Are you a Quiet Speculation member?
If not, now is a perfect time to join up! Our powerful tools, breaking-news analysis, and exclusive Discord channel will make sure you stay up to date and ahead of the curve.
If you've been reading my articles, you'll see that I’ve taken in a lot of data, smashed it, shook it, baked it in Excel, and it’s given us some perspective on the prices and demand of various cards. There is still a lot more to learn, and I hope, more data to be had. There is another very valid way to evaluate cards. Buy Lists are perhaps an even truer value of the worth of a card versus the "price" or "cost".
Before we go further, I should make a note about that last sentence. To many of you, it might seem as though I used a thesaurus to spit out the pricing equivalent of “me, myself, and I” but there are some pretty distinct differences in my mind. Though the term “cost” can get pretty overloaded with meaning. Here is how I (with the aid of neither Merriam nor Webster) would define each as in relation to the topics herein.
As with the ongoing development of the Demand Index values these Worth Matrix values are going to take some time to work out, and more data is going to be needed in order to lock them down to something we can really rely on.
The concept of the Worth Matrix is similar to the Demand Matrix in that it should give us some indication of the real worth of cards in an ever changing landscape. A few of the major differences are that we can establish a Demand Index value for any card in a given subset of cards, but in the Worth Matrix, we can only highlight the cards that are specifically in demand by the indexed sellers or the changes in those lists over time. Some of the cards that are on buy lists are there because their inventory has fun low, or because the cards are forecasted to become hot in the future. I would imagine that many of the cards are only needed regionally, and that different stores might have different needs. Some of these irregularities can be overcome by aggregating more information, some by looking at the values over time. As with the Demand Index values, more data will mean more accuracy. Another difference is that these results are going to be more directly related to their price. The Demand Matrix looks at the relative interest a card is generating and I hope that the proportional values extracted from the Worth Matrix might give us a baseline for establishing and predicting the pricing of certain cards.
This week I’ll show you some of the results from aggregating and sorting a few of the more accessible online buy lists. The lists currently being crunched are Star City Games, Cool Stuff Inc, Troll and Toad, BlackBorder and AdventuresOn. I’ll probably be adding more (please list out your favorites below) but am somewhat limited as to which I can automatically pull data from, so please bear with me. As I add and compile more and more buy lists and cross these values with sales history and pricing fluctuations, I think that not only will we be left with another potentially useful value (Worth Matrix), but we will also see stronger relational information between all the values.
Below we have some of the more straight forward information. These three lists are sorted according to the highest average buy price as listed in the above online stores published buy lists.
[iframe https://spreadsheets.google.com/pub?key=0AukC5EWiTvOpdE1fQjE4bFdmMHJ5ZE84d2tNbjhsN2c&single=true&gid=3&output=html&widget=true 100% 600px]
This list seems about right with the possible noted absence of Jace, The Mind Sculptor.
[iframe https://spreadsheets.google.com/pub?key=0AukC5EWiTvOpdE1fQjE4bFdmMHJ5ZE84d2tNbjhsN2c&single=true&gid=4&output=html&widget=true 100% 600px]
Star City Games and Cool Stuff Inc don’t seem to buy a whole lot that is strictly in the Extended block at this time.
[iframe https://spreadsheets.google.com/pub?key=0AukC5EWiTvOpdE1fQjE4bFdmMHJ5ZE84d2tNbjhsN2c&single=true&gid=5&output=html&widget=true 100% 600px]
The Legacy and Vintage list is mostly useless in this format but fun to look at. With Standard and Extended I can get down to the $1.50 card mark in about the top 50. For Legacy it’s almost the top 750 to get down to $1.50 and a list that large isn’t as easy to produce due to some of the limitations of Google Docs. I’ll come up with a better way to display the Legacy values soon though.
For a bit of a taste as to what the Demand Index values can do for us, the following list is a comparison between the cards that are specifically sought by the stores but whom have the lowest demand. In other words, the stores may want them more then you do, so now is a good time to sell them either on their own, or add them into the shipment you have going out.
[iframe https://spreadsheets.google.com/pub?key=0AukC5EWiTvOpdE1fQjE4bFdmMHJ5ZE84d2tNbjhsN2c&single=true&gid=2&output=html&widget=true 100% 600px]
As you can see some of the stores are buying some cards that might seem a bit surprising. These are all cards that have virtually no demand (Demand Index values) behind them, so they should be either lying about in your sorting boxes or perhaps the stores know something we don’t. I’m anxious to add Channel Fireball to the list of stores searched since they do so much speculative pricing and buying. As of right now their online buy list isn’t practical enough to access, but I hope that gets better in time. This list will get better as I refine the sorting parameters. For example, right now I’m only looking at cards that are at least $1 on average, but other thresholds and lists (perhaps a $5 min?) are something I’m considering.
