12 Jan, 2010, Zeno wrote in the 1st comment:
Votes: 0
Let's say I want to generate census data for my MUD from pfiles. It would look something like this: http://www.playonline.com/ff11us/guide/d...

What programming/scripting language should I use? Lua? Python? I know it doesn't really matter because the end result will be the same, but still. Throw out some suggestions.
12 Jan, 2010, David Haley wrote in the 2nd comment:
Votes: 0
I would use whatever you are most comfortable with to get the job done most quickly.

I probably would not use Lua for this as its standard library isn't very large. You will have to look hard to find libraries to generate pie charts, for example (or do it yourself with some kind of image rendering library – another usage that isn't super-common).

Personally, I would probably use Python. In another life, before I discovered Python, I might have used Perl. If Ruby floats your boat, use that too.

Personally, I would not use C++, Java, or anything of that nature.

If you don't really care, find a language that has good chart generation libraries, and pick that language. As long as it has decent text parsing (regular expressions are your friend) you should be fine.
12 Jan, 2010, shasarak wrote in the 3rd comment:
Votes: 0
In a commercial, web-based, Windows-based environment I would use C#/ASP.NET with Dundas Chart for the graphs.
12 Jan, 2010, Barm wrote in the 4th comment:
Votes: 0
Another vote for Python.

Check out some of the examples on pyGoogleChart. Since it uses the Google charting API you don't even have to generate/host images.
12 Jan, 2010, quixadhal wrote in the 5th comment:
Votes: 0
Perl.

Everyone hates it, but they will grudgingly admit it's the perfect tool for quickie scripts like this. I'll even point towards the GD library, which can be used to generate pretty pie charts with fairly little effort. For the login graph, you could probably just use MRTG or something similar, or you can roll your own as well.

The big reason perl survives is that you can find stuff in CPAN that does half your work for you. :)
13 Jan, 2010, Zeno wrote in the 6th comment:
Votes: 0
Going with Python, wasn't planning to use the same language to generate graphs since I already have Open Flash Chart and it's pretty nifty.

On the subject of census data, should I exclude players under say 4 hour playtime?
13 Jan, 2010, David Haley wrote in the 7th comment:
Votes: 0
I guess it depends on what exactly you're trying to do. I'm assuming you mean total playtime less than four hours? Those people can still be interesting. In fact, they are arguably very interesting as they are the ones who did not like it enough to stick it around, so perhaps it's worth finding out why. What were they doing during those four hours that they didn't like enough to stay? But if you're trying to present statistics to future players, and not for your own analysis, perhaps you should exclude people who "come and go" very quickly.
13 Jan, 2010, Zeno wrote in the 8th comment:
Votes: 0
Well I guess I mean like… if I have say 5000 players and 1/5 of those are people who never did anything (level 2, ~1 hour playtime, etc) then should I really include them in a census? The goal is not to find out why people stopped playing, but to show the playerbase a general idea of what the community is. How many are male, what is the most played race, etc. 1/5th is a lot and I figured it would be enough to skew the results of what it really represents.

Like WoW's census: http://www.warcraftrealms.com/realmstats...
Quote
Realm stats are easily viewable snapshot data taken from the overall database. The data is limited to only characters level 10 and up that have been seen in the previous 30 days.
13 Jan, 2010, KaVir wrote in the 9th comment:
Votes: 0
Zeno said:
Well I guess I mean like… if I have say 5000 players and 1/5 of those are people who never did anything (level 2, ~1 hour playtime, etc) then should I really include them in a census?

Personally I think its worth collecting the data from every player, but once you've got that data there's no reason why you couldn't create separate charts based on those who have connected within the last X days and/or played a minimum of Y hours and reached at least level Z.

A couple of months ago I collected data on playing time, although it was mostly out of idle curiousity: http://www.godwars2.org/playingtime.html - after reading this thread I'm tempted to collect more data and produce some charts.

EDIT: I wonder if it might even be worth having the mud keep track of some of the data internally, so that the website stats are always up to date. Some browser games show a player list with an indication of when they last logged on - it'd be interesting to see something like that for a mud, I think, perhaps having a section of the website showing all players who have logged on within the last 30 days, along with some statistical information. Okay so you could just set up a script to generate the same data once per day, but I still think it would be nice to show something that's always current.
20 Jan, 2010, KaVir wrote in the 10th comment:
Votes: 0
KaVir said:
I wonder if it might even be worth having the mud keep track of some of the data internally, so that the website stats are always up to date. Some browser games show a player list with an indication of when they last logged on - it'd be interesting to see something like that for a mud, I think, perhaps having a section of the website showing all players who have logged on within the last 30 days, along with some statistical information. Okay so you could just set up a script to generate the same data once per day, but I still think it would be nice to show something that's always current.

