Sulka's Game

Thoughts on in-game advertising

2008-09-07T19:00:07Z

One of the hot new ways to make money with games and virtual worlds is advertising. Or at least that's what a lot of companies want all game developers to believe. If visit website of any of the in-game ad network providers such as IGA Worldwide or Double Fusion, they have a bunch of press releases out on more and more games including ads, and what a wonderful world has become to the developers.

Nielsen and IGA also published a study in June where they found 82% of gamers didn't mind contextual ads in games. Curiously Nielsen.com doesn't have any more information about the study (contrary to the press release) so I don't know what games and methodologies were used to get the result. I would assume if this was on a racing game, the game would probably feel more real with real ads, so getting such a good result would be quite natural. Had this been about slapping a Coke board on a building in Thunder Bluff, I doubt the score would have been in this league.

Now, it seems the current message being broadcasted to game developers is that putting ads into your game gives you free money. Unfortunately, it's not that simple. Andrew Chen blogged an excellent piece about maximizing ad revenue a couple weeks back, which specifically deals with what websites can do to get more dough. If you're doing bulk advertising (banners in any for, be it adwords or video streamed to a medium rectangle), bringing in serious money requires serious volumes, which, if you read Andrew's posting, changes how you make the money on a given network.

Websites can tackle this by integrating multiple networks of choice, but I as far as I know (and I don't know much), there aren't enough of them for in-game advertising to allow for this to happen for games. With relatively little competition, the networks can require exclusive deals, and I'd be surprised if the CPM rates of the networks varied enough to call for this type of an optimization. Further, I'm not sure if too many game developers even think of this level of sophistication when integrating ads into a game, given that most websites don't probably do this. Anyway, this probably means that the average developer who integrates into the current networks should not expect a fat check in the mail, no matter what anyone tells you.

Talking about economies of scale, part of the reason Google is so darn effective in the online advertising is that they can sell targeted advertising to anyone willing to spend any money at all in ads. Joe Bob's Auto Garage from Oklahoma can decide to spend a hundred bucks on people searching for local auto dealers and it's possible. With in-game ads, I doubt getting access to those dollars will come any time soon, which probably cuts off a very significant portion of businesses putting money into advertising. And if your ad dollars come from a small set of big companies, you won't get a single ad in your game if they don't happen to like the content. Modern online advertising has a long tail, and the way to cash that is by serving Google Adwords.

Coming to think of it, it might be worth starting a business that offered developers technology that offered compatibility with any in-game ad network and enabled the developers to change the integration on the fly, rather than relying on a single network. This could be the beginning of making smaller scale ad-funded games a lot more viable than today.

Now, even if advertising in games is it's infancy, there's movement ahead that might make in-game ads a lot more appealing and lucrative in the future. PlayNoEvil had an extremely interesting speculation about upcoming ad war between Google and Microsoft. Given that games can offer a lot deeper exposure than most websites, if there'll be disturbances in the force on the web, advertisers might suddenly start to flock to media where ad-blockers and browser incompatibilities are nowhere to be seen. We live in interesting times!

What's happening with game pricing?

2008-09-04T19:54:57Z

I have hard time believing this. Warhammer Online is selling for the same price in Europe as in the States, once you do currency conversion. Amazon UK price is £26.98, and US charges $46.99. Rounded to closest whole number, those both amount to 33 . Europeans not paying extra for a game is amazing.

So here's the really, really amazing part. Wrath of the Lich King is cheaper in Amazon UK than in US! UK price for the item is 17.98 British pounds = 22.15 Euros, while US price is $39.99 = 27.87 Euros. I can't remember ever seeing this before, especially for software made in US.

More on Chrome performance and success metrics in general

2008-09-04T19:03:59Z

Given that I found Chrome, Webkit Squirrelfish and Firefox 3.1 performance to be on par, I started to wonder why so many sources are claiming Chrome is the fastest engine out there.

The reason seems to be this benchmark. Instead of using Sunspider or some other existing benchmark, the V8 guys have created their own benchmark and tuned V8 to perform well with that particular benchmark in mind. And of course, all the excited bloggers are using these numbers.

This has paid off on that benchmark, of course. Webkit scores 391 points, Firefox 3.1 gets 162 points and V8 goes all the way to 1927.

So what's wrong with this picture? The problem is, looking at the description of the benchmark, nothing in the V8 benchmark is meaningful for a web browser. I'm sure getting big numbers is a cool thing to present at work and gets you the Nerd Bonus, but that just means the team has forgotten the goal they should have had when implementing the engine - high quality web surfing experience. This is in stark contrast with the goals of Sunspider, which aims to benchmark real world use cases, so good score in Sunspider should translate to more comfort when actually using the browser.

