The Hotness
Games|People|Company
Dominion: Dark Ages
Total War
Mage Knight: Board Game
Fantastiqa
Libertalia
The Lord of the Rings: Nazgul
Descent: Journeys in the Dark (Second Edition)
Eclipse
Mice and Mystics
Doctor Who: The Card Game
Lords of Waterdeep
Star Wars: X-Wing Miniatures Game
Agricola: All Creatures Big and Small
Dungeon Fighter
Android: Netrunner
Virgin Queen
A Game of Thrones: The Board Game (Second Edition)
Glory to Rome
Infiltration
Collapsible D: The Final Minutes of the Titanic
Dominion
The Lord of the Rings: The Card Game
Twilight Struggle
City of Horror
Snowdonia
1989: Dawn of Freedom
Goa
Sherlock Holmes Consulting Detective
Agricola
Among the Stars
7 Wonders: Cities
7 Wonders
The Swarm
Through the Ages: A Story of Civilization
Arkham Horror
Village
Ora et Labora
Battles of Westeros: House Baratheon Army Expansion
Race for the Galaxy
War of the Ring
Trajan
Kingdom Builder
The Castles of Burgundy
Zombicide
Twilight Imperium (third edition)
Space Alert
Dungeon Command: Sting of Lolth
Hacienda
Battlestar Galactica
Ground Floor

playing and designing board games

Gaming blog about board games and board game design. Includes reviews and session reports, strategies and variants, thoughts on games in development, tips for hosting game nights, etc.
Recommend
22 
 Thumb up
 tip
 Thumb up

more BGG stats: fun with numbers!

Dave Ross
United States
Ames
Iowa
designer
Avatar
mbmbmbmbmb
Note: this content also appeared on my blog, playing and designing board games.


Ok, so I’ve been having more fun. First up is just the top twenty games that are “wanted in trade” on BGG:



Next up is the top twenty games that are being offered for trade:



And finally, the top fifty games with the highest want in trade / offered in trade ratio:



One caveat: I didn’t go through the entire BGG database looking for the twenty games with the highest ratio, instead I assumed that they would all be in the top 100 games wanted in trade. And this might well be a bad assumption. Still, you have to start somewhere, and I also wanted to avoid games that had an artificially high ratio due to a small sample size (1 wanting and 0 offering would, after all, be infinite).

How can a list like this be used? Well, it gives a pretty good idea what to look for when thrifting, for one thing, and it also gives you some idea just how desirable that mint copy of Antiquity you’ve got down in the basement is.

I guess I was surprised that Crokinole came out so high and that Goa didn’t come out any higher than it did. It didn’t surprise me that so many newer games were on the list, though, since demand is high and folks haven’t had time to tire of them yet.

All in all a fun little exercise. I hope you enjoy it.
Twitter Facebook
12 Comments
Subscribe sub options Wed Jun 8, 2011 9:28 pm
Post Comment
Paul Smith
United States
Lakewood
California
flag msg tools
Avatar
mbmbmbmbmb
An interesting measure of popularity vs. availability.

I took a look at some of the games below the top 100 wanted in trade, and there are a lot that stay around a 10:1 want/offered ratio (still with over 100 users wanting).
 
 Thumb up
 tip
 Thumb up
  • Posted Wed Jun 8, 2011 10:34 pm
    • Choose your Dice
      • Roll
      • Comment (Optional)
    • Reply
    •  
    • Quote
Dave Ross
United States
Ames
Iowa
designer
Avatar
mbmbmbmbmb
SaiSaysPlayGo wrote:
An interesting measure of popularity vs. availability.

I took a look at some of the games below the top 100 wanted in trade, and there are a lot that stay around a 10:1 want/offered ratio (still with over 100 users wanting).

I had wanted to find some way to automate this process -- I think pivot tables would have worked, but I use Open Office Calc, and that feature isn't supported. So what I did was go through the top 100 games wanted in trade by hand to find out how many people were offering those games in trade. Tedious, yes, but it worked.

I didn't have the patience to go through any more than the top 100. It's good to know, though, that some games farther down the "want" list do indeed have a higher "want"/"offer" ratio than some of the games listed here. Thanks for checking that!

