The Hotness
Games|People|Company
Among the Stars
Targi
Mage Knight: Board Game
Ace of Spies
Mice and Mystics
Eclipse
1984: Animal Farm
Virgin Queen
Lords of Waterdeep
The Big Bang Theory: The Party Game
Pirate Dice: Voyage on the Rolling Seas
Omen: A Reign of War
Eselsbrücke
Village
Descent: Journeys in the Dark (Second Edition)
The New Science
Vegas
Android: Netrunner
Kingdom Builder
A Game of Thrones: The Board Game (Second Edition)
Dominion
The Lord of the Rings: The Card Game
Nefarious
Dungeon Command: Sting of Lolth
Agricola: All Creatures Big and Small
Twilight Struggle
Glory to Rome
1989: Dawn of Freedom
War of the Ring
Quarriors! Quarmageddon
A Few Acres of Snow
K2
Thunder Road
Agricola
7 Wonders
Hawaii
The Convoy
Through the Ages: A Story of Civilization
Hemloch
Gladiatori
Zombicide
Trajan
Tammany Hall
Ora et Labora
The Castles of Burgundy
Arkham Horror
Terrain Game
Dominion: Dark Ages
Dominant Species
Core Worlds

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
7 
 Thumb up
0.03
 tip
 Thumb up

fun with BGG URLs, part III (and an update on El Chupacabra)

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


First up is a URL that takes you to a bunch of different ways to rank the games on BGG: http://www.boardgamegeek.com/topn.php3. My favorites here are sorting by number owned, by number wanted in trade, by number that users have put on their wishlist, and by the total number of pageviews. I was hoping to find a way to break down the number of views by year, to get an idea of when games were generating a lot of interest, but multiple attempts at URL hacking went exactly nowhere.

What I’d really like to do is come up with a relational analysis of all the games in, say, the top 200. Analyze the “recommends” data on each game, for example, to find links between games based on common ownership. Or find a way to do it based on “people who play / rate highly game x also play / rate highly game y.” In other words, first find out which games are “connected” to one another, then present this data graphically. If anyone has any ideas or suggestions for how to go about this (without losing what sanity I have), I’d love to hear them. I suspect that BGG’s xml api might be the way to go, but after an initial assessment of its features, I don’t quite see how. Again, though, I’d welcome any and all suggestions — enlighten me!

Since this relational analysis appears to be beyond me (for the moment), I’ll have to content myself with sharing something much more mundane: the ability to search BGG’s database for games of a particular weight, sorted in pretty much any way you want.

This, for example, looks at all the games with a weight between 1.8 and 2.0 (fairly light, includes both Carcassonne and Ticket to Ride) and sorts them by the number of people who rated the game: http://www.boardgamegeek.com/search/boardgame?sort=numvoters.... And yes, it’s ugly — but the important thing is that it works.

Now for a little URL hacking: change the 1.8 to 2.0 and the 2.0 to 2.2: http://www.boardgamegeek.com/search/boardgame?sort=numvoters.... This does the same search, only it bumps the weight up a little bit — this list includes Citadels, Alhambra, and 7 Wonders.

If you want to look at slightly heavier games again, then change the 2.0 to 2.2 and the 2.2 to 2.4 (this’ll get you games like Dominion, Pandemic, and Settlers of Catan). While you’re at it, change the “numvoters” to “rank” to sort the resulting games by their BGG ranking: http://www.boardgamegeek.com/search/boardgame?sort=rank&advs.... Or change “rank” to “numowned” to sort by the number of BGG users who own the game: http://www.boardgamegeek.com/search/boardgame?sort=numowned&.... Easy, huh?

I’ve always thought it should be easier to get at some of this info, but it’s nothing a little URL hacking can’t take care of. :-)



And, as promised, an update on El Chupacabra. Those who have been following my blog know that the game didn’t work as well with 7 as I had hoped, so I went back to the drawing board to come up with something a little different. I decided that the opportunity to “shoot the moon” offered by the role of El Chupacabra just wasn’t interesting enough to justify its overhead in terms of rules complexity, and I also decided that the mechanism just didn’t fit very well with the push-your-luck gameplay. So I took it out.

I’ve been thinking for some time that it would be fun to design a game that encourages people to cheat, or at the very least encourages people to call one another cheaters. The first step, then, was to check to see if that name had been taken, and sadly, it had: Cheater. I always hate it when a poorly-rated game is taking up a perfectly good name, but what do you do? I mean, aside from turning your copy of Cheater into a perfectly serviceable version of The Bottle Imp.

