wytshus

I was thinking just a simple multiplier based on completion percentage. However, not all sets are created equally, like you said. Standardization of sets would be needed to make it fair, and that would be a massive undertaking. In a perfect world, that would be totally reasonable. The problem is the registry does not represent the market. The score weights are based on CGC census data. Basing raw scores on market value was the best choice at the time the registry was created. I don't think anyone could have predicted the size of the registry as it sits today.

During the award selection process, we implement a freeze on changes to the sets. During this freeze, we simply stop modifying the registry. It has been stated earlier that the goalposts keep moving due to slot additions, set creations, and score adjustments. I agree with this completely, and my boss suggested extending the freeze time to eliminate this issue.

Keep in mind that the algorithm is SUPPOSED to weigh individual scores according to grading category, desirability, and scarcity. The raw scores would be consistent for Universal across the board. This will eliminate the outrageous point spreads in sets with high value keys. It will also give photos, descriptions, and completion percentage a bigger role in the competition, and the awards. I really like the idea of a completion modifier, and I will let you guys know how feasible that is. I would also like a modifier for photos and descriptions, but I don't have write access to the database, and I am NOT a DBA. I would have to make a request for the code change, and justify the labor time..... I don't expect this to please everyone. That is impossible. Something has to give on scoring, the status quo is not sustainable...

Yes, this would be the optimal solution, but I don't see it happening. I am going to push for a base score across the board for 9.0, maybe 9.2, not sure yet. This will resolve a number of issues with scoring. It will also make completion percentage, photos, and descriptions more important than the total score. This is the way it should be, in my opinion. If it is determined that we need another tier score for keys, we can look at ways to implement that. This would be difficult, due to the research time, and the criteria for what truly makes an individual book a "key". Keep in mind that this is for next year's "season", no changes will be made until after next year's awards are announced. We are also discussing a longer "freeze" time for the registry in order to limit the goalposts being moved. After that, I am going to go over the individual sets, and attempt to define a standard competitive set.

The web team reports that there are server issues on the back end, should be temporary, but there is no ETA.

I have informed the web team, I will let you all know when I hear back.

Let me know if you are still having this issue, please post screenshots with error codes if possible.

The main problem is that prices for keys do not stay static. Also, we are talking about open market prices, something that CGC shouldn't be involved in.. In the beginning, for most non key books, the 9.0 Universal score was the minimum cover price, before 2015, which was set at 3 After 2015 or so, the minimum price is $3.99 (4). Keys and vintage comics were scored according to Overstreet 9.2. The algorithm is supposed to weigh individual books according to scarcity, desirability, and grading category. However, it gets this data from the Census, (i.e. books that have been graded by CGC) So you get a weighted score for books within the Census, not in the open market. Which is why we say that scores do not equate to market dollars. (Even though we do get part of that data from the market...) o_0 Currently, there is simply too much volatility and FOMO hysteria to continually update individual scores every time there's a spike, it's not worth the time, and it raises a lot of questions about impartiality.

This is the way of things when it comes to the Registry. I have been working with the Collector Society website for about a year and a half now, and I still find it.....lets say.....clunky. I can just imagine that a new user would find it a bit overwhelming. I have been exclusively approaching the Registry as an Admin, modifying sets, adding slots, modifying scores, and resolving add fails. Now that I have my feet underneath me somewhat, I am starting to look at things from a collector standpoint (whats fair, fun, and interesting), and from a marketing perspective (how can we get new users, how can we make the competition more transparent) Let me state for the record that the Registry is NOT the Census. I do not have read/write access to the Census, the Registry gets its individual book data from there, and I input Universal Scores on the Registry side. Once that's completed, its out of my hands. What I am trying to do, and the reason I created this thread, is to get input, instead of springing drastic changes on everyone arbitrarily, and expect everyone to like it. Most of the comments in this thread have crossed my mind at one time or another, especially the comment about competitive sets being modified ad hoc. This is one of my biggest gripes about the Registry when it comes to the Awards, and I intend to do something about it next "season". I am heavily in favor of a base score for ALL books. It has been mentioned, and I totally agree that CGC should not be in the business of declaring FMV, especially since the modifiers for weighted scores only apply to books that have been registered BY CGC. Having one score for all books will level the playing field, and the algorithm can weigh scores for individual books. If there are tweaks to be made, we can do that in the "off-season". I am also leaning heavily towards fixed sets for award consideration. If a new slot absolutely must be added, it will be non competitive until the new season starts. But for the most part, the scores and slots do not change. I do really appreciate all the feedback, keep it coming. I read all the comments(and re-read the long ones).

