Archive for June, 2006

Final Casuality Notes

Friday, June 30th, 2006

I took fewer notes on the sessions I attended the last two days of Casuality. There were more panels and fewer individual speakers, and the panelists in general were giving out less chunky data nuggets (and more soft opinions) than the individual speakers. Here’s some stuff I did note down: (From various speakers and various sessions – I don’t remember in all cases who said what)


  • $5.99 is the average game sales price
  • 65% of revenue from women
  • Try Before You Buy is starting to hit mobile (10 minute trials versus 60 minute for PC), but it’s not THAT successful, because the games are often poor and users thus don’t want to buy.
  • 1/3 of mobile sales are puzzles
  • 75% are, more broadly, ‘casual games’
  • Other:

  • Lots of talk of bringing in more ad revenue, in part to make more money from the 98-99% of gamers who don’t buy (i.e. only doing the first part of ‘try and buy’).
  • AOL rep. said that AOL does not share (and apparently doesn’t plan to share) ad revenue from browser-versions of downloadable games, but DOES share ad revenue from games built exclusively for web. His excuse is that the former are designed mainly to sell the downloadable versions of the game and don’t generate many ad impressions, anyways. The crowd at this session was not terribly happy with his statements/position.
  • The AOL rep. also said that downloadable games with browser versions sell much more than those without (I think he said 3-5 times more but I may be mistaken). My take: First, probably only the higher quality/bigger budget casual games HAVE a corresponding downloadable version, and second, his statement also seemed to be part of his justification for not paying ad-revenue sharing on these games (i.e. – see, you’ll sell more downloadable games, so you don’t need ad-share). So I take his figures with several grains of salt.
  • Casuality – Shwag

    Friday, June 30th, 2006

    I’ve been going to GDC most years since 1995, and the quality of the shwag (free stuff) there has fallen from mediocre (free t-shirts, decent parties with cold-plate appetizers and beer) to almost non-existent (t-shirts that you have to compete for or stand in a 20 minute line for).

    Casuality is a much smaller conference, and perhaps that’s the reason the shwag there is vastly better than GDC.

    Most of the vendors/publishers with tables set up (~10 or so) were pushing free t-shirts or something comparable – almost throwing them at you. Trying to control the volume of stuff I was carrying around made me pretty selective though. There were also frisbees and other goodies.

    I only went to one party – the PopCap party. About 200 people were there, and it was nice – excellent appetizers, open bar (fancy drinks, not just beer), and a gift bag for each attendee with about $50 of assorted free stuff.

    But I missed the party with the best shwag – by Oberon. At Oberon’s party, they gave away a free Mini-Cooper (yes, the ~$20,000 car). There were apparently about 250 people at the party, so the cost to Oberon of that one party favor was almost $100/guest.

    Casuality Conference – Day One

    Wednesday, June 28th, 2006

    I flew into Seattle yesterday for the Casuality conference – the main U.S. conference focused on casual game development. Some impressions:

    Industry Size/RealArcade Size
    The first interesting data points came not at the conference, but in an article about the conference in the local paper (Seattle Post-Intelligencer – more articles here and here) that I read in the cab on the way over. The article quotes a DFC analyst saying the U.S. casual games market will grow from $314 million in ’05 to $458 million in ’06. That’s on the high side of other estimates I’ve heard, and shows really strong growth.

    Also in the article is the factoid that Real Network’s Q1 revenues from games were $18.6 million. If we annualize Real’s game revenue and assume a bit of growth for this year, we get to about $85 million for Real’s ’06 games revenue. But, according to Real, Europe accounts for ~25% of their revenue, and they have an Asian presence, too. So, let’s say only 70% of their revenue is U.S. That’s about $50 million for ’06, or about a 12% market share, if the DFC analyst is right. My general impression is that Real is even more dominant than that in the U.S. downloadable space – perhaps 25%+ market share, so I’d guess either the DFC analyst is overly optimistic with his overall revenue estimates, or that his industry definition is broad (i.e., including mobile games).

    Initial Show Impressions
    The show itself is in downtown Seattle, at the home theatre of the Seattle Symphony, which apparently does a side-business in conferences during the week. The main speaker for each time slot spoke in the concert hall – it was a very different vibe from the typical convention held in a utilitarian convention hall or hotel.

    I had been expecting as many as 1,000 people at the show, but my eyeball guesstimate was lower – maybe 400-500 people. Seattle is the home of a number of casual game companies (Real/Gamehouse, BigFish, PopCap, and others), and each of the biggies seemed to have about 50 people at the show, so there was sort of a lopsided representation of just these few companies.