I still think this list is interesting, even though it is in no way definitive. (And if anyone has Excel and wants to delve deeper into the database, feel free. Please post results below. )

 
 Thumb up
 tip
 Thumb up
  • Posted Wed Jun 8, 2011 10:54 pm
    • Choose your Dice
      • Roll
      • Comment (Optional)
    • Reply
    •  
    • Quote
Randy Cox
United States
Clemson
South Carolina
designer
1024x768 works just fine - Don't Wide the Site!
badge
The Back Alley gets no respect.
Avatar
mbmbmbmbmb
ddgdrs wrote:
(And if anyone has Excel and wants to delve deeper into the database, feel free. Please post results below. )
I'd love to delve deeper, but how will having Excel help? I mean, the difficult part is pulling the data and that, as you mentioned, is a tedious process. To my knowledge, there's no real-time means of pulling data for all games (or even most games) without going to each one separately.
 
 Thumb up
 tip
 Thumb up
  • Posted Thu Jun 9, 2011 12:14 am
    • Choose your Dice
      • Roll
      • Comment (Optional)
    • Reply
    •  
    • Quote
John Holder
United States
Centennial
Colorado
flag msg tools
designer
Avatar
mbmbmbmbmb
Neat work. I'm not too surprised Antiquity is the most desired / least offered king. Wish I had a copy too.
 
 Thumb up
 tip
 Thumb up
  • Posted Thu Jun 9, 2011 12:19 am
    • Choose your Dice
      • Roll
      • Comment (Optional)
    • Reply
    •  
    • Quote
Dave Ross
United States
Ames
Iowa
designer
Avatar
mbmbmbmbmb
Randy Cox wrote:
ddgdrs wrote:
(And if anyone has Excel and wants to delve deeper into the database, feel free. Please post results below. )
I'd love to delve deeper, but how will having Excel help? I mean, the difficult part is pulling the data and that, as you mentioned, is a tedious process. To my knowledge, there's no real-time means of pulling data for all games (or even most games) without going to each one separately.

I was originally thinking that if one were to list the top, say, 1000 games wanted (10 pages) and the top 1000 games for trade, then it would be possible to merge the data using pivot tables. But I see now that there would be way too many holes. You could probably get decent results with 5000 games, but (a) there would still be holes and (b) 50 pages is starting to get pretty tedious again.

It would be easy if there were a current snapshot we could download, but I believe the last one was taken in '06. And it would be easy if we could just add another column to either listing mentioned above. Anyone know of a URL hack that would add another column? Any admins want to make the view more customizable?

I had thought it would be relatively easy to dig deeper with the right tools, but I see I was mistaken. soblue

1 
 Thumb up
 tip
 Thumb up
  • Posted Thu Jun 9, 2011 1:33 am
    • Choose your Dice
      • Roll
      • Comment (Optional)
    • Reply
    •  
    • Quote
Gregor Samsa
Netherlands

mb
Thanks Dave, for the exercise. By the way, nice cat-avatar laugh.
Cheers.
 
 Thumb up
 tip
 Thumb up
  • Posted Thu Jun 9, 2011 11:05 am
    • Choose your Dice
      • Roll
      • Comment (Optional)
    • Reply
    •  
    • Quote
Pelle Nilsson
Sweden
Linköping
designer
Avatar
mbmbmbmbmb
Found your blog now and this is all fun. How do you get the data from the html pages? Do you just copy the table into openoffice and let it split it up for you? Unlike when browsing your collection I find no link to download csv data. As you mentioned in an older post the XML API is probably the way to go, if it is possible to figure out how to list games like this. Unfortunately that API has no option as far as I can tell to output some simple format like csv, so any analysis would have to start with parsing the XML. Other than that this sounds like a fun exercise for me to learn some more R. Adding it to my todo list. Worst case I guess it is possible to parse the html tables directly from those pages you link to.
 
 Thumb up
 tip
 Thumb up
  • Posted Thu Jun 9, 2011 7:27 pm
    • Choose your Dice
      • Roll
      • Comment (Optional)
    • Reply
    •  
    • Quote
Dave Ross
United States
Ames
Iowa
designer
Avatar
mbmbmbmbmb
pelni wrote:
Found your blog now and this is all fun. How do you get the data from the html pages? Do you just copy the table into openoffice and let it split it up for you? Unlike when browsing your collection I find no link to download csv data. As you mentioned in an older post the XML API is probably the way to go, if it is possible to figure out how to list games like this. Unfortunately that API has no option as far as I can tell to output some simple format like csv, so any analysis would have to start with parsing the XML. Other than that this sounds like a fun exercise for me to learn some more R. Adding it to my todo list. Worst case I guess it is possible to parse the html tables directly from those pages you link to.