In my opinion, the V8 team has defined their success metrics based on something they shouldn't have. They should have aimed for best possible browsing experience and hence used benchmarks like Sunspider and Dromaeo to tune the engine.

Setting wrong goals seems to be quite general issue with any industry, given that setting the right metrics to measure success requires a deep understanding of the problem at hand and objective approach to solving the issues. At worst, you define the goals to please someone in your organization through numbers. Someone I know was working as a networking system developer and switched jobs after being forced to optimize for throughput benchmarks (to please the boss of his boss), rather than optimizing for real world use cases.

Coming up soon, what questions I iterate to understand the problem.

Girl saves family from crashed car

2008-09-04T14:31:31Z

MyWebTimes.com reports a 11-year old girl saved her family from a crashed car, motivated by knowing an upturned car can explode because it happens in Grand Theft Auto.

I'll just say the parents should not have let the kid play GTA in the first place, but good think the family survived the crash. The Kotaku readers have much more interesting comments, such as this total classic from Nobuyuki: "The article fails to mention that after she saved her family, she jacked a car passing buy and ran over 3 cops."

So how fast is this Chrome thing

2008-09-03T20:22:15Z

The blogosphere seems to be chock full of Chrome news and benchmarks. My inner nerd likes benchmarks so I took note and tried to find a good comparison between the browsers. To much disappointment, I couldn't find a single one that actually included all the upcoming interesting browsers, namely Chrome's V8 and IE 8, Webkit's new Squirrelfish and Mozilla's Tracemonkey.

So I made my own chart. Overall speeds of the browsers are (time in ms, smaller is better):

What can we see here? IE 8 is going to be a massive speed boost to the benchmark, but only because they've fixed whatever caused the abysmally bad String score (as seen below), meaning I doubt real world applications will actually see any improvement. Opera is fine but not the cream of the crop. Current versions of Firefox, Opera and Safari are good already, but the new versions will improve the performance very significantly.

Interestingly, Chrome doesn't actually perform that well on the benchmark. I think my machine is having some odd Beta issue that's causing it to perform badly, given my Webkit score is better than on benchmarks that giv much better score to Chrome. However, even if I took the best score I've seen for Chrome online, it's not better than Webkit or Firefox's upcoming releases.

Detailed graph below. Please note the following uses logarithmic scale. Again, smaller is better.

What does this mean? What's at least clear is, the claims of having implemented ground-breaking new performance-shattering engines that Firefox, Chrome and Webkit teams have put out are true, but also marketing drivel given the engines are actually running at similar speeds. Microsoft is losing the browser game which hopefully will make them releases a high quality browser one day, but I'm not holding my breath that'll happen very soon.

And yes, Chrome user interface is nice. When the Mac version comes out, I'll consider switching right away. Using Webkit nightly as the main browser is like so last night anyway.

Emergent gameplay in WoW

2008-09-02T08:42:19Z

I've previously mentioned how WoW players are using the heavy leather ball to added fun - for example see this discussion on how someone comments spending two hours doing PvP in Warsong Gulch was fun because the group had the ball and couple play a game with the ball while slaughtering the poor Alliance members.

I was prompted to to write this after I saw some incredibly funny videos on YouTube about players doing very, very long jumps using the Rocket Boots Xtreme. For the uninitiated, this is a pair of boots that can be manufactured by players specialized in engineering, and for a for a very high price. I doubt the Bliz dudes really thought what they were doing when they implemented this feature - I'm immediately thinking how to abuse this for capture the flag PvP. Anyway, see the video below for a method of travel that probably wasn't ever supposed to make it into the game.

Who does digital distribution empower?

2008-09-01T11:51:50Z

Bruce has noticed the rise of digital distribution. This reminded me of commentary by Greg Costikyan - I recall he ranted in GDC a couple years back about how the next gen consoles will probably have no physical drives at all, as opposed to relying on Xbox Live and PS Network for all game sales. This in turn will mean that instead of having competing publishers who get to participate in decision making of which games get shelf space in stores, those marketing decisions will be 100% determined by Microsoft, Sony and Nintendo. If your game doesn't sell, it'll fly out from the front page of the store in a day.

Unfortunately I can't find which talk this was in. I thought it was the infamous Game Developer Rant from GDC 2005 so I dug up the MP3 and listened if Crys had missed a piece but no, it's not there. If anyone can point me to this, I'd love it.

The talk did have one classic line though from Greg, which wasn't blogged (on Nintendo's practice of relying on one aging designer over cheaply licensed devkits): "Iwata-san has the heart of a gamer, and my question is what poor bastard's chest did he carve it from?"