    I arrived during lunch break, and fortunately stumbled into a crew of regular posters from the IndieGamer forums, mostly smaller indies, so I felt a bit connected right from the start.

    After years of seeing GDC grow to massive size, it was nice to be at a conference small enough where I recognized a fairly high percentage of names on the name tags (even though of course I’m terrible with faces, even for those who I have seen in person before). There was the usual bit of wandering around the main gathering area, squinting at people’s name tags without trying to be too obvious about it (’cause in some cases it might be a person I really should recognize by face and not just by name tag…)

    The first session after lunch was given by Rob Glaser, CEO of RealNetworks. Some data points he gave:

  • Users download 750,000 games a day from RealArcade
  • 2% of downloads result in customers paying, either for the full version or a subscription (I’m pretty sure this does not mean a 2% average conversion rate by the typical definition – not sure how much of that is the subscription model)
  • He wants to monetize the other 98% – announced a plan to integrate ads between levels for the free 60 minute trial versions of games (based on a Zylom’s Clicktopia system)
  • 70.5% of games purchased by mobile phone gamers are what he would define as ‘casual games’ (i.e. Tetris, Bejewelled, Zuma, and Luxor are all big hits on mobile)
  • Problem with mobile phone development is proliferation of handsets, OS’s, etc.
  • He announced EMERGE ( It sounds like either a fancy framework or possibly a new language. It wasn’t entirely clear what it does – I think it will compile out a single source base and target the 10,000+ permutations needed to fulfill mobile carriers’ different hardware/software/network needs.
  • Next session was “Beyond the 60th minute” – a game design panel with Jason Kapalka of PopCap, and a guy from GameLab, and a guy who designed Tradewinds 2 and Tradewinds Legends (sorry – missed those names)

    Mostly a bit ‘touchy/feely’ – not so many hard data points so I didn’t take as many notes, but here’s a couple:

  • Someone made the analogy that casual games today are like arcade games back in their golden age. They’re almost free to try ($.25 for arcade games, a 5 minute download for casual games), so the game designer has to hook the user immediately, or he will move on. But the good thing is that the low cost of trying new games means that IP/branding is of lower importance – you don’t have to have a big movie license to get users to try your game. Since it’s free, they’ll try most things, but then stick with (and pay for) the ones with compelling gameplay.
  • There was general agreement that a $20 casual game should provide a minimimum of 5-6 hours of gameplay.
  • This session spilled over into a production session that I only caught the beginning of (had to leave for a meeting). Most meaningful data point:

  • Somebody from BigFish implied that dev. cost for Mystery Case Files : Prime Suspects was in the $500-600K range (i.e. much higher than the ‘standard’ ~$150K dev budget that had been prevailing for high-end casual games).
  • I went to one last session – a ‘State of the Industry’ panel with the honchos from many of the major players (PopCap, Microsoft, Pogo, BigFish, Real, Boonty, IIRC) It was an interesting panel, but again, rather ‘touchy feely’, and I didn’t write down any real data nuggets/notes from the panel.

    That’s it for today – I went to the PopCap party after the show and had a good time, but retired early (woke up at 6 am CST, 4 am Seattle time. Hopefully tomorrow I’ll have more on the show itself, and perhaps some of the more interesting (and non-confidential) water-cooler talk.

    Warren Buffett steps up to the plate.

    Sunday, June 25th, 2006

    Following last week’s announcement that Bill Gates is (slowly) retiring from Microsoft to focus on his foundation, comes news that Warren Buffett is giving away about 85% of his net worth (roughly $37 billion at current prices) to charity.

    That’s cool, and slightly newsworthy (he had originally planned to donate his fortune upon his death, not earlier).

    But what’s really interesting is that he’s giving 5/6 of the money to:

    Bill and Melinda Gates’ Foundation.

    I’ve followed a bit of what the Gates Foundation is doing, and I’m pretty impressed. Obviously, Buffett is, too. Most super-rich in Buffett’s position would want to form their own foundation, with their own name, for ego purposes. That Buffett is willing and able to put his money into someone else’s foundation is cool. And pretty smart too. Buffett is 75, Bill Gates is 50 – Buffett obviously trusts that Bill (and Melinda) will be solid managers long after he is gone.

    Bravo Mr. Buffett.

    (Original Slashdot blurb, which linked to longer CNN story)

    VS 2005 – Inferior to VS 2003?

    Thursday, June 22nd, 2006

    I just installed a trial version of Visual Studio 2005 (I had been using the previous version, 2003).

