Wata Pop Report (June 2023)

[inasuma | 1,203]

Hey all. Figured I'd share some interesting data on the new pops.

This is where I've compiled all the data. For each console, you need to scroll down to the bottom-most content to see June data. Sorry it's annoying, I'm not an spreadsheet expert. lol

• Between December 2022 and March 2023, roughly 3000 new games were added to the pop. Then between March and June this year, over 5000 games were added (not including Switch, PS3, and DS).
• Sharp increase in new 10s over the last few months.
  • For comparison, there were only an additional 10 examples of 10 A++ assigned between Dec. 2022 and March 2023 (from 34 up to 44).
  • There's now been a 2x increase (from 44 to 103), excluding new platforms, between March and June.
  • More 10s have now been added than the entirety of previous 10s in given in Wata's history.
• There's now 349 total 10s on the pop if you include Switch, DS, and PS3.
• A++s saw a sharp increase too.
  • Between December 2022 and March 2023, only 212 A++s were added.
  • Now between March and June, an additional 1062 were added - a 5x increase.
• Of the new platforms added, Switch has an enormous population of 9.8 A+ and higher grades. Not surprising - it's the current generation and dealers were/are submitting sealed factory cases of new games.

Anything interesting you see? Please share. I always enjoy seeing trends in this data.

Edited August 10, 2023 by inasuma

[inasuma | 1,203]

Relevant video

 

[3rdStrongestMole | 430]

On 7/30/2023 at 12:37 PM, inasuma said:

Hey all. Figured I'd share some interesting data on the new pops.

This is where I've compiled all the data. For each console, you need to scroll down to the bottom-most content to see June data. Sorry it's annoying, I'm not an spreadsheet expert. lol

• Between December 2022 and March 2023, roughly 3000 new games were added to the pop. Then between March and June this year, over 5000 games were added (not including Switch, PS3, and DS).
• Sharp increase in new 10s over the last few months.
  • For comparison, there were only an additional 10 examples of 10 A++ assigned between Dec. 2022 and March 2023 (from 34 up to 44).
  • There's now been a 2x increase (from 44 to 103), excluding new platforms, between March and June.
  • More 10s have now been added than the entirety of previous 10s in given in Wata's history.
• There's now 349 total 10s on the pop if you include Switch, DS, and PS3.
• A++s saw a sharp increase too.
  • Between December 2022 and March 2023, only 212 A++s were added.
  • Now between March and June, an additional 1062 were added - a 5x increase.
• Of the new platforms added, Switch has an enormous population of 9.8 A+ and higher grades. Not surprising - it's the current generation and dealers were/are submitting sealed factory cases of new games.

Anything interesting you see? Please share. I always enjoy seeing trends in this data.

I feel like, unfortunately,  consumers are partially to blame for this. We all say we want a grading company that grades fairly, but then patronize the ones that are more likely to give higher grades for personal gain. I guess this is just being a normal human being, like, it's hard to expect anyone to behave differently,  but these companies know that higher grades mean higher sales (for everyone) regardless of accuracy. 

Same thing in card grading.  If I have a card that I believe is gem mint 10, I might sub grade with BGS. However,  it's much safer for me to submit to PSA and get a 10 (9 at worst) knowing the card value will be retained better than getting a 9 or even 9.5 at BGS.

And I guess this is where we run into the worst problem with grading.  It's bad to have steep grading variance within one company, and having that same variance between different companies just muddies the water even more. 

They shouldn't be changing their grading criteria,  but they obviously are, and haven't made that public. I grade through WATA so all my sealed games look the same, but damn, this makes me want to swap everything back to VGA. They still seem to be the most consistent with not changing their grading parameters.

[Code Monkey | 2,010]

1 minute ago, 3rdStrongestMole said:

I feel like, unfortunately,  consumers are partially to blame for this. We all say we want a grading company that grades fairly, but then patronize the ones that are more likely to give higher grades for personal gain. I guess this is just being a normal human being, like, it's hard to expect anyone to behave differently,  but these companies know that higher grades mean higher sales (for everyone) regardless of accuracy. 

Same thing in card grading.  If I have a card that I believe is gem mint 10, I might sub grade with BGS. However,  it's much safer for me to submit to PSA and get a 10 (9 at worst) knowing the card value will be retained better than getting a 9 or even 9.5 at BGS.