As always, please comment as to what other information and results you’d like to see.
I'm interested in how you're automatically pulling buylists. I cooked up a script a while ago that scrapes websites for buylist data and spits it out to me every morning so I know what to trade for. I currently have a few local stores, BlackBorder, CoolStuffInc, TrollandToad, and CardKingdom parsing correctly, but I'm having a pain with parsing SCG's layout. Any pointers? (I haven't bothered with parsing ChannelFireball because their buylist is almost always worse than the other buylists)
May I call you Windle? Ok good.
@Windle
I'm still learning Python so that I'll be able to ditch the current method, but right now it's all done entirely in MS Excel. My main spread sheet for this data is about 65MBs and takes about an hour to update. That's not me entering data, that's just for formulas crunching away after the new data has been pulled down.
SGC's layout was a bit of an issue, but since the information stays in the same place (vertically) I just sorted the lookup tables into 3 separate series of columns, checked for the set names based on the position of the word "Card," ran the set names through a lookup table to convert them to something usable and filtered out the ones I don't want to check (like foils or promo card).
I can't wait to learn how to use something with more power.
@Windle
sounds good. I know a few people that are python coders, but someone with code doing exactly some of what i'm looking to do would be good. BTW, pulling down the data is pretty quick, maybe a minute or two to grab around 178,000 lines of text, it's the sorting that takes all the real time. I think though that the vlookup and index functions are linear so the time is not at all efficient. There is a lot to parse and filter. I could probably rewrite it to be more economical, but thinking of doing so makes me want to spend the time on Python.
Great series and article!
I'm trying to get better at creating spreadsheets and analysis like this, but I've got alot to learn. I'm way behind Michael Poon – I've got no idea how to automatically pull info from a website. 🙂
@D-Roy
Nearly everything I learned about Excel was from the help files. Took a while, but I got there. One thing that makes this kind of work a little easier is the Index/Match style lookups.
Spent a good deal of time yesterday trying to learn some of this advanced excel stuff. I figured out how to get the external web data to the spreadsheet & auto updating, but I've hit a wall as to using the data how I want to/ using formulas in my spread sheet. I never thought getting into Magic would involve learning Excel. 🙂
This is going to be a challenge for me. I think my head might asplode. Time to check out Excel 2007 books from the library!!
This was an extremely enjoyable and timely article. I just sat down to sort through a few boxes of old cards with the intention to sell them to Troll and Toad's buy list. Your article made me realize I need to do some more data analysis to identify which buy list is best for a particular card. The damand index buy list is extremely useful! I would like to know more about how you are pulling the information from the buylists. I think I have a pretty good idea and can see the inconvenience of Channel Fireball's list. Did you seek permission from the respective sites to pull this information? Thanks for the article, keep it up!
@coveredinbees
I didn't ask permission since they are online buy lists and they probably don't feel too bad about someone sending business their way… As for pulling down the data, it's all Excel.
Regarding permission – If a site doens't want the free traffic, pubicity, and isn't interested in the transparency of competition that something like this creates, I would question the motives of their buy list. It's just a standing offer to a free market. Shouldn't be an issue.
Thank you, I was just curious as I began thinking about programmatically accessing this data. I have been looking for a pet development project and would like to give this a try since I am actively trying to liquidate old cards.
This style of data aggregation analysis is immensely useful.
Is there any chance of this evolving into an on-demand queryable list where we'd be able to quickly get the highest store buy price across all included stores? (And could the aggregator sites, bidwicket.com and allmagiccards.com be tapped as additional data sources?)
@Lackey
That is definitely something I am looking to do eventually, and I'd like to take it further as well. I'm a bit limited by my programming knowledge, so I need to get that going first and then I'll be able to build some tools. I might be able to get some help in building these kinds of tools before then, but for now we'll have my spreadsheets.
I only bring this up because I'm not certain of it myself, but aren't BlackBorder and AdventuresOn run by the same people? Every time I've bought from BlackBorder it's been shipped in an AdventuresOn envelope. In addition, the prices they have are identical in every category you posted. It looks like you might be giving double weight to a single buyer/seller.
That may be true, I don't think I've personally made a purchase from either of them. I'm familiar with Black Border through some friends who swear by their pricing accuracy, and AdventuresOn was recommended to me by Kelly. The lists looked to same to me as well, but I thought one might be a copy cat. I'll check into it and fix the spread sheets to account for the duplication if there is an issue. Thank you for bringing this to my attention.
Michel Poon would you be willing to share your Python script for gathering b buylist data? If so please email me @ ymmij1950@gmail.com.
3FEyB7 Comments do help sites flourish, especially w