    Theoretically, new compilers can provide the following advantages (that are significant to me, anyways):

    1) Better interface/tools
    2) Faster compiles
    3) Smaller compiles
    4) Faster code

    Item 1 will take a while to evaluate – I’ll use the new compiler for a few weeks and see what I think. But 2-4 are fairly easy to test. I tested all of the following on my current game, which is fairly similar in scope to Bonnie’s Bookstore.

    Compile Time (Full Solution Rebuild):
    VS 2003_____:20_________:30
    VS 2005_____:26_________:34

    Compile size, KB (Release Mode Only)
    _________Game_EXE________DLL____Total ____Total (zipped)
    VS 2003_______988________340_____1328_______574
    VS 2005_____1,016________612_____1628_______717

    Code Speed (Release mode - milliseconds to do 30 passes through my main draw loop, EXCLUDING the blit to the video card)

    VS 2003___335*
    VS 2005___358*

    *These are averages of 10 attempts on each version, and there was significant variability between tests, with a range of 287-375 for the VS 2003 version, and 299-396 for the VS 2005 version. Also, I didn’t hand tweak the optimization parameters for VS 2005, but left them as they were imported from my VS 2003 project. Possibly, more tweaking would improve the 2005 results.

    All tests on my main dev box, with a dual-core AMD 4600+ CPU.

    I’m surprised that the build time got noticeably slower in 2005 – In theory, the new compiler will use both cores for compiles, which should nearly double the compile speed. Perhaps I don’t have the option properly enabled, but I looked for info on it and couldn’t find any way to explicitly enable or test that feature.

    Compiled size is another minor drawback. Since my games are built for download, I like to keep the EXE (and accompanying DLL – i.e. MSVCR71.DLL or MSVCR80.DLL) small. This isn’t a huge issue, since, with graphics and sound, the download will likely be in the 13 MB range. Still, it’s a minor consideration, and perhaps more important for those shooting for very small download sizes. (When I did the Active X version of Bonnie’s Bookstore, I was targeting an 800KB initial download, with further data streamed. OTOH, that version used a static version of the C run-time, and given that most of the extra bulk is in the run-time DLL, the issue may not be THAT severe)

    Finally, VS 2005 apparently produces slower code than VS 2003, though possibly the results are just due to noise/variability.

    Overall, the raw specs of 2005 look inferior to 2003. Still, if the new version enables me to be noticeably more productive, that will dwarf the comparitively small differences in the above metrics. Hopefully I’ll post an update in the future on the interface differences/improvements.


    Edit: I just spent a bit more time playing with the various code optimizations in VS 2005, and now I’ve got the performance to where the draw loop executes about 5-6% faster than in VS 2003, at a cost of slowing down the release mode full rebuild from 34 seconds to 50 seconds, and increasing the file size a bit (from 1016 KB to 1048 KB).

    Held Hostage

    Thursday, June 22nd, 2006

    I hate signing up for monthly recurring services that are automatically billed to my credit card. They’re like the old Roach Motel – once you check in, you can’t check out.

    One person recorded his call to cancel his AOL service, stubbornly resisted by the AOL rep. (Full transcript here, originally seen on Slashdot).

    Key Excerpts:

    VINCENT: I don’t know how to make this any clearer, so I’m just gonna say it one last time. Cancel the account.

    AOL: : Well explain to me what’s, why…

    VINCENT: I’m not explaining anything to you. Cancel the account.

    AOL: Well, what’s the matter man? We’re just, I’m just trying to help here.

    VINCENT: You’re not helping me. You’re helping me…

    AOL: I am trying to help.

    VINCENT: Helping… listen, I called to cancel the account. Helping me would be canceling the account. Please help me and cancel the account.

    AOL: No, it wouldn’t actually…

    VINCENT: Cancel my account…

    AOL: : Turning off your account…

    VINCENT: …cancel the account…

    AOL: : …would be the worst thing that…

    VINCENT: …cancel the account.

    [It goes on further…]

    I’m normally a pretty calm guy. I think the last time I lost my temper to a salesman was a few years back when I was trying to buy a video camera at BestBuy, and the salesman WOULD NOT SHUT UP about the extended service agreement.

    Over the course of the conversation, his tone shifted significantly from the initial sale (“This is a great camera…”) to his push for the extended service agreement (“This camera is junk and will almost certainly break on you one month after the basic warranty is up”).

    I was about as clear as the guy in the above conversation, probably 8 or 10 times, then lost my temper with the salesguy and started peppering my refusals with words that would make my mother blush.

    I should have just walked out, but the camera was ~$50 cheaper there than down the street, and really, how hard should it be for a salesperson to leave a customer who’s buying a $900 camera alone?

