GDC 2013 clewis1's notes

Post date: Apr 02, 2013 9:48:53 PM

FTL Postmortem: Designing Without a Pitch

There were a few cool things I gained from this talk.  The first was further support that starting with a single core experience and making sure that all the design decisions support that core experience can result in a pretty darn successful game.  With FTL, the developers wanted to capture the feel of being Captain Kirk at the ship helm, making high-pressure strategic decisions about things going on inside the various ships.  With every iteration of FTL, they evaluated what was fun, what wasn’t, and how the game could further match the core experience.  A lot of initial design features were scrapped during the process because of the answers they got from these questions.  However, they had an extremely thin design doc, so the potential risks involved with moving from one design state to another were low.

In fact, the primary thing I took away from the FTL talk was that while it’s important to have a strong understanding of the abstract/emotional experience you’re trying to instil with your game, it’s useful to keep design docs and specific expectations of how a game is going to work to a minimum.  The more work investment you have in a particular mechanic, the harder it is to remove it from your game when the mechanic stops supporting your core experience.  It became quite a bit harder for them to make design decisions that responded to the game rather than their fan’s response once they had their successful kickstarter.  More people had expectations of what the game was supposed to be, and they ended up burning several weeks on features that they knew they wanted to get rid of from the beginning because of fan response.  On the flip side though, they talked about several instances where fan response (or lack thereof) helped them save time in the polish phase.  Several placeholder animations from earlier prototypes ended up in the final version of the game because none of their audience viewed the animations as out of place.

SpaceChem Postmortem

At the beginning of the SpaceChem talk, they discussed three types of single player puzzle games.  The first is puzzle games like Portal, where each puzzle only has a very small number of carefully designed solutions (usually just one).  The second type was puzzle games without a solution (or ending), such as Tetris and Bejeweled.  The final type was what they called design-based puzzles, where a puzzle can be solved to varying degrees of success.  The value of these design-based puzzles is that you can feel a discrete sense of accomplishment by completing a puzzle, but you can always return to a puzzle to do better job later on.  To encourage the player’s desire to improve their solutions to puzzles, they added histograms to the results screen to get a sense of how well you did overall compared to everyone else who played the game.  A nice side effect of having the puzzles be design-based was that as long as you could prove that the puzzles were at the very least solvable, you didn’t have to actually beat them.

The two main shortcomings they discussed in their presentation were a lack of visual sophistication in the art, and having the tutorial be too long.  Their view on tutorials was that games should never tell the player how to think, because that’s no more fun than listening to a lecture.  Instead, tutorials should be set up so that there is a relatively high success rate for solving the problems through trial and error, so that making the mental connections and understanding of the system is left up to the player.

Free-to-play, in-app, ad-supported analytics

This talk focused on the financial numbers of a free-to-play iOS game.  Given the nature of the indie summit, the audience wasn’t too thrilled with this talk, including myself.  I didn’t really gain any valuable insights from the talk, but here’s the notes I took from it:

Talking about Chip Chain, a puzzle/casino game

made a little bit of money

IAP helped a lot financially

This talk is really depressing

Put things on sale, perceived value can be based off of prior prices

US is paying more for vegas puzzle games than other countries

Interstitials - cross promotions - full page game recommendations

About 5 seconds were spent on talking about the game, rest is financial analytics for a low-grossing game and how it's a good model for stuff.  I mad

"Encouraging people to spend is a game design problem" aaaaaaaaaaaaaaaa D:

This makes me so mad

Advertising potential purchases works

This is like the opposite of the FTL presentation

There's a good chance you can make more money by making your better than spending most of your time juggling around the financial model (this was not brought up by the speaker)

Talk is just about results of how certain pre-existing financial concepts worked, nothing new, just numbers

There's a slide titled 'buying users'.  I'm serious.

WHY NOT JUST MAKE GOOD GAMES FFFFFFFFFFFFFF

My blood pressure rose more than my understanding of game development

"We totally overdid the analytics" - darn straight you did

"Didn't have time to write a slide about a thing" - good, I want this talk to be over already

There’s room for professionalism in my notes.

Indie Rants

This was probably the best part of the indie games summit.  They started off with Steve Swink talking about the current state of the IGF, including the new voting systems that had been built for it and statistics of just how many more games the judges are actually playing and evaluating.  The most active judge, Tim W, had played over 600 of the games submitted to the contest.  That's pretty darn impressive.

Next up, a programmer who worked on Fez talked about all of the crazy things that went wrong with the release of Fez, including a patch that accidentally corrupted save files.  The main takeaway from his talk was that things inevitably go wrong; make sure you're prepared for the worst.

Bennet Foddy gave one of the most entertaining talks of the day after that; discussing why freemium games suck and ways to make them better.  I relished in the fact that this seemed almost like a direct attack against the Chip Chain talk (in retrospect, I should have brought popcorn).  To summarize, existing models make you look bad to your audience and make your design suffer.  He suggested for games that want to go freemium to integrate the monetization in a way more meaningful and honest to the design and players, such as pay to have your score ranked (tournament model), never pay if you play well (reward model), and pay to change the game for everyone (builder model).

Chris Hecker talked about how nobody knows about your game, and used his own game Spy Party as an example.  He pointed out that all of the really successful "AAA" indie games had a long-term slow-burn grass-roots campaign.  A key quote was that "you cannot overhype, you can only underdeliver."  As a developer of Spore, Hecker understands this very well.  Still, when your marketing budget is zero as an indie, you want to hype yourself as much as you can.  A developer from Vlambeer later reiterated that indie developers really need to get people to play their games more and bring themselves more to public-facing events like PAX instead of just developer-facing events like GDC.

A creative director at Linden Lab talked about the value that text still holds in an ever-more visual gaming medium.  Aside from text content being way cheaper than graphics, it can communicate personal subjectivity more effectively than games with graphics have so far.  I saw this more as a challenge than a constraint, though.

Disasterpeace gave a talk about how music in games can be designed to be more organic and concert-like instead of playing the same track on loop forever.  To him, music gains meaning when we let it die.  I think that generative music can be really effective for certain game experiences, but it all depends on the game.

Miscellaneous Things I Learned

• Alcohol is networking

• Every new person you talk to at GDC is worth it

• Don't spend a billion dollars on business cards