Naming concerns aside (I can’t decide between Russian Rollette and Cheater’s Dice), I decided to gut the original game, take out all the fussy bits having to do with scoring, and make it a betting and bluffing game pure and simple. Think of it as a dice version of Bullshit (a.k.a. I Doubt It / Cheat) with gambling thrown in. It’ll fall most definitely into the Beer & Pretzels category of gaming….

More anon.
Twitter Facebook
9 Comments
Subscribe sub options Sun Jun 5, 2011 4:29 pm
Post Comment
Mark Klassen
Canada
Mission
British Columbia
designer
Avatar
mbmbmbmbmb
I have been hoping for a good relational analysis tool like the one you mention for a while. Here's my suggestion from about two years ago: Better Similarly Rated feature.

Your blogs are quite interesting.
1 
 Thumb up
 tip
 Thumb up
  • Posted Sun Jun 5, 2011 5:17 pm
    • Choose your Dice
      • Roll
      • Comment (Optional)
    • Reply
    •  
    • Quote
Dave Ross
United States
Ames
Iowa
designer
Avatar
mbmbmbmbmb
manchuwok wrote:
I have been hoping for a good relational analysis tool like the one you mention for a while. Here's my suggestion from about two years ago: Better Similarly Rated feature.

Your blogs are quite interesting.

Yeah, what I'm looking for is a kind of real-time taxonomy of games, presented visually. In an ideal world. In a not-so-ideal world, it could be a static feature, updated every month or so.

But the idea is that related games would be connected visually, so that people could tell at a glance which games were (a) similar to other games and (b) the "hub" of that particular branch.

Kind of like a dynamically-created family tree. If someone liked Puerto Rico and wanted something similar, they'd take one look at the graph and say, "based on the connections, x looks good." Or if a person wanted to explore a new branch of the tree, they'd look at the picture and say, "based on size, y is the first game to get."

Thanks for the encouragement.
 
 Thumb up
 tip
 Thumb up
  • Posted Sun Jun 5, 2011 6:10 pm
    • Choose your Dice
      • Roll
      • Comment (Optional)
    • Reply
    •  
    • Quote
Gordon Stewart
United States
Bowie
20715
flag msg tools
designer
Avatar
mbmbmbmbmb
Somewhat related, David Dokter (Herr Dr) does interesting
graphs w/BGG statics; for instance:
1 
 Thumb up
 tip
 Thumb up
  • Edited Sun Jun 5, 2011 6:47 pm
  • Posted Sun Jun 5, 2011 6:46 pm
    • Choose your Dice
      • Roll
      • Comment (Optional)
    • Reply
    •  
    • Quote
Kevin B. Smith
United States
Margate
Florida
designer
Avatar
mbmbmbmbmb
Could you use the XML API to download a bunch of raw data, stuff that into an rdbms, and do the analysis there? Or does the XML API not expose enough user-specific data to do what you want (I haven't studied the API)? Or are you just not an XML API geek?
 
 Thumb up
 tip
 Thumb up
  • Posted Mon Jun 6, 2011 1:28 am
    • Choose your Dice
      • Roll
      • Comment (Optional)
    • Reply
    •  
    • Quote
Dave Ross
United States
Ames
Iowa
designer
Avatar
mbmbmbmbmb
peakhope wrote:
Could you use the XML API to download a bunch of raw data, stuff that into an rdbms, and do the analysis there? Or does the XML API not expose enough user-specific data to do what you want (I haven't studied the API)? Or are you just not an XML API geek?

Well, I'm not sure what all is possible (most definitely not an XML API geek), but looking at it initially I didn't see that I could get, for example, the list of recommended games (which is based on common ownership). And I don't think I could get enough information to infer connections based on common play habits, either (i.e., "x percent of players who have played y have also played z").

I looked into a social network analysis program called socnetv which can crawl websites to determine which sub-sites are linked to which (and thus generate a network based on, again, the "recommended" data), but I didn't see a way to limit how fast their bot crawled (crawling too fast violates BGG's terms of service). I could maybe get permission to do that, but I'd feel more comfortable asking if I knew better what I was doing (and had more experience doing it).

So I'm kind of at a dead end unless someone offers some concrete advice, though I'll certainly keep thinking of ways to try to do it.

I could possibly go through the recommended data by hand for the top 100 games, but I doubt I'd want to do that more than once. It would provide a nice snapshot of the so-called family tree, in other words, but it would soon become obsolete.
 