    Play Your Game As Your Users Will

    Tuesday, June 20th, 2006

    I’ve been designing games for over a decade, and I always fall into the same trap:

    I come up with a game design, and a list of things I need to do to implement that game. Then I focus so much on adding these features and testing them in isolation, that I forget to test the whole game.

    I spent 10 minutes playtesting the new game today (testing it as a whole, not the little bit that I was working on), and came out of that short session with a list of 15 more to-dos, all of which will have a vastly more positive impact on the game than the tasks I added to my spreadsheet weeks ago that I was plowing through.

    Sometimes, the issue is even more obvious. When I’m programming the game, I always run it in windowed mode, because then it’s easier to debug. But as soon as I switch to full-screen mode, I see little graphic glitches and artifacts that aren’t noticeable in windowed mode.

    Play it the way the users will…

    Bill Gates’ New Job

    Thursday, June 15th, 2006

    If you haven’t seen, Bill Gates just announced that by 2008, he’s stepping down from day-to-day involvement at Microsoft to work full-time at his charity, focusing on world health, poverty and other issues.

    I say Bravo Bill.

    Whatever I may think of the actions of Microsoft over the last three decades, I applaud Bill for focusing both his money and his time on trying to make the world a better place, and for doing it while he’s still young enough (50), to really get involved and make a (hopefully) long lasting, positive impact.

    If Bill can apply the same level of intellect, energy, and financial resources that he did at MS against some of the relatively intractable problems of the world, I suspect he will make more of a difference than a thousand government bureacrats.

    Game Design Drift

    Wednesday, June 14th, 2006

    I’ve been working these last few months on my next casual game. The new game is a match-3 style game with what I consider to be a fairly interesting twist on it’s mechanics.

    This time around I’ve been a little more close-to-the-vest with the game than I was with Bonnie’s Bookstore. The new game is within a more competitive genre, and the key innovation of the game could easily be copied by others.

    I know, I know, who would ever copy/clone a game concept in the casual market?


    The downside of this is that I haven’t been seeding alphas around like I did with Bonnie’s Bookstore. Two months into BB’s development, probably 50 people had seen the game, but for the new game, at about the four month mark, I’ve shown or sent it to less than a dozen people. (One of those people was my 6 year old daughter – she liked it, and was better at one of the mini-games than me.)

    When I don’t get regular feedback on what’s good or bad about the core elements of the game, my focus tends to drift. It doesn’t take long for the game designer (me) to get bored with the basic game, and waste, err, allocate time to peripheral aspects.

    So I’ve done some quasi-goofy stuff:

    1) Added 7 mini-games aside from the main game
    2) Created a global high-score system, rolled from scratch (teaching myself a bit of PHP along the way)
    3) Added an overly complicated flame system to the game, used for one small-ish effect (a torch and something it burns)
    4) Added an overly complicated system to track the beat of the background music for the game, in use for a single interface element (twin flowers which vibrate in time with the beat).

    None of these things really add anything to the core game. Still though, while each individual element seems like a slightly foolish waste of time, in the aggregate, I think these (and others) do add value to the game – taking it from an interesting game mechanic with pleasant graphics, to, hopefully, a polished game that hooks the player in any number of ways.

    My reference point here is Super Collapse 3, recently released by Gamehouse. It doesn’t add much, gameplay-wise, to Super Collapse 2, but the overall package is a lot of fun. It has a better UI, more mini-games, a great quest mode, some splashy particle effects, and lots of other things that kept me hooked on the game through the 60 minute trial, got me to buy the game, and play it to completion (~5 hours of gameplay).

    There’s a thoughtful interview here with indie developer Svero, a longtime poster on the indiegamer forums, and incidentally, one of the handful of people who’ve seen my new game.

    Knobs > Digital, part 2

    Friday, June 9th, 2006

    This evening, we were making cookies in our new oven (initial discussion), and the oven shut off about 2 minutes before the cookies were done, due to some confusion in using it.

    How could an oven be confusing? Turn it on, insert cookies, monitor cookies until done, remove. Easy.

    Here’s how an oven display SHOULD be designed. This is our toaster oven, which is what I like to use. Twist the knob to the desired temperature. The light goes on. Remove your food when done and twist back.

