I'm guessing there is something seriously wrong with gocollect15767
 I hear their hourly rate is outrageous!
|
sportshort private msg quote post Address this user | |
| first off I don't have the pay gocollect subscription just the free membership so I could not dig deeper. I looked up Spider-Man 75 and this is what I saw - Seems like something is wrong but gocollect also shows that 344 comics have been sold. what's up???  |
||
| Post 1 • IP flag post | ||
 Forum Crier
|
OGJackster private msg quote post Address this user | |
What do you want to see? |
||
| Post 2 • IP flag post | ||
|
Forum Crier
|
OGJackster private msg quote post Address this user | |
The other half... |
||
| Post 3 • IP flag post | ||
|
I hear their hourly rate is outrageous!
|
sportshort private msg quote post Address this user | |
| This book (from what I can tell) has to reason to be this expensive all of a sudden. the last time I checked a 6.0 was probably 70 dollars. what happened? why is it listed as this expensive? | ||
| Post 4 • IP flag post | ||
 " . "
|
Davethebrave private msg quote post Address this user | |
| I just noticed some odd artifacts. Their FMV model is messed up (always has been?) but when I checked on a couple books earlier today the FMVs were rEALLY mesSeD uP. | ||
| Post 5 • IP flag post | ||
|
Forum Crier
|
OGJackster private msg quote post Address this user | |
| "Death" of Silvermane John Romita Cover | ||
| Post 6 • IP flag post | ||
|
I hear their hourly rate is outrageous!
|
sportshort private msg quote post Address this user | |
| You can find a CGC graded 5.5 copy on ebay for under 220 dollars. way lower than gocollect's FMV of 725. | ||
| Post 7 • IP flag post | ||
|
I hear their hourly rate is outrageous!
|
sportshort private msg quote post Address this user | |
| @OGJackster, You'd think right? but you can buy it all day long (currently) for half or less than the fmv's your seeing on gocollect. Are there other sites that can confirm these crazy prices? | ||
| Post 8 • IP flag post | ||
|
Forum Crier
|
OGJackster private msg quote post Address this user | |
https://covrprice.com/ |
||
| Post 9 • IP flag post | ||
|
I hear their hourly rate is outrageous!
|
sportshort private msg quote post Address this user | |
| Now that's more like it. Thanks @OGJackster, As I said gocollect has lost their ever-loving-mind. | ||
| Post 10 • IP flag post | ||
 Where's his Bat-package?
|
Byrdibyrd private msg quote post Address this user | |
Looks like a CGC 9.8 copy just sold at Heritage a couple days ago for US $3,960!!! It's a documented sale, too. Dang!  Anyway, that would completely screw with any site that uses algorithms based on actual sales (that would be GoCollect) or even sites with values based solely off actual sales (that would be Covrprice & Nostomania [I think]). That is a CRAZY price someone paid for that book! CGC census shows 14 copies in 9.8, so that's not a whole lot, but surely it's not so rare to justify nearly $4,000. |
||
| Post 11 • IP flag post | ||
|
I hear their hourly rate is outrageous!
|
sportshort private msg quote post Address this user | |
| @Byrdibyrd I get where your coming from but if this is based on 344 sales, it should be real but I’m not a believer. | ||
| Post 12 • IP flag post | ||
|
Where's his Bat-package?