What I did for this was just to copy and past the table from the "wanting" web page into OOCalc, then manually look up the "trading" info for each game. But I've been looking at various ways to try to get deeper into the database without too much hassle. And you mention some good possibilities. I'm currently looking into either javascript or a greasemonkey script that would display more results on a single page, then grabbing the html from that and parsing the data. wget (for linux) is a good way to grab pages, but so far I haven't been able to figure out how to encourage wget to save everything in one big file (it really wasn't meant to do that). Parsing the html will be tricky, but do-able.

Oh, and once I get the data into OOCalc, I think it'll be possible to use the "data pilot" feature to merge the two lists. (Didn't know that was possible, but I think it will work.)

One thing, too, that wget will allow -- I can (slowly) grab all the game pages, and then I should be able to analyze the connectivity using socnetv (another project I'm working on). As you say, fun stuff. And, like you, I see it as an opportunity to learn new things.

I hope I'm not being dense, but what's R?
 
 Thumb up
 tip
 Thumb up
  • Posted Thu Jun 9, 2011 9:04 pm
    • Choose your Dice
      • Roll
      • Comment (Optional)
    • Reply
    •  
    • Quote
Dave Ross
United States
Ames
Iowa
designer
Avatar
mbmbmbmbmb
Gregor Samsa wrote:
Thanks Dave, for the exercise. By the way, nice cat-avatar laugh.
Cheers.

Thank you. I like your avatar, too.
 
 Thumb up
 tip
 Thumb up
  • Posted Thu Jun 9, 2011 9:13 pm
    • Choose your Dice
      • Roll
      • Comment (Optional)
    • Reply
    •  
    • Quote
Pelle Nilsson
Sweden
Linköping
designer
Avatar
mbmbmbmbmb
R is sort of a scripting language to do statistics. So you can write a script that loads the data and outputs graphs and stuff. Like a spreadsheet but without the spreadsheet. Actually makes a lot of things easier than having to enter formulas in lots of cells, imo.

I found the BGG XML Snapshot, but unfortunately it has not been updated for years.

Here is a description of how to reasonably quickly grab info about all games: http://boardgamegeek.com/article/4111860#4111860

But if I have to mess with loading lots of pages, parsing the xml, merging it into one "table" (csv file, most likely) my R skills are far too weak. I would cheat and use python instead for that.
 
 Thumb up
 tip
 Thumb up
  • Posted Thu Jun 9, 2011 9:37 pm
    • Choose your Dice
      • Roll
      • Comment (Optional)
    • Reply
    •  
    • Quote
Pelle Nilsson
Sweden
Linköping
designer
Avatar
mbmbmbmbmb
Yes, it seems quite easy to grab all the xml as described in that post. My python script is now downloading 450 games at a time. Will be 200-something requests to grab everything, so this isn't something to do repeatedly I guess. A quick look at the XML shows that the trade info is included, so it should be quite easy to parse out what is needed to calculate the ratio as above.
 
 Thumb up
 tip
 Thumb up
  • Posted Thu Jun 9, 2011 10:38 pm
    • Choose your Dice
      • Roll
      • Comment (Optional)
    • Reply
    •  
    • Quote
Dave Ross
United States
Ames
Iowa
designer
Avatar
mbmbmbmbmb
pelni wrote:
Yes, it seems quite easy to grab all the xml as described in that post. My python script is now downloading 450 games at a time. Will be 200-something requests to grab everything, so this isn't something to do repeatedly I guess. A quick look at the XML shows that the trade info is included, so it should be quite easy to parse out what is needed to calculate the ratio as above.

Excellent! It'll be interesting to see the results.
 
 Thumb up
 tip
 Thumb up
  • Posted Fri Jun 10, 2011 9:21 pm
    • Choose your Dice
      • Roll
      • Comment (Optional)
    • Reply
    •  
    • Quote

Subscribe

Categories

Contributors

Front Page | Welcome | Contact | Privacy Policy | Terms of Service | Advertise | Support BGG | Feeds RSS
Geekdo, BoardGameGeek, the Geekdo logo, and the BoardGameGeek logo are trademarks of BoardGameGeek, LLC.