VectorStorm Blog

Quick note that due to a mistake I made in previous OS X builds, MMORPG Tycoon wouldn’t run on OS X 10.4 as I claimed it would. I’ve put up an 1.0.7 build for OS X which adds 10.4 compatibility.

If you’re running OS X 10.5, there’s no reason to upgrade; there are no major changes other than 10.4 compatibility.

So it’s been about a week since I released MMORPG Tycoon for the TIGSource procedural generation competition, so it’s probably about time I wrote about what went right and wrong.

But all the same, I don’t feel right calling this a “Post-Mortem”, as I’m not actually finished with MMORPG Tycoon.  In fact, I’m taking a break from coding on it right now.  But I’d like to talk a little about where I think it is right now, and where I see development going.

What went right:

• Having the “Procedural Generation” theme nicely focused my attention on the procedural aspects of MMORPG Tycoon, including the prebuilt level configuration, the names of the various players, the names and sizes of the competition MMORPGs, etc.  This theme also convinced me to avoid spending time letting the player change any of these procedurally generated factors.  This is why, for example, the player cannot unzone water, or change the names of player classes or monster types, simplifying the UI work dramatically.
• Substantial hustling to get the basic game framework up and running quickly.  I had the zoomable map interface and basic zoning within the first week.  The second two weeks were taken writing the user AI, and the final week was spent writing UI classes.  Most of the reports you see were written within the final 48 hours of the competition.
• Grabbing this game concept on the first day of the competition really meant that I didn’t waste any time trying to figure out what to do.  Being able to think about it for a few days before the compeition started also helped me hit the ground running, and get the most work possible done during the one month limit.
• Also, I happened to have some time off from work during the competition period.  This was purely by chance;  I’d arranged it long before the competition was announced, but it gave me some extra time to work on my entry, and that polish really showed up in the final entry.

What went wrong:

• The big “went wrong” issue here was that I hadn’t accounted for how long it would take me to implement the mouse control and UI features.  I wasted a lot of time trying to get the mouse scroll wheel to behave properly (needed input system modifications to support multiple ‘clicks’ per frame), and to implement a vector graphic UI system from scratch.  This UI implementation time thing really sucked up most of my time, and stopped me from tuning the MMORPG simulation nearly as well as I should have, and prevented me from implementing some very important “Quality of Life” features.

What’s still to do:

• The big obvious features that are missing are:  Fullscreen, changing video resolution, changing music volume (or disabling it altogether), and the big one:  save game support.  I’m working on these right now, and hopefully should have a 1.1 version within a week with all of these implemented.
• As for the game itself, there are some real bits that should get fixed.  Raising the box price of the game doesn’t slow people from purchasing the game as much as it should.  Raising the subscription price of the game makes people unsubscribe more than it should.  Class and monster attack/defense ratings aren’t taken into account when generating “too easy/too hard” complaints.  And the “Issues” report is much too vague anyway;  I need to find a way to really hilight the problem areas on the map view.  I’d like to revamp the way that roads work, so you actually get a real road network, instead of merely paint roads directly between two buildings.  User AI should be more aware of roads.  Users should have marital and social statuses displayed in the “Inspect” dialog box.   The game should generate names for regions and towns.  The player should be able to change the names of virtually anything in the game.  Maybe allow more in-depth construction of towns, or at least procedurally generate real town floorplans, instead of the current iconic “there is a town here” indicator.  Etc.

So overall, I’m extremely happy with how the compo edition of MMORPG Tycoon came out, but there’s still plenty of stuff left to do.  But I’m still actively working on it, and so hopefully there’ll be some exciting updates soon! 

Just a quick note for those who follow my front page articles (you know who you are), that I’ve made a number of crash fixes and game tunings to MMORPG Tycoon.  Fixes a couple of crashes, a number of niggling interface bugs, and completely overhauls player reactions to your MMORPG.  You can grab the updated build from the usual spot in the sidebar.

I’ll post a post-mortem once I’ve had a few days away from it.  Though I’m sure I’ll eventually return to this one to fix up some of the more obvious shortcomings (save games, etc)

Well, it was a bit of a push, but my entry for the TIGSoure Procedural Generation Competition is finally done!  (Actually, it was done about an hour ago.  I always forget how long it takes to update pages everywhere on the web)

Go grab MMORPG Tycoon binaries from the sidebar, under “Full Games”, and leave me feedback about what you like an don’t like from it!