|
Byrdibyrd private msg quote post Address this user | |
| One huge outlier can skew the median and the averages, and depending on what sort of statistical model being used to base the algorithms off of (does it give a lower value for older sales, or do they retain the same value over time; do they give the highest value to the most recent sale, even if the most recent sale in a particular grade was years ago when the book was valued very differently, etc.), it can hose the entire thing, which is pretty much what we're seeing here. My guess would be that the algorithm tracks the most recent sales and uses those to assist in calculating values for grades that have few sales, no sales, or just sales that are very out-of-date. Throw in an insane number like 4,000 and everything from that point on is toast. | ||
| Post 13 • IP flag post | ||
 Collector
|
Stardust_Memories private msg quote post Address this user | |
| Yes, gocollect has a habit of inflating prices and manipulating the market. | ||
| Post 14 • IP flag post | ||
|
" . "
|
Davethebrave private msg quote post Address this user | |
Quote:Originally Posted by Stardust_Memories Nope. They often undervalue massively. In other words, their FMV tool is junk. Edit: not pure “junk” but certainly unreliable and should be used with CAUTION. It is a good starting point to help quickly get the underlying data and in aggregate across multiple books the data becomes more useful. Sometimes their FMV is disconnected on the high side and sometimes on the low side. A fair market value calculator should have a prioritization of data inputs and some sophistication behind it. Their system appears to have neither but instead a basic formula. I haven’t bothered to derive it but I can tell by the outputs generally that it is extremely “one size fits all.” |
||
| Post 15 • IP flag post | ||
|
" . "
|
Davethebrave private msg quote post Address this user | |
| @Byrdibyrd - their system has some type of normalization but it doesn’t appear to be appropriately weighted. I agree with your characterization. I focus more on SA and GA that transact infrequently. I have seen books that transact infrequently but with high visibility that GC undervalues in their FMV (the last 2 or 3 sales occurred at a much higher price point in a LOWER grade than their FMV shows in a higher grade) because they will retain transaction data from a long time ago in preference to recent data or in combination with recent data. That simply doesn’t work for low frequency / rare books. The result is their FMV is clearly much, much lower than true FMV. On the flip side they have highly transacted books (literally one or more sales per grade per day) that have outliers due to either ill informed bidders, auction manipulation?, unlisted variants etc, and those outliers (usually high) will have an impact on value. Simple example of the latter case is Eternals 1 with the cover price variant copies being lumped in and massively distorting the FMV calc in very high grades. At the end of the day they need a real valuation expert (someone good at math and stats) to fix their tool IF they cared about its quality. I don’t think they care and to be blunt “the market” generally doesn’t seem to care either. |
||
| Post 16 • IP flag post | ||
|
" . "
|
Davethebrave private msg quote post Address this user | |
| Oh, and I do think go-collect needs to make clear that their FMVs are not FMVs. I am not going to look now but I do not recall big disclaimers. Maybe they have them up. Their numbers are definitely not FMVs but rather gocollects calculated view of values. It does a disservice to the entire community to portray them as otherwise. Maybe worse than disservice. I have my own value model (not an FMV model) that used some of their data as input. The reliability is distorted a bit because of their FMV unreliability so I have discussed some of the above issues with them in the past. Because of how my own model works there isn’t tremendous reliance or need for perfect accuracy but it has meant that I have looked more closely at their data than I otherwise would. At the end of the day I appreciate their site as a nice summary or portal to other inputs but would always use HA and ebay listed sales data directly as well. |
||
| Post 17 • IP flag post | ||
 Collector
|
Jabberwookie private msg quote post Address this user | |
| @Davethebrave I’ve noticed GoCollect has some unusual values, but like my news, I don’t get my values from one source. Do others go “one and done” that often? |
||
| Post 18 • IP flag post | ||
|
" . "
|
Davethebrave private msg quote post Address this user | |
Quote:Originally Posted by Jabberwookie Great point. Personally for my news I use Reuters. Most of what news outlets now provide is editorialization. For GC I think the danger is threefold: 1) often the gocollect fmv are indeed spot on 2) they specifically use real data as inputs and that data is easy to audit and 3) they call it fmv. As a single source goes, in theory it could be very reliable. When working on a trading desk you can see recent transactions through your Bloomberg terminal. Anyone would feel comfortable relying on that. One step “closer” is you can also use its options pricing calculator as it stands or tweaked with some adjusted assumptions. You can still generally rely on what you are looking at (the latter is formulaic vs equity prices being purely reference transactions). GC calls their outputs “FMV” but FMV despite the word “fair” is generally considered what the current market will bear. They are instead overriding with formulas in ways that distort vs any market value. The most visible prices in the market today are asking prices. Horrible data point overall. GC is a useful repository and if it worked properly could be a quick single source. Graded comics are relatively easy to value (for market values). It is a shame GC doesn’t work properly. |
||
| Post 19 • IP flag post | ||
|
Where's his Bat-package?
|
Byrdibyrd private msg quote post Address this user | |
| @Davethebrave I very much agree with you that what GC claims are their FMV's are not FMV's but just an interpretation of them calculated via whatever program they are using to compile data. Not surprisingly, the more data points they have (as in the more items sold), the more 'accurate' their estimate of the real FMV. Naturally, if one of the items sold winds up being an outlier, then that's a spanner in the works and the whole thing becomes invalid. It also gets skewed when things like price/cover/error/etc. variants are not taken into account and get lumped in with everything else. I do notice that they seem to be trying to fix that with separating out newsstand editions and some variants, but they can only do that if the auction/sale item's entry title contained words like 'newsstand' or the accurate name of a particular variant. It's a real can of worms. | ||
| Post 20 • IP flag post | ||
|
I hear their hourly rate is outrageous!