And I guess this is where we run into the worst problem with grading.  It's bad to have steep grading variance within one company, and having that same variance between different companies just muddies the water even more. 

They shouldn't be changing their grading criteria,  but they obviously are, and haven't made that public. I grade through WATA so all my sealed games look the same, but damn, this makes me want to swap everything back to VGA. They still seem to be the most consistent with not changing their grading parameters.

I'm working on a machine learning computer vision model that will objectively grade games. It's the only true way and the future of grading in my opinion.

[Gloves | 11,779]

12 minutes ago, Code Monkey said:

I'm working on a machine learning computer vision model that will objectively grade games. It's the only true way and the future of grading in my opinion.

I suspect that this would require a relatively hefty setup for photo automation, especially considering the near-microscopic nature of certain things. For instance let's say you're simple about it and just do a single photo of each "face" of the box, you're going to miss creases/cuts/etc. which would only shot up under certain light. There's photogrammetry setups I've seen before which would be handy for this, they autorotate a DSLR around the object on a pedestal, taking enough high res photos all around at every angle. 

You're going to run into things also which to you may be "objective", but bare some subjectivity to them. I for instance would argue that a sticker or signature diminishes a copy of a game, but some would say that of the two a signature at the least either adds or doesn't impact the value. So let's say you determine that a signature is a good thing, and you tell that to your ML model. What if the signature is customized (e.g. see image below)? What if the "signature" is just the previous owner writing their name on the game? How will it decide? Some celebs have pretty shit signatures, and game devs especially will tend to just kinda scribble and be indistinguishable from literal chicken scratch. Will it know if a signature is real or not, and if so how? 

[Code Monkey | 2,010]

14 minutes ago, Gloves said:

I suspect that this would require a relatively hefty setup for photo automation, especially considering the near-microscopic nature of certain things. For instance let's say you're simple about it and just do a single photo of each "face" of the box, you're going to miss creases/cuts/etc. which would only shot up under certain light. There's photogrammetry setups I've seen before which would be handy for this, they autorotate a DSLR around the object on a pedestal, taking enough high res photos all around at every angle. 

You're going to run into things also which to you may be "objective", but bare some subjectivity to them. I for instance would argue that a sticker or signature diminishes a copy of a game, but some would say that of the two a signature at the least either adds or doesn't impact the value. So let's say you determine that a signature is a good thing, and you tell that to your ML model. What if the signature is customized (e.g. see image below)? What if the "signature" is just the previous owner writing their name on the game? How will it decide? Some celebs have pretty shit signatures, and game devs especially will tend to just kinda scribble and be indistinguishable from literal chicken scratch. Will it know if a signature is real or not, and if so how? 

For the former, I have that figured out.

For the latter, you're talking about 2 different things here .We're discussing the model providing an objective grading of the game but then you start talking about value. The model doesn't account for any sort of value, it specifically gives a grade and you can do what you want with it to determine value. The signature would detract from the grade (as it should) but when you look at an item to consider purchasing it, you can add value back for the signature if you wish to do so.

Objectively, the model has done its job, separate from value.

[Gloves | 11,779]

Just now, Code Monkey said:

For the former, I have that figured out.

For the latter, you're talking about 2 different things here .We're discussing the model providing an objective grading of the game but then you start talking about value. The model doesn't account for any sort of value, it specifically gives a grade and you can do what you want with it to determine value. The signature would detract from the grade (as it should) but when you look at an item to consider purchasing it, you can add value back for the signature if you wish to do so.

Objectively, the model has done its job, separate from value.

I said value offhandedly meaning the same basic thing, but in this specific case you've gone the "detracts from" route which sounds good to me; was really just an example regardless. Not that there are many particular examples. I feel the same way that an ML trained to weed out flaws and provide an objective grade is certainly more valid than any human element within the process (ignoring the human providing the training), so all good there. 

Another concern were I in your boots would be whether people would actually value such a service. Likely anyone with a signed copy of a game would simply not submit such a game to your service, knowing that it would detract from the grade. You'd really need to be the de facto grading option for people to be universally sending such things in. 