    Toaster Oven Knob

    Here’s our new oven:

    New Maytag Buttons

    Whereas it only takes 1 knob to interact with our $40 toaster oven, it takes 11 buttons to interact with our $400 oven. I guess they figure they need about one button for each $40 in price. I’m glad we didn’t get the $800 model…


    This morning, another poorly designed button interface nearly cost me $180. I had borrowed a power washer from a neighbor, which fortunately only had 4 buttons. On and off, on the unit itself, and Test and Reset, on the oversized power thingie you plugged into the wall. Understandably, having a built-in circuit breaker is a necessary thing on an electrical device meant to spray water all over the place.

    After using the device successfully for a while, I moved it to a different location and plugged it in again. Nothing happened – the engine didn’t come on. I tried various combinations of toggling the on/off, the test/reset, trying different wall outlets, etc – nothing. I left it alone overnight, figuring something had gotten wet and it needed to dry. This morning – still no luck. Had I broken it? Was I on the hook to replace the $180 device for my neighbor?

    Finally, I called the manufacturer’s help line #, helpfully printed on the unit. Although it was an automated system, it actually got me a useful answer fairly quickly. Turns out I had to press and HOLD the reset button for 4 seconds. Yes, being a computer geek, I suppose I should be used to press and hold. But I usually work around it on my PC just by unplugging and replugging it (as I had tried with the power washer.)

    Consumers are used to pressing a button, then releasing it. Any button that requires you to press and hold it for multiple seconds is broken.


  • Don’t use 11 buttons when 1 knob will do.
  • Don’t try to turn a single-state button into a multi-state device (normal press, and LONG press).
  • Use the right device, in the way that your customers are used to.
  • =========================================

    P.S. – After I wrote this, I realized our toaster oven has two other buttons – ‘Toast’, and a toaster darkness setting (a knob, fortunately). Still, I think my success rate with the toaster oven is about 99.5% (the other .5% is me turning the toast darkness setting and forgetting to press the button).

    P.P.S – When I cropped the pictures above, I considered cranking up the contrast so that you could actually read the words on the oven buttons clearly. But then I realized that the button contrast is part of the problem. Note that the toaster oven manufacturer, who probably had the factory owner’s high-school graduate nephew design the interface, realized that labels need contrast – thick lettering – white on black or black on white. The oven manufacturer’s buttons look sleeker, but thin gray letters on a tan backgrounds does not make for easy readability…

    Casuality Conference

    Wednesday, June 7th, 2006

    I will be attending the Casuality Conference on Casual Games will be in Seattle June 27-29.

    If you’d like to set up a meeting with me there, drop me a line (PSTEINMEYER A T NEWCRAYON D 0 T C0M.

    WordPress Deferred Posting

    Wednesday, June 7th, 2006

    Here’s a tip for all you WordPress users out there (this blog runs on WordPress). You can write a post and have it automatically posted at a later time by changing the ‘Post TimeStamp’ in the lower right.

    Now when I’m in a writing mood (like now, 11:21 pm on a Tuesday night), I can write a post and have it appear, say late Wednesday, thus not crowding out the post I made 15 minutes ago and making it appear I’m writing this blog on a semi-regular basis, rather than bursts of posts all at once, then nothing for a few weeks…

    More Badly Written Articles

    Tuesday, June 6th, 2006

    …or, the difference between 80% uptime and 99.78% uptime.

    Last week, a badly worded NPD report led various web sites to say that NPD had pegged Casual Gaming as a $52 million industry (in fact, NPD was only talking about subscription revenue – relatively small potatoes within this biz).

    Here’s another example, made a little worse by the fact that the bad wording is repeated by Robert Scoble, who should know better.

    The Yankee Group released a report stating that Windows Server 2003 had “nearly 20% more annual uptime” than Red Had Enterprise Linux. This was then reported by Yahoo, and in turn picked up by Scoble .

    Scoble, at least, should know better. Claiming that Windows Server had 20% more uptime than Linux implies the average Linux Server was down about 20% of the time, or ~70 days a year or ~5 hours a day. That’s ludicrous and anybody who knows anything about tech should know that.

    The brief report synopsis on the Yankee site does not clarify things any further, but the Yahoo article contains more details, presumably from the full report. Further down in the article, it states that the various servers (Unix servers were also tested), “experienced 3 to 5 failures per server per year in 2005, generating 10 to 19.5 hours of annual downtime for each server”.

    So I think it’s safe to assume that all the articles should have stated that Linux had 20% more downtime (annoying, but not alarming), rather than 20% less uptime.

    i.e. With (24 * 365 =) 8760 hours in a year, the worst case is that Linux servers averaged about ((8760-19.5)/8760) = 99.78% uptime, versus approximately 99.82% uptime for Windows 2003 servers. Rather less alarming, and rather less worthy of Scoble’s trumpeting post.