|
sportshort private msg quote post Address this user | |
| Aren't GC's FMV's based on sales (except of course where there arent' any, like some grades that have no sales). Isn't an FMV (example) 3 of the same books in the same grade, one sells for 100 one, for 200 and one for 300 so the FMV is 200? (100+200+300=600 then divide by the 3 books = 200) I always assumed this was the case. I understand this does not work well with books with very few sales but it would seem like the best scenario. My question is, what (if anything) went wrong with their valuation of Spider-man 75? | ||
| Post 21 • IP flag post | ||
|
" . "
|
Davethebrave private msg quote post Address this user | |
| @sportshort well yes and no as they interpolate a lot and will override sales data. Your example is good and I will expand a bit but either way it is very busted (their model is). They interpolate in the absence of data. They also appear to do other calcs (some nonsensical) even as new fresh data is added. I know this because I have seen some live impacts (pre/post transaction) that showed a decrease for a FMV category following a documented (public) sale that transacted at a higher-than-prior-“FMV” level. So a real input to their model, at a higher than their previously calc’d FMV dropped the price lower. Of course I can totally see how a poorly constructed algorithm can result in this happening but it is absolutely illogical. For example, if they retained and calcd based on last 3 trxns and those trxns went back 3 years at $300, 100, 200 they might calc an FMV of $200. Now a new trade happens at $250… a poorly constructed model taking a 3 year (datapoint here) trailing average would show a drop in FMV from $200 to about $185. Bad logic all around in the above example. Their pre-trxn FMV would have been overstated vs FMV and their post-trxn understated AND the trend movement would be wrong. My guess is something like the above was done. Likely by someone who thinks they are doing something logical but are not. Then the model is doing the interpolations between and across grades blah blah blah. Same lack of true expertise… |
||
| Post 22 • IP flag post | ||
 Why just the women? I like bears.
|
Gaard private msg quote post Address this user | |
| One sale in the last 7 years. That was $100 in 2017. What would the (GC's) FMV be? | ||
| Post 23 • IP flag post | ||
|
" . "
|
Davethebrave private msg quote post Address this user | |
Quote:Originally Posted by Gaard $1 million? |
||
| Post 24 • IP flag post | ||
|
I hear their hourly rate is outrageous!
|
sportshort private msg quote post Address this user | |
| @Gaard, I would think there's not enough data to support the posting of any fMV. | ||
| Post 25 • IP flag post | ||
|
Why just the women? I like bears.
|
Gaard private msg quote post Address this user | |
Just curious how they come up with FMVs. Is it posted somewhere? |
||
| Post 26 • IP flag post | ||
|
Where's his Bat-package?
|
Byrdibyrd private msg quote post Address this user | |
Quote:Originally Posted by sportshort Yeah, and that's where the questionable math comes in to calculate an FMV from too few data points. Not gonna lie, I use GoCollect, but I use them to see the lists of documented sales (very useful) compiled over a relatively long period of time. Because eBay and other auction/sales sites don't display sales after a few months go by, GC is one of the few places where that data is archived, and that's what makes GC useful to me - not their odd conclusions about FMVs. |
||
| Post 27 • IP flag post | ||
|
I hear their hourly rate is outrageous!
|
sportshort private msg quote post Address this user | |
| @Gaard, Actually that one makes sense to me. so few are out there (I think) that any sale of any kind that got recorded would probably need to be used to extrapolate the fmv's for other grades. Since there are only 97 total books graded they probably don't change recorded hands enough to give a true fmv | ||
| Post 28 • IP flag post | ||
 I live in RI and Rhode Islanders eat chili with beans.
|
esaravo private msg quote post Address this user | |
Quote:Originally Posted by Byrdibyrd This is exactly why I have had a Go Collect subscription for the last 2 years. Without the subscription, you have to take their values with a large grain of salt. Some are pretty accurate, and others are absurd. With the subscription, you can see the sales, and not just from eBay - you get Comic Link, Heritage, and others too. So I find that useful when purchasing or selling a comic, especially something that doesn't have lots of sales. However, I am getting tired of waiting for them to link up with the CBCS census. You can enter a CGC code and it will fill in the grade, page quality, etc. It will also track values for your list of CGC books. These features do not work with CBCS slabs. If something doesn't change by the next time my subscription is due, I will probably drop them. |
||
| Post 29 • IP flag post | ||
|
Where's his Bat-package?
|
Byrdibyrd private msg quote post Address this user | |
Quote:Originally Posted by sportshort That's it exactly. I may not be the biggest math expert out there, but I still took statistics/sadistics, and even a little study there will tell you that accurate maths require A Lot of data points. Even if all 97 books out there were to have sold at one time or another, it wouldn't be enough data to estimate an accurate FMV. |
||
| Post 30 • IP flag post | ||
This topic is archived. Start new topic?