NDL party posted 3/14 at 1:58:00 PM PST by Sande
I just want to say: I really dug the invites! It's a little lizard inside an egg drinking a martini. It was put into one of those film negative envelopes so it would be like an ultrasound. It was a way of promoting gamembryo (which is a hard name to say). The women from the NDL booth remembered me and let me in with my friends, the aforementioned Brian Hawkins (Soma Consulting) and Ivanna Kartarahardja. The NDL party was in the room next to the Microsoft Women Celebrating Women in Game Development party. Food from Il Fornaio, this party was really nice and not loud like the Suite Parties.
Women in Game Development Group Gathering posted 3/14 at 1:51:39 PM PST by Sande
Well, I missed those guys doing their session on independent game development because I went to the Women in Game Development Group Gathering! There were many people, some of whom weren't on the list, but it was great to attach faces to names. Most of this gathering was focused on: what next? Should the mailing list be partitioned? Do we need a webpage? Or maybe some kind of mentoring program? Of note, there are guys on the Wigd list and I saw Eric Cha of Xenopi Studios, Inc. hiding behind some women!
Learning Goals in Sports Games posted 3/10 at 10:07:51 AM PST by Kevin Hawkins
With my interest in AI and sports being equally satisfied, I decided to attend the "Learning Goals in Sports Games" lecture by Jack van Rijswijck of the GAMES Group from the University of Alberta. The gist of the lecture was figuring out how to have adaptive AI in sports games that learned from its own mistakes. The particular approach presented involved using EA Sports' FIFA soccer game with stripped out graphics and sounds so the focus would remain on the AI.
The lecture began by defining AI. According to Jack, various developers define AI as "anything that's not graphics". He defines AI as "any decision the gamer would make". When thinking about it, this definition provides a strategy for developing a good sports game AI.
Next, the benefits of learning AI were discussed. The primary benefits included:
Potential concerns were also raised, and included learning the wrong things from bad examples, learning unnatural things, too much processor power, and difficulty in testing. Jack gave an example of a tournament he had with various AI type's playing FIFA soccer while he was doing his research. The AI that won the tournament had a rule where it would immediately kick the ball as soon as it had control. Obviously, this is something that needed to be addressed.
So, the idea of his method is to optimize fun, but not necessarily wins for the AI, within realtime and hardware constraints. The challenges involved in sports games in particular is that the AI characters don't die, the sports games emulate real world sports and not events that have been fabricated for entertainment, and that the sports games emulate specific athletes and teams with their strategies.
Typically, AI is developed using states, but van Rijswijck proposed defining behaviors in contrast to the states (interestingly, a statechart defines behavior with concurrent state machines). The reasoning behind this is that behaviors are less rigid and more fluid, they are more natural, they are easier to extend, and there are more opportunities for learning.
Finally, the force field model was proposed as a way of handling AI in sports games. In this model, forces are pushing and pulling each player to help determine where a player needs to go on the field at any time during play. In the soccer example, the forces are placed in the spots of the ball's trajectory where a ball might have been intercepted after a scoring play (for defensive AI). In these areas, the force becomes an attractor, pushing everyone toward the trajectory of the ball.
The implementation involves a discrete grid of points with two inputs: the ball position and the player position, where the force fields are stored for each position. After each iteration, a new force field is calculated and additively combined with the main (current) force field, so in the end you only have one field.
There's a little bit more to this than I explained in this summary, so if you're interested in learning more about sports game AI, I suggest checking out the GAMES Group website at http://www.cs.ualberta.ca/~games. I don't know if Rijswijck was planning to post his slides to the site, but there is plenty of information there on sports game AI research. Overall, this was probably the best lecture I attended in terms of a balance between technical info and enjoyment, and if it was given again I'd highly recommend it.
My and Mason's Lecture posted 3/9 at 6:45:53 AM PST by John Hattan
3pm Friday was the time for the lecture "Leveraging Online Resources to Make it as an Independent Game Developer" run my Mason McCuskey and myself. At first I was worried, because when I asked the first two guys who sat down if they worked for a small company, they replied "the biggest". That worried me, because nobody from EA or Sony would be much interested in a trick that can save you $30 on a copy of Paint Shop Pro.
Anyway, it was rather a 2-lectures-in-one deal because our material was quite different. I started out with a pile of products and hints for doing your project on the cheap. Things like free compilers, free paint programs that don't suck, cheap install programs. Stuff like that. Mason then followed up with a higher-level lecture on how to run a small project, write press-releases, find cheap help, etc. People were taking notes more furiously than I'd seen in other lectures, and lots of folks talked to us at the end, so I think we were well-received. I think next year, though, I'm going to pitch our lecture as a roundtable discussion, as there were several other folks who had their own suggestions for cheap products and such. If everyone could've better volunteered their favorites, it would've been even better.