Give me some lovin'

2008-08-29T12:59:34Z

Scott J points to some pretty harsh commentary directed to the Age of Conan team.

I have to say I'm familiar with the phenomena. When you have millions of people using your site, each of whom has a lot of time, you get a collective results of million hours worth of feedback every month. Now that's fine and useful most of the time, except when it turns to silly things like defacing wikipedia pages about the product (there's a bit over 5000 edits on english Wikipedia Habbo page alone), or scamming other users to "get even". So, I feel the pain the AoC dev team must be experiencing.

It's been interesting to read the commentary of AoC all over the web, given I've not played the game myself. The bit that I disagree with, is the notion that the company intentionally put out a bad product out because they thought they can just rely on the franchise. I don't believe that for a second. What they were probably aiming for is a killer product, and they just didn't have the skill to deliver. The biggest fail in in the project seems to have been overspending on the look and feel, compared to the resourcing in the programming and design of the actual game. That in turn indicates whomever made the calls on the resourcing failed, so most of the team really is not to blame.

(Incidentally, I apologize for yet another breakage for some readers. I had accidentally not converted the URL most people see to use the Feedburner version, so most readers were still getting the old crappy feed. I promise to stop fiddling this for now.)

News from the DSLR-land

2008-08-28T07:57:23Z

I'm a digital camera junkie. I've had 6 (3 point and shoots and 3 DSLRs), and am always looking to upgrade. However, the reasons for upgrading aren't primarily more features, but convenience. The biggest shift that's happened with digital cameras over the past six years that I've been on board, has been the current products have more of the good stuff from the film cameras than the early digicams, while also embracing the digital capabilities better. In concrete terms, my current DSLR body (EOS 40D) operates extremely fast (like film did) but also gives me very high sensitivity at good image quality (which film never delivered, at least at decent cost per frame). Compared to film cameras, cheaper DSLRs still mostly have inferior viewfinders and the autofocus systems are pretty sad compared to what affordable film cameras used to have.

Every digital camera I've bought has been a Canon and I have a considerable arsenal of lenses for the Canon EOS system, which would make switching expensive. This obviously results in myself being more interested in what Canon is releasing, than what competing companies do. There's new product out now from Nikon, however, that makes me wish I'd gone the Nikon route. The Nikon D90 shoots HD video at 720P resolution. This is massive news. For the uninitiated, digital video solutions with similar imaging sensors and ability to swap lenses have have previously a hell of a lot more, and now you can get that "for free" as part of an awesome DSLR.

Chase Jarvis seemed to like the product a lot:

Not all is lost, however. Canon Rumors, and some other sites, are reporting the upcoming EOS 5D mark II also shoots HD, and maybe even at the 1080P resolution. The site has been slapped with cease and desists, which adds to the credibility of the rumor. If this is true, I'm going to fork out for the camera immediately, even if I have to sell something to get it. Video is cool, and having proper control over the depth of field when shooting would be awesome.

Subscriptions encourage good game design??!?

2008-08-27T19:25:44Z

Ben Zeigler wrote a continuation piece to his MMO economy musings where he mentions me (thank you!) and an interesting piece from Scott Jennings.

While I do agree to the if you give players too many choices, they end up choosing none" sentiment in Ben's post, the piece has some interesting assumptions:

"However, when playing a microtransaction game, instead of one simple decision, you are confronted with a shitload of complicated decisions. Is this sword really worth $2? What about this other sword that costs $5? Oh crap what about that haircut?. When presented with microtransactions, my eyes usually glaze over, and I end up buying nothing."

This makes me wonder if Ben has actually played micro-transactional games. In my experience, what happens is that users first decide on what they want, and then get the currency to get the object of desire. Nobody ever experiences the situation where you'd first shell out to get some game points and figure out what to get with them afterwards - simply because you need to build the desire to buy before you get the cash.

"In a pay-for-play system its clearly in the financial interest of the developer to make a game that gets played as much as possible."

While this is true for games that actually charge for time spent, I don't think there are too many games out there that use the model. "Micro-transaction" does not equal pay-for-play, and what this means is that while users who buy into your game will buy more, one has to remember the free to play games are indeed free to play. This means you have a lot of people who don't pay, and some of these users might play for a lot too - especially if you incentivize long sessions. What you really want to do when designing a game that has paid currency, is make the paid items as desirable as possible, without adversely affecting the game balance.