@Qalyar No problem, I welcome all of the input. I know it seems like I am all over the place, but it all ties in together. I'm not ignoring what has been said, I am processing, and trying to figure out the pros and cons from an admin/marketing/collector point of view.(what's fair, what's possible, and what's efficient) I talked to a colleague in the Trading Card registry last week, they don't have the "spread" of grade points like comics, but they do have a fixed score for a particular grade, and then individual cards receive a weighted score from that source. This is where I would like to see comics go. If it's determined that keys need to be given a higher score, we will do that, once a year. With the comics registry, once a slot is created, the score applies to that slot, regardless of which set(s) it's in. This is why I think a base score for 9.0 Universal(Maybe 9.2) is the way to go. Let the algorithm determine the final score for individual books, and have fixed sets for the awards.

It would be a LOT of work on the front end, but long haul, it would make my life much easier. Plus, I wouldn't have to request any fancy coding, I could do it by hand.

Another random thought: "Award Eligible" Tag We designate certain sets eligible for award consideration, sets not designated are still competitive. Once a set is designated as eligible, the set list will not be modified. Sets that are not eligible can become so in the future, if it's fleshed out enough...

@Rosland Yes, the source value is Universal 9.0, *for the most part. We are still talking over 2500 possible scores for just 1 book. I used ASM 15 as an example. Not all titles in the pop report have slots in the registry. I don't know if 49 is the max number of categories, or the categories that have been graded. It was just a random thought that pops into my head from time to time...

Not that I'm aware of. if I implied that, it wasn't my intention. Just a quick thought: There are 297,985 separate titles in the population report 1st Print ASM #15 alone has 49 different grading categories There are 25 different "grade points" Just think of how many different possible combinations there are! It's mind blowing and soul crushing at the same time...lol

@MAR1979

@Sauce Dog The minimum requirement for set creation is for NEW sets. I really don't want to delete existing sets or slots. My intention is not to get rid of entire sets/series. I would very much like to eliminate duplicate sets, where ever possible. Take Amazing Spider-Man for an example. There are 33 separate competitive sets for ASM alone. There is a Complete Set, a Complete with Variants set, a set for 1st appearances, issue specific sets, issue run sets, foreign issue specific sets, etc. Now, I understand this, because the title has so many individual issues, across 40 years. Once you include variants, it becomes impossible to put them in one set and expect someone to collect them all. And this is just for one title. Once you take into consideration other popular titles, you can see the problem. I have never seen the value in having a Complete set, and a Complete with Variants Set for the same title. But, like I said, I am not going to make any drastic changes until after next year's awards. I am just looking for input from the community.

I completely agree with this. Although, I would still like to have 1st Print only sets, but that's just a personal preference. To be perfectly honest, I do not like the (Complete) sets at all. I prefer to have individual slots for each variant, it soothes my OCD...heh. The main problem, and the reason I started this thread, is individual issue scoring. The other issues tie into this. On the registry side, when I create a slot, I input the Universal score for a 9.0 Grade(This is usually the minimum cover price). The table in that slot then calculates Universal Scores for the other grades. Once that is done, there is an algorithm on the back end that is supposed to adjust for scarcity, desirability, and value across all other grading categories and their individual grades. I believe it gets those variables from the census. The 2 main issues are: There is no automated system in place to update the Universal Scores. There seems to be an issue with the algorithm, it is not always calculating scores correctly.

Not ALL variants, just foreign. Sorry

A few thoughts on set standardization: After scoring, this is the biggest concern I have. There are too many competitive sets in my opinion. And there are a lot of sets that are too small, or too obscure, to ever qualify for an award. On the other hand, I think if someone takes the time and money to get a book graded and slabbed, they absolutely should be able to add it to the registry. I have tried to mitigate the bloat by requiring a minimum of 3 first prints to qualify for a set. This helps a little, but there are still too many sets that have less than 10 slots. To cover ground already addressed by @Qalyar Again, with Golden Age, this isn't really an issue. The problem starts with Silver Age and beyond. There are some Silver Keys that have slots in way too many sets. This is understandable, because those keys may have first appearances, crossover between different series, etc. Then you have multi packs, newsstands, one shots, and foreign/price variants. It only gets worse with Modern, due to the variant cover insanity... The competitive sets really need to be cleaned up and streamlined, but it's hard to know where to begin, honestly. I don't want to make any drastic changes until after next year's awards. One thing I'm thinking about is just making all FOREIGN issues/variants non competitive. (I never want to delete existing slots/sets, but it might be necessary in the future.) I am also going to ask for a "Series Year" Field in the cert lookup. Another idea I've been kicking around is new award categories. 1st Print only, Artist Centric, and a People's Choice Award, where we let the community vote on a winner. To be clear, these are just random thoughts that bounce around in my head, nothing is set in stone. More to come. (Edited for clarity)

