect it to be far more consistent and explainable than PSA. I definitely would not expect a TAG 7 become a TAG 10 and vice versa. Would love to see data, though.
3 Likes
@thurco genuinely nailing it 1 by 1.
brother please dont tag me in the replies of everyone who has a relatively calm and measured take on TAG v PSA 
17 Likes
PSA 6 → TAG 10 → TAG 5
7 Likes
Whether you mean to or not, when you bring this up from 12 years ago, you are effectively applying the credibility I had as a new collector to who I am today. Bringing this up basically just gives the reader reason to dismiss my opinions and ignore me. Who are they going to trust - me or the founder of the forum?
To be honest I don’t think you really even know how picky or not I am today. I have a very large PSA slab collection and can count on a few fingers the number of slabs I don’t agree with the condition. Funny enough they are super old certs where variance seemed to be much higher. 99% of my slabs were either graded by me or purchased with a simple front and back picture. And the past 2 years, I would say 90% of my purchases I spent probably 30 seconds or less looking at the pictures before buying. So am I really that picky? You and Dyl seem convinced I’m as picky today as I was 12 years ago, but that’s simply not true.
1 Like
Let me just say I was being hyperbolic with the 10+ photos comment. I may have made those requests 12 years ago (although I doubt I asked for a specific number), but I can’t even remember a time when I asked for more than like 1 or 2 extra photos and even that’s extremely rare. The vast majority (probably 90+%) of my large slab collection I either graded myself or I bought with just the PSA front and back scan or the seller’s own front and back picture/scan. I’m not nearly as picky as you may want to paint me.
My point with the large number of photos was to say, some condition issues are only seen with either a few high quality photos or lots of photos from different angles. And let’s be honest you are never going to get that. Hence why PSA’s (or whatever grading service you use) opinion on condition does matter so much because you have to trust them in large part when you buy. The point I was making was not “this is what I need for every slab I buy”, it was “it’s not as easy as you imply to buy the card and not the grade with the information given”. That’s it. That’s all I was trying to say.
Bingo. This 100%
Man I just had 1700 seconds of my time wasted
16 Likes
Disregard the old reference. The point still remains that most people are fine with front and back scans. You have to understand your post inherently reads as for lack of a better term as, “picky”, since most buyers don’t need extra photos of an already graded card. That’s the only point being made, which has now been exhausted and I don’t want to belabor it any further.
Anyway to try and get this back on topic, I’ve been grading the same 1st ed base cards from the same collection, most being pulled from packs, and the 10 rate is definitely lower. I don’t have the excellent data as op, but I’m sure I’m preaching to the choir. Also I’ve been grading with PSA since 2008-09. The new ownership was the beginning for most of these changes. Some make sense, others like this topic don’t; either way the consumer carries the cost.
19 Likes
I would like to see some literature from PSA on this. I don’t know how long it’s been in their terms, but clearly it’s a recent policy change to enforce this.
Would also like clarification on if this is a result of a formal change in standards, or retroactive application of the existing standard. For example, I believe Nat recently said that the change in wording of centering requirements did not reflect a new change, and that the standard has been the same for a number of years.
Also, are all reholders being regraded, or just cert numbers before any formal change? What about recent misgrades?
4 Likes
Also how I sleep, knowing my card binders are in a drawer with my jeans so I don’t have to worry about grading
10 Likes
Thanks for sharing the video. That’s not what I would have expected from TAG.
The creator cracked 7 cards and submitted each to TAG twice (same cards, resubbed after cracking) to test TAG’s “machine-consistent” grading. Here is a summary of the results:
| Card | Original slab grade (company) | TAG #1 | TAG #2 |
|---|---|---|---|
| Lugia VSTAR Gold | PSA 8 | 8 | 7 |
| Squirtle Art Rare (Pokémon 151) | CGC 8 | 9 | 10 |
| Mew EX Gold (Pokémon 151) | PSA 6 | 10 | 5 |
| Charizard VSTAR (was already TAG) | TAG 8 | 6.5 | 6.5 |
| Magikarp Art Rare (Triplet Beat) | CGC 7.5 | 10 | 9 |
| Ivysaur (Japanese Classic Collection) | PSA 8 | 8.5 | 8.5 |
| “Serge” Art Rare (BGS Pristine card) | BGS Pristine 10 | 10 (Gem Mint) | 10 (Pristine) |
9 Likes
Like most things with AI, TAG’s grading is overhyped. The system clearly has flaws and has been demonstrated to be inconsistent at picking up certain issues (I believe surface issues are the main culprit).
AI may indeed be the future of grading, but the technology is not ready yet or where it needs to be to provide consistent and reliable results.
7 Likes
TAG is an advertising company that grades cards. Look thru this forum, plenty of TAG 10s have been posted with dozens of scratches on them
1 Like
Just so you know, that’s not astroturfing. Astroturfing would be if a TAG employee came in here and posted fake reviews pretending to be unrelated to the company. @timewaster1700 very transparently said that he hasn’t graded with TAG before.
13 Likes
Definitely agree. The only vintage card I have got back as a 10 was an Erika´s Venusaur that was so minty (pulled myself) that it could have well gotten a black label.
POP control? To be completely honest, I´m not too sour about it. It does make the 10´s that I have more valuable.