I thought about the above idea for a few days, and finally decided to add it. It's only been in place for about 10 hours, and I didn't initialise it with any data, so I expect it to keep growing for the next few days. But this is what it looks like:

http://www.godwars2.org/mwi/players

It only adds you to to the list when you log on, and only if you've got at least 3 hours of playing time, so it won't include one-time visitors. You can hover over the coloured dots to see how long ago the player was online (relative to the time the page was last refreshed), but the colour code is:

  • Cyan: Online right now.

  • Blue: Online in the last 24 hours.

  • Magenta: Online in the last 3 days.

  • Red: Online in the last week.

  • Black: Online in the last 30 days (after that you vanish from the list).


  • The names of the players and clans are hyperlinks, but I will also include some more data about the active players as a whole (such as total/average playing time, how many of each class, etc). I'm not quite sure where I'll add it yet, perhaps it could be a section at the top or bottom of the page, or maybe it should be another page entirely.

    It might be nice to have some graphs as well, but I'd have to look up how to do that - my web design skills are mediocre at best.
    20 Jan, 2010, Zeno wrote in the 11th comment:
    Votes: 0
    Quote
    It might be nice to have some graphs as well, but I'd have to look up how to do that - my web design skills are mediocre at best.

    Check out http://teethgrinder.co.uk/open-flash-cha...
    Know how to code and it's pretty easy. I cranked out this in a day: http://biyg.org/graphs/
    21 Jan, 2010, elanthis wrote in the 12th comment:
    Votes: 0
    If you're against Flash for any reason (some people are, some people aren't), there's also things like http://www.deensoft.com/lab/protochart/i..., http://www.liquidx.net/plotkit/, and many others. Most of them are even out-of-the-box IE6 compatible using Google's canvas->VML translation library. We use Open Flash Chart at work and it's pretty slick, but I'm planning on switching to one of the JS libraries to help alleviate the loading problem Open Flash Chart causes some of our users (for some reason on some setups it can take the Flash widget a long time to start and then to load the data from the server).
    03 Feb, 2010, KaVir wrote in the 13th comment:
    Votes: 0
    elanthis said:
    If you're against Flash for any reason (some people are, some people aren't), there's also things like http://www.deensoft.com/lab/protochart/i..., http://www.liquidx.net/plotkit/, and many others.

    Thanks - I finally got around to having a play with the first one of those yesterday. I've been having trouble displaying the chart properly with firefox (although it works fine with IE), so I may try out the second one later and compare the two. I'm still just playing with ideas at the moment, but you can get a rough idea of the sort of thing I'm aiming for here:

    http://www.godwars2.org/mwi/statistics

    I think the piechart looks a bit inconsistent with the other stats, so I may end up removing it from that page and using it elsewhere.
    03 Feb, 2010, David Haley wrote in the 14th comment:
    Votes: 0
    The 'active players by gender' stats makes it look like you're talking about the player gender, but I would imagine that you're actually talking about the character gender considering how the numbers line up with the other character numbers (like class). Out of curiosity, why do you have that stat there?

    Also, are the active player options really for the entire game, or just people right now? You have mentioned at several points that you have a lot of blind users who use the config options and yet that number says that only two people have the option set.

    I also find it somewhat confusing that the 'active in the last XYZ' stats are not cumulative. I would expect that everybody active in the last 24 hours would also have been active in the last 3 days, for example.
    03 Feb, 2010, KaVir wrote in the 15th comment:
    Votes: 0
    David Haley said:
    The 'active players by gender' stats makes it look like you're talking about the player gender, but I would imagine that you're actually talking about the character gender considering how the numbers line up with the other character numbers (like class). Out of curiosity, why do you have that stat there?

    Yes, all of the stats are for "characters" rather than actual "players". Perhaps the term "player characters" might be better.

    As I said before, I'm really just playing with ideas at the moment. There's no particular reason why I included gender, except that I vaguely recall it being mentioned on another forum (TMC?) when someone was asking mud owners about stats for their muds. Occasionally people write articles about male/female ratios in different types of online game, so maybe including that data will get my mud listed as a reference somewhere - I don't see any harm in including it.

    David Haley said:
    Also, are the active player options really for the entire game, or just people right now? You have mentioned at several points that you have a lot of blind users who use the config options and yet that number says that only two people have the option set.

    It's been collecting information for about 3 weeks, but I only added the sound and screen reader options last night (about 16 hours ago). I should probably run through the existing active player list and update the data, but at this point I'm really just trying to get a feel for what sort of information should be included, and how it could be presented.

    David Haley said:
    I also find it somewhat confusing that the 'active in the last XYZ' stats are not cumulative. I would expect that everybody active in the last 24 hours would also have been active in the last 3 days, for example.