Just a thought that I've been kicking around. I think it would bring some sanity to the scoring system. I could always go back and set Keys higher, then update twice a year. There's just so many variables, and so much volatility. Golden Age isn't so bad, they don't move much, and are stable for the most part. It's the Silver Age Keys that are giving me fits. Every time there's a new movie or T.V. show, the spec market freaks out. I'm trying to strike a balance between what's fair, and what's actually possible. I have been scoring Universal 9.0s for Modern at 4 for a while now. It cuts down on research time.

A few thoughts: I'm considering making all 9.0 Universal Scores 4 points, the minimum cover price. It used to be $2.99 (3), but after 2015 or so, it's $3.99. The problem with this is there are some modern 1st issues that go for $4.99, and others that go for higher. But again, points do not equal market dollars.....ugh. Not sure if this can be done globally in any reasonable amount of time. Or if it can be done without crashing the system. It would probably have to be done in increments. I have a dog and pony show with the new CTO Friday, and I will run this past him and my boss. I want the algorithm to do the heavy lifting when it comes to points. If I can deal with a standard Universal Score across the board, it would take a lot of weight off my shoulders. As far as completion % goes, we do see that on our end. It's the deciding factor when we choose award winners, along with descriptions and pics. We usually start with the top 10 or 20 sets, generated by points, and the above criteria. Then we look at if the owner and/or set has won in the recent past. Another consideration is how often, and how recent changes were made. We also look for originality, and the passion that comes through in the descriptions. The system only makes suggestions, my colleagues and I pick the ultimate winners. If you take a look at this year's winners, you will see what I'm talking about. A 100% completion badge or indicator is a great idea, I just don't know if it's feasible coding wise. Again, I will run this past the powers that be. More to come. As for standardization of the sets.... There have been a lot of hands in this soup...heh.

@Iconic1s I had no input on this decision. I was told of this change only after it was announced. I haven't seen any requests for newsstand slots, and it's still too early to tell how much interest there is going to be in relabeling/re-slabbing existing newsstand editions. There are plenty of sets with newsstand variant slots, and to be honest, variants don't impress me at all. Newsstands impress me even less. Personally, I think they dilute the value of the 1st Prints. I find it ridiculous to have 50+ variant covers for 1 issue, Please keep in mind that points in the registry are for entertainment purposes only. We took great care in this year's awards to avoid picking winners based on how much many points(i.e. money) they have.

First thing we need to do is get a handle on the algorithm. I have discovered several anomalies where scores don't match up with other parts of the database. I input Universal Scores, and the algorithm is supposed to adjust for other categories. This isn't happening correctly in some instances, and it really concerns me. I am going to work with our new CTO to see if we can identify where the problem lies, and what we can to to resolve. After that, I am seriously considering creating "Invisible" sets for the Silver, Copper, Bronze, and Modern keys. Probably max 100 books in each set. The scores in these sets will be updated at designated times, probably twice a year. This will take a lot of time and research, but once completed, it will take a load off my mind. More to come.

@ADAMANTIUM As promised, I would like to open a discussion about the Registry Scoring System. Let me start off by saying this: Registry scores do not equate to market dollars. I understand that this is a competition, and you can't have a competition without points. But, as far as the Registry Awards go, we do not pick the award winners based on points. If you look at this year's award winners, you will see that we gave awards to sets that have close to 100% completion, lots of photos, and descriptions that are not cut and pasted from other sites. We put in a lot of time and effort into choosing the award winners, and I am very proud of the job we did this year. I would ask that everyone that is concerned about registry scoring read this thread: You will notice that the thread is over 15 years old, when the registry was a fraction of the size it is now. As it currently stands, we are approaching 600,000 books,, with over 9500 competitive sets. There is no automated system in place to update individual scores. This is why the score update thread was created. However, due to inflation, market volatility, FOMO hysteria, and anomalies in the scoring algorithm, I have locked that thread. I will fulfill the existing requests in the thread when I have the time. It simply is not feasible to adjust the scores every time there is a price spike for 1 book. I have more thoughts, but I wanted to get this thread going and get some input from you all. Thanks!