« The Gamer Bloc | Main | Computer Games Magazine to Launch New MMO Publication »

Mar 15, 2006

Comments

1.

Wow...this sounds like an amazing project, but at the same time, it also sounds extremely difficult to keep your scraping software up to date *and* efficient upon patches. I know the mod community struggles wit this, and sometimes a 'clunky' line of code to get a mod out the door after one patch leads to lines and lines of spaghetti code 3 patches later.

Three cheers for pushing the boundaries and utilizing WoW's APIs for relevant data collection...hopefully they won't put a ban on the user created addons anytime soon that will make this sort of thing prohibitive in the future.

2.

This reminds me... haven't players made a number of individual-use scrapers for economic data, specifically Auction House prices? They certainly don't capture all the information about a transaction (notably, you can't capture the actual price something sold for!) but they do scrape a lot of information about how much various kinds of items are put on the market for.

3.

hopefully they won't put a ban on the user created addons anytime soon that will make this sort of thing prohibitive in the future
I think the problem is if these types of addons become popular with the general public for some use they will be putting extra strain on the game databases and increasing bandwith usage. The auction house scanners are one example that is already popular.

4.

So, I know this may sound silly, but why not just put together a research proposal and pitch it to Blizzard? Then you don't have to write silly scrapers... you just get the *real* data. Of course, you have to actually offer a compelling *reason* for your research that Blizzard will find of direct benefit to their business (That's always seems to be the hard part). I mean, wouldn't most of these queries take about 15 minutes if you had the actual system logs?

While I love the desire to understand online population dynamics and SNA (it was my business for a while), the research methodology seems a little weak, the census math missing, and the data collection is riddled with an appalling amount of noise.

Sorry to be party-pooper.

5.

I guess my question is similar to Michael's above: To what extent has Blizzard directly been a help or hindrance? Have they even acknowledged your existence?

6.

Nice idea here!

Why don't you develop a CLIENT for your data collection stuff. The client would send message in a chat channel (or whisper) the info to your BOT.

You could trigger the client to send infos like grouping status, gold amount, PvP kills/death and stuff like that.

This would permits you to collect more data. But the problem with this system would be that it might be "abused" if someone hack into your client to change to data transmitted.

But this could be overcome by having a lot of client software running. Or you could have a characters whitelist from whom you accept data.

I think that your project would greatly benefits from this.

And I really don't think that Blizzard will ever let you tap into their data. That would be too "dangerous". And I personally think that Blizzard already have tools that analyse data. This data is probably leading them in their development decision. Which type of content to add?
Which zone/instance is less used?
Which spells/class is less used?
What's the inflation rate?
Boting analysis...

On and on...

7.

thanks for this Eric... i'm intrigued by this stuff from a surveillance studies point of view actually. Data mining of this kind is an imperfect science at best (even when the pros do it) for many of the reasons you seem to have encountered. I've chatted with marketing folks who have similar experiences with mined data from consumer databases - moving targets, inconsisent data, constantly changing search parameters.

I have always wanted to challenge the ethical aspects of the PlayOn scrapers (do I really want to be scraped in WoW anymore than I want my buying patterns analyzed on Amazon?) but the process intrigues me enough to encourage the effort. Indeed on the ethical front - your efforts may lead to useful counter-surveillance tactics that players can make use of against the undoubtably more sophisticated surveillance apparatus that Blizzard maintains.

Of course I really love what you are doing but I am also dying to challenge the meaningfulness of this kind of data (in a friendly way) whether collected and acted upon by Blizzard or by PlayOn.

Perhaps its time for you guys to host a workshop at Parc.

8.

It takes a philosopher of science to properly understand what an empiricist outlook constitutes. As a general rule of thumb, if your interpreting engine never stumbled upon unexpected or unprocessable data, then one should naturally be suspicious of those observations.

9.

Naomi: Yes, as a player, I'm a big fan of Auctioneer by Norganna (auctioneeraddon.com). In poking around, however, we discovered that it was not possible to bring up an auction window programmatically with addon code. To get auction house information, it's necessary to right-click on one of the NPC auctioneers first.

Michael: (Warning! I'm going to edge awfully close to blatant self-promotion here.) We're right with you. Working with server-side data trumps scraping. We have tried pitching a research proposal to Blizzard -- providing tools that would lead to direct dollar savings -- and ... let's just say that we're now pursuing opportunities with other companies. But, hey! If you know of any game companies willing to cut a deal (Parc is a for-profit corporation, after all) to get top-drawer understanding of their customers, please send them our way.

Russ: In our brief discussions with Blizzard, one person told us that a "number" of people there read the PlayOn blog.

10.

If you're looking for a new world to get info about, take a look at Guild Wars. I'm an alpha tester for them, and the devs are very friendly.

11.

Interesting post, but it was the replies that really got me thinking. There is a big moral difference between observing people's actions and influencing them. If you mix them, you get something like ATITD's "Little Shop of Horrors", which I fled. If you are honestly trying to figure out how people in MMOG's behave, you must never, *ever*, participate in the game design.

12.

Did the people scraped get to give their consent?

13.

Hi ! Your site is very interesting. Thank you.

14.

Hello ! This is very [url=http://www.google.com/bb497]good[/url] site !!

15.

Interesting project, too bad you chose such an uninteresting MMOG to study. Of all the MMOG's I've play W0W is most like a Disney Land ride. There is no real individuality possible. Whatever race, class and level you choose you will be doing exactly the same thing as everyone else that chose that race class and level. In short there is no Meta game to make you feel as though your avatar makes a difference in how the world behaves ... YAWN!!! IMHO you're monitoring something that lacks reason to be monitored. I'm sure Disney knows how many tickets they sell a day and how many people ride each ride but, besides Disney, who cares?

16.

I think the problem is if these types of addons become popular with the general public for some use they will be putting extra strain on the game databases and increasing bandwith usage. The auction house scanners are one example that is already popular.
Valentine's Day Gifts

17.

I guess my question is similar to Michael's above: To what extent has Blizzard directly been a help or hindrance? Have they even acknowledged your existence?

Currency trading, forex market

18.

Of course, you have to actually offer a compelling reason for your research that Blizzard will find of direct benefit to their business with no wow powerleveling (That's always seems to be the hard part). I mean, wouldn't most of these queries take about 15 minutes if you had the actual system logs? Get world of warcraft power leveling here!

The comments to this entry are closed.