    Yeah that probably would make more sense.
    03 Feb, 2010, David Haley wrote in the 16th comment:
    Votes: 0
    KaVir said:
    There's no particular reason why I included gender, except that I vaguely recall it being mentioned on another forum (TMC?) when someone was asking mud owners about stats for their muds. Occasionally people write articles about male/female ratios in different types of online game, so maybe including that data will get my mud listed as a reference somewhere - I don't see any harm in including it.

    I guess my issue with character gender is that it's either misleading, or provides little (or even no) useful information. I'm assuming that on your game, gender is a cosmetic choice and doesn't affect gameplay. In this respect, it's basically the same as tracking hair color. But it looks like it could be tracking player gender, which is something that people might actually care about (because it could affect your interactions with other people, or something along those lines). I can easily imagine that some people would prefer to play with people of their gender in some cases, or the opposite in others.

    The male/female ratios I've seen are usually w.r.t. actual player genders, although there are also articles written about when people play characters of the opposite gender. (Both cases require knowing the player's gender.)

    KaVir said:
    It's been collecting information for about 3 weeks, but I only added the sound and screen reader options last night (about 16 hours ago).

    Ah, I see then. That makes sense.

    KaVir said:
    I should probably run through the existing active player list and update the data, but at this point I'm really just trying to get a feel for what sort of information should be included, and how it could be presented.

    It might be nice to see player distribution by GMT time, so that I could see how player activity relates to the time I have available to play.

    Since you have a coordinate-based world, another interesting statistic might be how concentrated players are in the game; perhaps you could measure this as standard deviation of distance between players. This might be more interesting to you than to people visiting your website, though.
    04 Feb, 2010, KaVir wrote in the 17th comment:
    Votes: 0
    David Haley said:
    I guess my issue with character gender is that it's either misleading, or provides little (or even no) useful information. I'm assuming that on your game, gender is a cosmetic choice and doesn't affect gameplay. In this respect, it's basically the same as tracking hair color.

    Mechically, yes. But gender has much more of a social impact - even just the perception of gender (which is really what people talk about in most cases, as there's no way to know what percentage of the characters are played by actual men and women). Perhaps it would be more interesting if I tied the gender to other data though, such as the frequently of PK deaths (i.e., are female characters less likely to be attacked?).

    David Haley said:
    It might be nice to see player distribution by GMT time, so that I could see how player activity relates to the time I have available to play.

    MudStats provides the same information, but it does a single check each hour. Perhaps I could give proper hourly averages instead.

    David Haley said:
    Since you have a coordinate-based world, another interesting statistic might be how concentrated players are in the game; perhaps you could measure this as standard deviation of distance between players. This might be more interesting to you than to people visiting your website, though.

    The problem is it's an unrestricted PK mud with no concept of grouping. If two players find themselves in close proximity, one will usually kill the other.
    04 Feb, 2010, David Haley wrote in the 18th comment:
    Votes: 0
    KaVir said:
    Perhaps it would be more interesting if I tied the gender to other data though, such as the frequently of PK deaths (i.e., are female characters less likely to be attacked?).

    This would indeed be a fascinating thing to know, as it would indicate that people have some kind of reluctance to treat perceived gender differently than others…

    KaVir said:
    MudStats provides the same information, but it does a single check each hour. Perhaps I could give proper hourly averages instead.

    I think you'd have to in order to give good information, yes.

    KaVir said:
    The problem is it's an unrestricted PK mud with no concept of grouping. If two players find themselves in close proximity, one will usually kill the other.

    But isn't it still interesting to know that players tend to congregate in the same places, even if they end up killing each other? That would be different than players wandering all over the map looking for each other.
    04 Feb, 2010, Barm wrote in the 19th comment:
    Votes: 0
    David Haley said:
    This would indeed be a fascinating thing to know, as it would indicate that people have some kind of reluctance to treat perceived gender differently than others…


    Early in Everquest, I was the only blacksmith on my server selling sets of dyed fine-plate armor. This was years before they added the player appliable dyes and the only way to get custom colored gear. The stats were lousy, it took days to collect the ingredients, the vendor purchased components were stupid expensive, and failure rates were brutal so we're talking pure vanity gear and I charged accordingly. As a shaman, I could make the dyes via alchemy but not wear the armor and, naturally, people wanted to see it before handing over a ton of cash. So I made a male barbarian warrior to model and sell it.

    The downside of selling brightly colored sets of expensive vanity gear were the non-customers who wanted to haggle over the cost, get angry, then come back an hour later and haggle some more. This went on for about a week until I made a female barbarian warrior to model it. The difference was amazing. Guys do not want to appear cheap in front of women, even virtual women. People were far more polite and I sold things with hardly any debate at all.
    0.0/19