So I’ve been doing a bunch more development on my MMORPG Tycoon project, and wanted to put up some new screenshots.

These screenshots were taken about twenty seconds into a new game.  You’ll see that I’ve got about 300 subscribers (most of whom are currently connected), and have earned a little money on top of the $1,000,000 starting loan (mostly from the purchase price paid by my subscribers).

The first shot is a zoomed in view, and shows the build toolbox which lets me place towns, respawn points, and starting areas.  Or move them (top icon), or remove them (bottom icon).

The second shot is zoomed out more, and shows me placing a mid-levelled area using the new zoning interface.  I’m not 100% happy with it yet, but it’s better than it was.  You can see the level range for each zone drawn in the middle of each region, for easy reference.

Most of the work has gone into UI stuff;  making buttons work as one would expect, etc.  But there’s also been a lot of work on closing the loop.  Players are a lot smarter, and are now unsubscribing when they’re unhappy, leaving forum posts even while they’re still logged in, and new players joined based upon how much buzz the game has, and how positive/negative it is.

So my next task is to find a way to communicate to the player what problems players are encountering;  as an MMORPG grows, it becomes harder and harder to figure out what the problems are.  And worse, once you’ve identified the problems, it’s not always obvious how to fix them.  I need to make players try to avoid heavily congested regions;  that might help some of the problem.  And of course, I still need to let the player place key quests, to move players around.

But in general, I’m pleased with progress.  And tremendously scared that I only have a week left to finish the project! 

So I’ve wired up my initial “Forum Buzz” calculation and discovered.. that while my plan for its calculation is probably a good mirror of what happens in the real world, I’m not convinced that it works in the context of a game.

Basically, here are the game rules as they currently stand:  Only players who are addicted to your MMORPG will generate any kind of forum buzz, on the theory that nobody else will bother to post on your forum.  The addicted player will add either positive or negative buzz, based upon their current satisfaction with the game.  (So if you’ve just nerfed their class, they probably will be generating negative buzz).  Buzz decays over time (currently at a rate of 25% per day.. I’ll probably increase this to 50% per day, but it needs more testing).  I then base the rate of new subscribers on a weighted buzz value which is equal to:  (2 * positive_buzz) + (1 * negative_buzz).

So as a rough rule of thumb, if you have two hundred people posting negative comments about your game, you’ll get half as many new members as if you have two hundred people posting postive comments about your game.  This works off the assumption that any sort of exposure will bring you more customers than no exposure, regardless of whether the exposure is good or bad (although obviously good will be more effective than bad).  This seems to make sense, and actually appears to be more or less what we see happening in real MMORPGs.  The problem is that I’m not sure whether this works well in a game format.

See, what it effectively means is “big games get more buzz, and more buzz generates more players for the big games”.  Big games get bigger, small games get smaller, almost regardless of whether the buzz is good or bad.  It really can’t be argued that that’s what’s happening;  certainly most of the new MMORPG players are going into the big MMORPGs like WoW.  But I don’t think that’s going to be fun to play;  it kind of makes having lots of subscribers become the only requirement for success at the game;  if you have lots of subscribers, your new subscriber base will grow massively, pretty much regardless of whether or not your player base actually likes what you’re doing, because either way, they’ll be generating a lot of buzz on the forums which will bring a lot of new players into the game.  This is, of course, assuming that you manage to keep your players addicted.  But so far, that’s actually not turning out to be at all difficult to do.  Maybe I need to improve the ‘addiction’ algorithms, to make that a little trickier to generate.

Need to think about this more.

Not much to say right now;  gotta get sleep.  But I wanted to post about the MMORPG simulation progress from the last few days.

