Current Spider-Man Sales

[DT Winslow]

MisterMets answered your question in his initial post here:

It just means that does not be fully exhibit of all the data in the 3000 stores across the United States, it merely is a small idea of what they could be.

Remember, there are two types of people in this world:

1. Those who can extrapolate from incomplete information

[CaptainUniverse]

Remember, there are two types of people in this world:

1. Those who can extrapolate from incomplete information

2. Those who cannot?

[Rift]

Remember, there are two types of people in this world:

1. Those who can extrapolate from incomplete information

2. Those who recognize that Piper is best ASM girl.

[bob.schoonover]

ICv2 interrogates 125 stores, but it's actually measuring (approximately) the behavior of all of the customers of those stores. Which is to say, the sample isn't 125 out of 3000, it's all of the customers at 125 stores compared to all of the customers total. If you think each store has only 20 customers (on average, which is incredibly low), that gives you 2% error at 95% CI. It drops below 1% if the stores all have around 100 comics customers. The only argument against ICv2 data not being a fair reflection of reality would be that the selection of stores somehow biases the data irrevocably - but their statement "... do represent a variety of locations and store emphases" certainly cuts against that argument.

Now, since the data ICv2 displays is just the ranking, that allows for specific pathological outcomes that *could* happen but wouldn't be consistent with all of the pre-order data we have from many years from Diamond. That is, a 1% error might be enough to mis-identify the #1 book by one or two places in the ranking, but only in a month where there were no big event books. Our years of Diamond data strongly suggest that a 1% error in sales data on either side is not enough for the book ranked #1 to actually be 8th, or 20th, or "boy these sales suck" territory.

[Spider-Tiger]

MisterMets answered your question in his initial post here:

It just means that does not be fully exhibit of all the data in the 3000 stores across the United States, it merely is a small idea of what they could be.

Sort of. Basically, there are certain steps that could be taken to ensure that a subset population used for data collection is representative of a larger population under examination (such as randomization to ensure that there was no bias during the selection process.) Not necessarily size itself. The data presented here is from a non-random selected subset of stores that were not necessarily designed to be a representative sample, and so the data itself only reflects the customer buying habits from these particular stores.

(source: student who has performed research and is routinely tested on this bs for standardized exams.)

[blank]

Remember, there are two types of people in this world:

1. Those who can extrapolate from incomplete information

I don't get what you are trying to say that because I said its a small idea of what it could be that I can't extrapolate data from incomplete information? Because if so you are being weirdly rude about an a poor explanation, in a population of over 3000 stores globally(I wrote in the US earlier but thats wrong re-read how they conduct it) using data from a non-random(high possibility of bias) sample of 125 stores(which are most likely all in the US) is merely a small(it represents less than 5% of the population size of 3000) idea of what it could be globally and while its completely fine to extrapolate based on that but it still remains a only possibility and not a certainty, its not exactly. Fun little note it would be remarkably easy to skew the data in this sample if even one person were to find which exact stores they use to create said data.

Sort of. Basically, there are certain steps that could be taken to ensure that a subset population used for data collection is representative of a larger population under examination (such as randomization to ensure that there was no bias during the selection process.) Not necessarily size itself. The data presented here is from a non-random selected subset of stores that were not necessarily designed to be a representative sample, and so the data itself only reflects the customer buying habits from these particular stores.

(source: student who has performed research and is routinely tested on this bs for standardized exams.)

Thanks for the correction, I'm not all that well versed on the exact definitions of terminology for statistics.

[DT Winslow]

I don't get what you are trying to say that because I said its a small idea of what it could be that I can't extrapolate data from incomplete information? Because if so you are being weirdly rude about an a poor explanation, in a population of over 3000 stores globally(I wrote in the US earlier but thats wrong re-read how they conduct it) using data from a non-random(high possibility of bias) sample of 125 stores(which are most likely all in the US) is merely a small(it represents less than 5% of the population size of 3000) idea of what it could be globally and while its completely fine to extrapolate based on that but it still remains a only possibility and not a certainty, its not exactly. Fun little note it would be remarkably easy to skew the data in this sample if even one person were to find which exact stores they use to create said data.


Thanks for the correction, I'm not all that well versed on the exact definitions of terminology for statistics.

So, number two, then.

[Xenon]

All I can say is that to the extent these are accurrate people need to do a better job and not bad bad comics. Hopefully people were just waiting for answers and now that they're done they'll drop off.

ICv2 interrogates 125 stores, but it's actually measuring (approximately) the behavior of all of the customers of those stores. Which is to say, the sample isn't 125 out of 3000, it's all of the customers at 125 stores compared to all of the customers total. If you think each store has only 20 customers (on average, which is incredibly low), that gives you 2% error at 95% CI. It drops below 1% if the stores all have around 100 comics customers. The only argument against ICv2 data not being a fair reflection of reality would be that the selection of stores somehow biases the data irrevocably - but their statement "... do represent a variety of locations and store emphases" certainly cuts against that argument.

Now, since the data ICv2 displays is just the ranking, that allows for specific pathological outcomes that *could* happen but wouldn't be consistent with all of the pre-order data we have from many years from Diamond. That is, a 1% error might be enough to mis-identify the #1 book by one or two places in the ranking, but only in a month where there were no big event books. Our years of Diamond data strongly suggest that a 1% error in sales data on either side is not enough for the book ranked #1 to actually be 8th, or 20th, or "boy these sales suck" territory.