Besides, "The Cheapass Roundtable" has a nice ring to it :)
Memory Optomization posted 3/7 at 11:59:44 PM PST by Gaiiden
This was another tutorial that bombed for me. The lecture itself wasn't too bad, it's just that the slides were very visual - there were a lot of charts and code samples and such that after a while I just gave up taking notes because there was no way that I could keep up. The fact that there were no handouts of the more visually complex slides was kind of annoying. Christer Ericson (Sony) announced in the beginning that he would make the slides available online, so if you're interested in memory optomization with a focus on the cache, check out this site for a copy of the slides. (Christer says they'll be up there eventually, so keep checking)The slides alone will be able to teach the lecture to you. Heck I'm getting the slides.
IGDA Annual Meeting posted 3/7 at 11:46:22 PM PST by Gaiiden
The IGDA's Annual Meeting was held this afternoon. The Annual meeting is the time for the IGDA Board of Directors to come together with the rest of the IGDA community and review the year and discuss plans for he coming year. The meeting was kicked off by Graeme Devine as he gave his retirement speech - he's stepping down as Chairperson this year.
After Graeme's speech, Jason Della Rocca took the podium to present the annual report. In short, IGDA membership rose by ~70% during the year, and 20 new chapters were formed worldwide, bringing the total to ~60. New partners and studios also affiliated themselves with the IGDA. The IGDA's various industry actions were also discussed:
Next the floor was opened to discuss several topics related to the IGDA, followed by the Board of Director Nominees speeches.
GarageGames Quit Your Job Fair posted 3/7 at 11:18:13 PM PST by Gaiiden
GarageGames was holding a self-sponsored session call Quit Your Job Fair: Can you Make a Living as an Independant? The session was run by Jeff Tunnell, the man behind GarageGames and a forerunner in the independant games community. The room the session was being held in wasn't packed to the max, but there weren't many empty seats either.
Jeff started off the lecture by ensuring everyone in the room that their quest was valid - you can make a living off being an independant developer. We're dealing with a $10 billion industry now after all. If you want a chunk of the profits there are three ways you can go about it:
Why Go Indie Now?
After Jeff was done breaking down your options into the industry he went over why it's more attractive for people to go indie today than it was a year or two ago.
How to Raise Money?
Next Jeff went over the biggest problem facing the indie developer - money (or funding, more specifically). The simple truth is that if you go indie then you should expect to self-fund pretty much all your games. Venture capitalism is simply not an option anymore, expecially with the economic downturn, so finding money can be the toughest part of starting independant development. Jeff strongly cautioned against taking money from friends and family, because most likely you won't be able to pay them back. So we're still stuck aren't we? The answer is to not quit your day job (at least not immediatly). Many people back up their decision to go indie by quitting their normal jobs. Maybe that makes them feel more independant but in reality they'll more often than not end up going back to that job so they can have money to live off of while they continue their independant development. Keeping your job, at least until you start generating some revenue, will just ease the stress and the risk involved with striking out on your own. You may end up selling your nice car (Jeff had to go from a BMW to like a Saturn or something) or living with your parents again, but Jeff made the point of saying that you can always make it work - you can find a way.
Can You Make a Living?
Looking at companies like Pop Cap, Small Rockets, Game House... all of them are highly successful independent companies that have put out games like DX-Ball, Collapse, and Combat Mission. Jeff even quoted some figures from a GarageGames' game called Marble Blast, which sold 1,000 units in three months. Jeff predicted that the game would make $150k - $200k throughout the course of its lifetime based on the unit sales.
What Does GarageGames Offer?
Jeff went on to explain the GarageGame publishing options. He said how he wanted to turn the traditional publishing method on its head, so GarageGames offers its developers royalty rates starting at 65%. In addition, if they think you can help out with Torque engine development or if you contribute to the community Realm Wars project, then GarageGames will consider increasing your royalty rates even more. Not only that, but you, the developer, get to keep your Intellectual Property (IP). This gives you the freedom to market and sell and do whatever with your game characters and any franchise that you may develop. GarageGames also offers various web tools for when you publish your game, such as bug tracking, access to your sales page, forums, etc.
5-Step Game Plan
It was a great session and I stopped off by the GarageGames booth later in the day to check them out. Check the Expo page - I'll put up shots tomorrow, cause I just realized I forgot to get some today, heh.