 Thumb up
 tip
 Thumb up
  • Posted Mon Jun 6, 2011 4:13 am
    • Choose your Dice
      • Roll
      • Comment (Optional)
    • Reply
    •  
    • Quote
Kevin B. Smith
United States
Margate
Florida
designer
Avatar
mbmbmbmbmb
ddgdrs wrote:
Well, I'm not sure what all is possible (most definitely not an XML API geek), but looking at it initially I didn't see that I could get, for example, the list of recommended games (which is based on common ownership). And I don't think I could get enough information to infer connections based on common play habits, either (i.e., "x percent of players who have played y have also played z").

I just looked at the API, and it looks like you are right. For a given game, you can retrieve a list of everyone who has commented on it (with their rating), but I guess you can't get a list of users who have rated it (regardless of comment). Bummer.

With a list of users, you could query each to find their total number of plays. You couldn't track that over time, but you could do the "highly rated and also played" analysis.

I suppose living with the subset of "users who have commented on" rather than "all users" might still give you some interesting data.
1 
 Thumb up
 tip
 Thumb up
  • Posted Mon Jun 6, 2011 5:02 am
    • Choose your Dice
      • Roll
      • Comment (Optional)
    • Reply
    •  
    • Quote
Mark Klassen
Canada
Mission
British Columbia
designer
Avatar
mbmbmbmbmb
ddgdrs wrote:
manchuwok wrote:
I have been hoping for a good relational analysis tool like the one you mention for a while. Here's my suggestion from about two years ago: Better Similarly Rated feature.

Your blogs are quite interesting.

Yeah, what I'm looking for is a kind of real-time taxonomy of games, presented visually. In an ideal world. In a not-so-ideal world, it could be a static feature, updated every month or so.

But the idea is that related games would be connected visually, so that people could tell at a glance which games were (a) similar to other games and (b) the "hub" of that particular branch.

Kind of like a dynamically-created family tree. If someone liked Puerto Rico and wanted something similar, they'd take one look at the graph and say, "based on the connections, x looks good." Or if a person wanted to explore a new branch of the tree, they'd look at the picture and say, "based on size, y is the first game to get."

Thanks for the encouragement.


Sounds neat. I hadn't considered a visual representation of the "family tree" I was looking for. Most useful, as you note, would of course be that it is dynamically-created. I was also looking for variable input parameters. Users could decide what type of connections they want graphed (community ratings is the main suspect here, but many other pieces of data could be used) and filter based on any relevant criteria (game weight, mechanics, playing time, etc).
1 
 Thumb up
 tip
 Thumb up
  • Posted Mon Jun 6, 2011 7:46 am
    • Choose your Dice
      • Roll
      • Comment (Optional)
    • Reply
    •  
    • Quote
George Leach
United Kingdom
Salford
Greater Manchester
designer
El Chupacabra sounds, at least in principal, very interesting. At a guess all or nothing scoring may be causing problems. If it's quick then scoring over multiple rounds might be more enjoyable.
 
 Thumb up
 tip
 Thumb up
  • Posted Wed Jun 15, 2011 4:11 pm
    • Choose your Dice
      • Roll
      • Comment (Optional)
    • Reply
    •  
    • Quote
Dave Ross
United States
Ames
Iowa
designer
Avatar
mbmbmbmbmb
Jugular wrote:
El Chupacabra sounds, at least in principal, very interesting. At a guess all or nothing scoring may be causing problems. If it's quick then scoring over multiple rounds might be more enjoyable.

The game most definitely had its strengths -- it was interesting to see if you could play in one way (trying to get a low score) while trying to make everyone else think you were playing in another way (going for a high score).

It did make the game more complicated, though, and fussier, too. I ultimately took it out because I wanted more immediate tension with less overhead, but it's certainly an idea I wouldn't mind recycling for some other game.

Also, before taking El Chupacabra out completely, I made it so the player in that role could score either way -- not just by shooting the moon. It made the role much more powerful (possibly too powerful) and made the game more interesting. El Chupacabra was often the player who was pushing the hardest -- if she succeeded, she got high score, and if she failed, she got low score. It was a win-win for her, though other players would often accuse her (not a problem if she went high, but definitely a problem if she went low). The role encouraged players to take more risks.
 
 Thumb up
 tip
 Thumb up
  • Posted Thu Jun 16, 2011 1:15 am
    • 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.