I’ve very nearly “closed the loop” on the simulation (though not quite yet);  I’m hoping to finish it tomorrow.  Here’s the status:

• Users connect and disconnect from the MMORPG on an approximate schedule (this schedule slightly varies from day to day, and based upon what the player is actually doing while connected.  Users won’t disconnect in the middle of combat or if they’re really close to gaining a level, for instance.
• While online, users pick up quests, go hunting, gain levels, maintain a home town, etc.  Users also maintain a “satisfaction” value and an “addiction” value, which describe how happy they are with the game, and how much they’ll continue to play, even if they’re unsatisfied.  (And no, I’m not going to reveal how that’s calculated, yet.  )
• Each region (the outlined areas in the screenshots in previous posts) is run by a single server.  If too many users enter a single server’s area at one time, that server will crash, dropping all its players offline.  Which decreases user satisfaction a lot, as you might expect.  Keeping too many users from clustering together on one server is going to be one of the player’s big challenges in this game.
• Each simulated day, the player pays his employees one day worth of wage, and earns one day worth of subscription fee from all subscribers.  Additionally, when a new subscriber joins, the player earns the “box cost” he’s specified for the MMORPG.  With a $15 subscription fee, it takes 2000 subscribers to match the salary of one employee.  Or fewer, if you’re getting new subscribers.  The game plays with an accelerated clock, such that one day passes every two minutes of real time.  (I’m still fiddling with this, though)

The last little bit that I need to finish in order to complete the simulation feedback loop is to make users’ satisfaction and addiction to the game create a “Forum Buzz” value, and then set the frequency of new subscribers signing up based upon the Forum Buzz (adjusted by the box price).  Once I have that, the core simulation code will all be in place and hopefully working reasonably well!

Of course, I’ll still have to write a bunch of user interface code to let the player tweak with the various variables in the simulation.  And I need to write some sort of “Information” view that will show information about the objects in the simulation.  But that’s all detail work, once the core simulation is implemented. 

It’s been too long since my last post, so I’m just going to drop a couple more screenshots and stuff.

The first one is a zoomed-in view of one corner of a map, showing a few zoned regions. We’re close enough in that you can make out a few towns, a starting point, and a respawn point. (The towns are shaped like little houses, the respawn point is the octagon in the southernmost green region, and the starting point is the small square in the northernmost green region).

The second shot is a much more zoomed in view of a different map. This is showing the corner between four regions (three green, one yellow), and we’re close enough in that not only can we see the town and starting zone, but we can just make out the simulated users (the little white dots). These users are mostly all users who have this town set as their home; they’re currently doing quests and fighting monsters to gain levels right now.

I need to come up with a good way for the player to get information about these users. Right now each has a whole lot of data associated with them (what their AI is doing, the player and character names, player gender, health levels, satisfaction with the game, etc), but there’s currently no way to see any of that. I haven’t even written a way to let the player click on them yet.

But the most critical thing is to finish up the basic game UI first. I need to let the player switch between several different bits of functionality in the left toolbar, instead of always being at the “Zoning” tool. The player needs to be able to change the values for the many other input variables in my MMORPG simulation (subscription costs, monster densities, player class definitions, etc)

So there’s plenty of stuff still to do!

Okay, I’ll confess that I’ve been a little intentionally misleading for the last week. 

The MMORPG game I’m working on isn’t an MMORPG;  it’s a single-player game about MMORPGs.  The player takes on the persona of the new owner of a software company with an unpopular MMORPG, and it’s up to the player to bring them back into profitability.

The player does this by adjusting the zoning of the map, by adjusting the player classes available, placing key quests, adjusting subscription costs, hiring and firing programmers/designers/artists, watching the forum buzz and news coverage about the game and company, etc.

At the moment, the very basic MMORPG mechanics are implemented;  the player can set up zoning, simulated players log in to the game, fight monsters, and pick up and complete minor quests.  I’m hoping that by the end of tomorrow, I’ll have simulated players accepting key quests, and then can move on to tuning how all the various parameters and systems interact with each other.

Just for fun, here’s a very early screenshot of my MMORPG game project.  Apologies for the lack of anything really visible in it;  almost all the work thus far has been behind the scenes.

This is a somewhat zoomed in view of the map screen.  The toolbars and suchlike haven’t been filled in yet, but will eventually provide a number of controls.  The user can drag with their right mouse button to scroll around the map, and use the mouse wheel to zoom from a full-world map to a close-up view of the action.  Again, this is a bit like World of Warcraft’s map, but with smooth zooming control.

Each square you see in this view is a “zone”, which would be about a five minute trek from one side to the other.  Theoretically, if this was a full-on commercial MMORPG, each zone would be simulated on its own server to manage player load across the world.  In practice, I’m sure that I won’t actually have to worry about things like that. 

Not visible here, I’m using the “Meat” sprite from Muncher’s Labyrinth for players.  But that’s probably only temporary.