[3rdStrongestMole | 430]

2 hours ago, Gloves said:

I said value offhandedly meaning the same basic thing, but in this specific case you've gone the "detracts from" route which sounds good to me; was really just an example regardless. Not that there are many particular examples. I feel the same way that an ML trained to weed out flaws and provide an objective grade is certainly more valid than any human element within the process (ignoring the human providing the training), so all good there. 

Another concern were I in your boots would be whether people would actually value such a service. Likely anyone with a signed copy of a game would simply not submit such a game to your service, knowing that it would detract from the grade. You'd really need to be the de facto grading option for people to be universally sending such things in. 

I honestly don't even know how current grading companies approach signatures. I've asked WATA in numerous emails, and I've had signatures on my games authenticated by JSA, and WATA support basically shrugs and tells me they don't know how it will impact the grade.

I think a programmed grading service would be cool. God knows those high grades would be even more valuable and exist with more objectivity than any other avenue right now.

I guess it'd be like my BGS vs PSA example. I'd send the cream of the crop to the objective grading service, and go with WATA or the like for all my question marks.

[Code Monkey | 2,010]

3 minutes ago, 3rdStrongestMole said:

I honestly don't even know how current grading companies approach signatures. I've asked WATA in numerous emails, and I've had signatures on my games authenticated by JSA, and WATA support basically shrugs and tells me they don't know how it will impact the grade.

I think a programmed grading service would be cool. God knows those high grades would be even more valuable and exist with more objectivity than any other avenue right now.

I guess it'd be like my BGS vs PSA example. I'd send the cream of the crop to the objective grading service, and go with WATA or the like for all my question marks.

I've seen games with signatures and also PlayStation games with holes punched in the UPC get above 9.0 grades. They didn't affect the grades at all.

[Gloves | 11,779]

17 minutes ago, Code Monkey said:

I've seen games with signatures and also PlayStation games with holes punched in the UPC get above 9.0 grades. They didn't affect the grades at all.

The hole punch/drill thing is another subjective thing where you'll just have to make your call and people will either agree or disagree. 

They don't ding them because that's how they came from the employee store direct from the publisher, or the promo. It's basically a really shit version of the Mario 3 on NES with the big NFR sticker tacked onto it. It's JUST a regular Mario 3, but it has this big fuck off sticker on it, and to my knowledge graders don't ding it for that fact. 

[GPX | 1,369]

Another possible reason why we’re seeing more 10s is due to collectors having better eye and better understanding on what to send for grading. I can imagine in the first year or so, people new to the scene would be sending random sealed stuff just to trial the process and to see if they would continue with graded games. 

[inasuma | 1,203]

On 8/5/2023 at 10:59 PM, GPX said:

Another possible reason why we’re seeing more 10s is due to collectors having better eye and better understanding on what to send for grading. I can imagine in the first year or so, people new to the scene would be sending random sealed stuff just to trial the process and to see if they would continue with graded games. 

I would have the opposite impression actually. I figure the nicest stuff would get sent in at the beginning, but then that rate of high grades stays at a constant rate as people learn the grade scale over the years.

[Code Monkey | 2,010]

On 8/5/2023 at 9:59 PM, GPX said:

Another possible reason why we’re seeing more 10s is due to collectors having better eye and better understanding on what to send for grading. I can imagine in the first year or so, people new to the scene would be sending random sealed stuff just to trial the process and to see if they would continue with graded games. 

No doubt, check the reports for the number of cartridge only copies of The Flintstones: Rescue Of Dino And Hoppy graded. Obviously people that didn't realize there were 2 separate Flintstones games on the NES.

[GPX | 1,369]

On 8/9/2023 at 4:39 AM, inasuma said:

I would have the opposite impression actually. I figure the nicest stuff would get sent in at the beginning, but then that rate of high grades stays at a constant rate as people learn the grade scale over the years.

If you already have experience with graded games, it’s a lot different to those new to the grading scene. I remember when I first started with VGA I was quite anxious on how it would end up, if the package might get damaged or stuff gets lost etc.

After sending 1 or 2 batches, I started to feel more comfortable and would then send my rarer/higher end stuff. I also felt more comfortable in knowing what kind of condition I wanted and what to expect with the grading process.