In an ideal world this will result in users paying more if they play more. This is in stark contrast with the subscription market. According to Nick Yee's WoW demographics, the average users spends 22.7 hours / week in WoW. This is about 90 hours a month. Now, the sub price is about 12 euros a month, which means the average WoW player is paying around 13 cents / hours for the entertainment. That's not very much, especially if we compare to traditional industries. Say, a CD that costs 10 euros and contains a hour of music has to be listened for 76 times before the price of each second of listening goes below what the current best (yeah yeah) offering of the game industry can charge. I have very few songs in my iTunes library with a play count higher than that, and 0 full records where all songs would approach anything near this.

So - the question is - how do we change games and related monetization models so that we can start charging anything close to what the older entertainment industries are charging for entertaining people? I think subscriptions won't get us there, while I do believe the entertainment we're offering is better.

Feed fixed

2008-08-27T17:00:20Z

Adam told me my feed sucked, which I hadn't noticed. The feed should now include the full stories with full formatting. If you have issues with the feed, please ping me. If you get the old stories in duplicate - sorry.

Making the most revenue out of an MMO

2008-08-25T09:43:42Z

I read two blog posts this morning which talk about MMO revenue models but go on wildly different tangents. These were:

Why Subscription-Based MMOs Make Sense over at Double Buffered, where Ben Zeigler points out that there is no publicly known cases of wildly profitable MMOs that use the free to play, pay for stuff model, versus many games that make a lot of profit, which use subscriptions.

Bruce Everiss, on the other hand, argues that the western games industry is losing $1 billion in revenue due to only offering subscriptions, which I wholeheartedly agree with. As a WoW player with limited amount of time on my hands, I'm considering ending my subscription since I can't compete in the game with teens who have endless supply of time. I'd have no trouble dropping down some cash to get gold if Blizzard offered the option, which would allow me to get on fair level with the players who have more time.

If I was to design a new product, I'd aim to implement both models. One way to do this could be:

Allow people to play for free, but implement some artificial strict restriction like not earning XP past some early level, unless you subscribe. This is essentially the current try-before-you-buy system, except I wouldn't enforce any 10 day limit and I'd let people keep they accounts in some dormant mode even after ending the subscription. I'm fully aware this can be argued to lose some revenue, but I believe enforcing subscription to keep data is akin to blackmail.
Implement a dual-currency model, where one currency is earned, and one bought. It should be possible to play without ever using the bought currency, but for most fun one would spend both. Users would be able to trade the currency with no pre-defined rate. The bought currency would be used for primarily vanity items, which incidentally is exactly what Blizzard is doing with their collectible WoW cards.
Allow users to pay the subscription either by paying, say, $10 a month for just the sub, or $20 a month for both the sub and $20 worth in the currency.
Allow a user to also pay for the monthly sub by spending $10 worth of the paid currency.

Doing something like this might enable getting the subscription revenues, as well as be close enough to "free play" that you'd get the large player base. The reason for allowing users to pay for subs using the paid currency is simple - I don't care who's paying the sub for someone else and what for, as long as the money keeps flowing in. The $20 option means you'd be giving out three subs for the price of two, but I bet most of the money would actually go towards spending it in something else. And of course, it would allow people who grind to grind, and the people who want to purchase the results of the grinding pay for the grinding, in a manner where the money goes through the publisher.

Dunno if that'd work. Feel free to criticize! :)

Gamedevelopertools.com

2008-08-22T20:07:08Z

I happened to bump to Game Developer Tools site which apparently is pretty new. GDT claims to be the "comprehensive library of game development resources" so I'm blogging this partly as a personal bookmark.

I didn't check the other parts of the site comprehensively, but the book section looks useful since each book appears to be listed only once. This is much better than doing searches on Amazon where you get every darn print of every book appearing in the results. Some of the contents are a bit dubious though, making me think maybe it's all been scraped from Amazon, after all... :)

Photosynth

2008-08-22T12:05:01Z

I just saw the friendliest "we don't support you computer" message ever - and it came from Microsoft:

"Unfortunately, we're not cool enough to run on your OS yet. We really wish we had a version of Photosynth that worked cross platform, but for now it only runs on Windows. Trust us, as soon as we have a Mac version ready, it will be up and available on our site."

I'm impressed even before I've seen the service itself.

Blizzard counters with Wrath of Lich King cinematic trailer

2008-08-22T08:54:14Z

Blizzard has released a cinematic trailer of Wrath of Lich King. I suggest you get the downloadable HD version for full screen goodness.

The video looks very pretty but is very thin on actual content. Compared to the Warhammer trailer I blogged about yesterday, the WAR trailer has tons of fun stuff on the expected gameplay of the game, in addition to looking great. I think Paul Barnett knew what he was talking about when he said they've fully embraced the rich background of Warhammer, while the WoW folks always need to come up with new IP.