That said, non-randomized and static mean the results are of limited value. It really only reflects the buying habits of the custoemrs that go to these particular 125 stores. They might be in different places with different emphasises, but as it's the same locations there's tons o missing data it could not be accounting for. Additionally, beacuase it's being tracked through this one specific system, there may also be an effect that this system is having on the results. ComicHub users may be more static than non-hub users for instance Or more prone to buy one company over another. Or maybe not. But it's impossible to tell without randomized sampling.

[Spider-Tiger]

Thanks for the correction, I'm not all that well versed on the exact definitions of terminology for statistics.

No problem! just thought I'd help clarify for the person that asked

[Celgress]

My Two Cents (for what little it is worth)

Oh, well, "Reality" TV shows are very popular too but they are trash-tier in terms of quality. I suspect much of these sales are from ride-or-die buyers who buy the comic book out of habit more than anything else; the same principle applies to Batman. 616 Spidey no longer holds my interest. But to each their own.

[Venomsaurus]

I know nothing of sales and how accurate they are or are not on the data. All I can say is, my town is fairly 'nerdy' with a few comic shops. I regularly visit them all, and I can say every shop always has a massive over-abundance of ASM no one is buying. So I can't help but feel the "Marvel inflating stuff with variants" theory is more on point. Do I know it for sure? Nah, not really, and the fact is that not anyone here knows the hard data for sure. We can argue day or night on info we don't have. All I can say is what I see with my own eyes, and this run doesn't seem to be selling well wherever I go.

[Celgress]

I know nothing of sales and how accurate they are or are not on the data. All I can say is, my town is fairly 'nerdy' with a few comic shops. I regularly visit them all, and I can say every shop always has a massive over-abundance of ASM no one is buying. So I can't help but feel the "Marvel inflating stuff with variants" theory is more on point. Do I know it for sure? Nah, not really, and the fact is that not anyone here knows the hard data for sure. We can argue day or night on info we don't have. All I can say is what I see with my own eyes, and this run doesn't seem to be selling well wherever I go.

I do wonder how many of these "sales" are actual customer sales versus comic book store sales. How many of these books are simply gathering dust in stores? I often wonder about this in general with American Comic Books given the direct sales business model.

Edit - The sales numbers have been steadily declining for years now but are they actually far lower than we are led to believe? Given the increasingly sad state of the comic book store part of the industry, I strongly suspect there is something very rotten in Denmark as they say.

[Venomsaurus]

I do wonder how many of these "sales" are actual customer sales versus comic book store sales. How many of these books are simply gathering dust in stores? I often wonder about this in general with American Comic Books given the direct sales business model.

THAT'S the data I want to see. However many shops are ordering does not equate sales whatsoever.

[blank]

So, number two, then.

I think I need to clarify my position, while I freely admit stats aren't exactly my strong suit, its not wrong to say that there still exists the small possibility that the extrapolating the ranking to apply on a global scale could be lead to an incorrect estimate of the standings, and that the only certainty with the data provided is that it at least provides a extremely accurate image of sales in these 125 stores(and when extrapolated it only provides a extremely blurry image what of sales may be in the wider 3000 stores).
However, I do actually think the rankings are a pretty decent indication(but they are and never will be perfect) of how well the books are actually selling but there obviously will be the potential for positional differences if we were to have more data available.
Nothing I wrote was against the idea of extrapolation or indicative of whether I am able to estimate the possible standings based on incomplete information, you clearly have some weirdly arrogant belief that you know everything about the capabilities of a person based off two rather haphazardly written forum replies.

I may dislike the current run but I do think #28 placing so low has much more to do with people in the USA going away by taking 4-5 day weekends by either taking Monday or Monday and Friday off and the 3 day sale period(although #26 proved that isn't too huge a hurdle), than it has to do with a drop off in interest but we'll find out next month.

[Mister Mets]

THAT'S the data I want to see. However many shops are ordering does not equate sales whatsoever.

This data is based on sales and not pre-orders.

I know nothing of sales and how accurate they are or are not on the data. All I can say is, my town is fairly 'nerdy' with a few comic shops. I regularly visit them all, and I can say every shop always has a massive over-abundance of ASM no one is buying. So I can't help but feel the "Marvel inflating stuff with variants" theory is more on point. Do I know it for sure? Nah, not really, and the fact is that not anyone here knows the hard data for sure. We can argue day or night on info we don't have. All I can say is what I see with my own eyes, and this run doesn't seem to be selling well wherever I go.

It makes sense to have more copies available of a best-seller. If a retailer buys 50 copies of Amazing Spider-Man and 5 copies of Savage Dragon (picking a random title that doesn't sell as well), and sells 90 percent of inventory, they'll consistently have 5 copies of Amazing Spider-Man left and every now and then one issue of Savage Dragon.

I think using non-representative sales data doesn't work. You could argue any of those possibilities and nobody could say whether it's true or not.

We're always dealing with uncertainty.

These figures would really need to be off for the title to be a flop. It's possible but it does seem unlikely.