New Warhammer cinematic trailer

2008-08-21T18:55:49Z

Just as I scorned Warhammer Online, Massively pointed me to a new Warhammer cinematic trailer. I suggest you check the HD version. My favorite part is the Goblin Squig Herder summoning the Squig for riding into the city.

Just Leap In

2008-08-21T08:17:58Z

Raph just blogged about Just Leap In - a new embeddable 3D world client.

In addition to agreeing on Raph's comments about the high barrier of entry to authoring content on a world with 3D assets looking like this (see the video below), the demo video made me wonder what purpose has the product been designed for? The demo video has zero social interaction. This, combined with the technical mumbo jumbo for the content authoring, makes me think the project is run from purely technical perspective and nobody has thought the users as part of the design process. Sad, really. I thought the collective wisdom for building virtual worlds would soon have reached the point where neutron bomb worlds are the way of bye bye.

Lesson in software security for game programmers

2008-08-20T20:09:38Z

PlayNoEvil pointed to an interesting old story about analysing a seeming innocent-looking shuffle algorithm for Texas Holdem'. Makes for a very good read. And proves that the behavior of most code is not fundamentally understood even by the person who wrote the code.

This reminded me of an old bug I reported on PHP, which was causing similar huge security issues on a number of website, the owners of whom ended up never knowing they had had a problem. The issues was that a number of open source PHP-based applications used the following algorithm (which was also suggested in PHP website) to generate random passwords who registered or wanted their passwords reset:

password = "";
$array = array('a','b','c','d','e','f','g','h','i','j','k','l','m',' n','o','p','q','r','s','t','u','v','w','x','y','z',0,1,2,4, 5,6,7,8,9);
shuffle(&$array);
for ($i=0; $i<6; $i++) { $password .= $array[$i]; }

Looks innocent, right? An array of wanted characters is shuffled to random order and the first six are concatenated to form the new password. The issue here is, the programmer assumes the shuffle is actually random. And on some systems, it was, some it wasn't.

The c-language implementation of PHP's shuffle sorting at the time was the following:

(php_rand() % 2) ? 1 : -1

where php_rand was actually using the platform's native rand() call. The cunning reader notices the algorithm really only uses the least significant bit of the generated random number, and hence relies on the stream of least significant bits being random. This is the point where the chain of assumptions failed.

On Solaris and Linux of the time (and maybe still!), the least significant bit of the generated random numbers wasn't random at all. If you generate a sequence of random numbers on these systems with rand(), the LSB always follows one of the following four patterns, depending on which point in the sequence in the grand scheme of things you start iterating: 1001, 0011, 0110 or 1100.

Now where does this leave our password algorithm? The answer is, on the affected systems the above algorithm only ever generated four different passwords, because there only were four different shuffles. These four looked random, but weren't. And any user could generate the four combinations by just requesting a password reset a few times, after which you could request the password of someone else be reset.

To resolve the issue, PHP was changed to use an algorithm that generates a more random sequence of LSB's. One could also point that sending passwords to users when they register or want one reset is exceedingly stupid, but I'll leave that as an exercise to the reader. I'd be interested to know how many pieces of software still use the LSB from rand(), though.

Update: reading what I wrote made me think the story is a bit strange, coming from someone working as a concept designer. But hey, knowing a broad range of stuff can't hurt.

Emily

2008-08-20T19:19:32Z

Times Online has a video clip of a wholly computer generated talking head which some people have gotten excited about. The page itself claims the animation leaps the uncanny valley but I totally disagree. It's impressive, right, and looks like a step into the direction, but uncanny it is. The individual still are very good but the movement just isn't right. Maybe they hired a robot to play the woman for the capture?

WAR!

2008-08-20T10:34:53Z

The interweb is full of WAR related stuff now that the NDA was lifted. I have to say I _love_ some aspects of the stuff that are coming up. Like the copywriting! The advancement paths for Orc tanks are:

Path of Da' Brawler - You'z gud at stabbin' things.
Path of Da' Toughest - You'z gud at survivability. Mor' killin' later!
Path of Da' Boss - You'z gud at da group support, cause everyone 'as some killin' ta do.

The bad parts of the deal is of course that the game is again split to separate US and Europe servers so I can't actually play in a group with friends from both continents. And for some reason the WAR Europe website hasn't been tested on any browser except for Internet Explorer (just failed to register due to crap javascript forms that FAIL on Firefox). Looks like it's going to be an awesome game, but again a project that's failed to realize the web exists. Shame.