Google weighs in on how gamers use search

I didn't know it when I posted Measuring MMO popularity with Google Trends, but Google had earlier issued a white paper and blog post on how gamers use search. Both the blog post and white paper are in support of Google's ad business and aim to convince game publishers to advertise with Google.

The blog entry, How the digital era has changed video game launches, summarizes the white paper.

The white paper itself, Understanding the Modern Gamer, has the details and is based on an analysis of "…video game searches…for the top 20 selling games of 2010 and 2011."

Quoting various bits:

Game searches included title keywords, such as ‘Elder Scrolls V’, along with all relevant game title keyword variations, such as ‘Gears of War 3 trailer’, ‘Battlefield 3 website’, ‘Batman Arkham City review’, and more.

Search data analysis is representative of gamer behavior for 2 reasons: (1) millions of gamers use search engines, providing us with an unparalleled data set and, (2) gamers are incredibly savvy Internet users whose searches reveal an extraordinarily high level of intention.

Desktop engagement patterns closely align with game unit sales trends (.92 correlation)…

In 2011, desktop searches per gamer grew 20% y/y. Additionally, desktop search volume for marquee titles increased 29% y/y, outpacing the 8% growth y/y in console gamers online.

Our data reveals a .92 correlation (on a scale of -1 to 1,1 signifies perfect correlation) between clicks during the 10 month game launch cycle and game units sold during the first 4 months post-release…

More importantly, our data demonstrates that 84% of sales can be predicted by all clicks during the 10 month launch cycle. We used the regression coefficient from our analysis to create a predictive model and found that if a game accrues 250,000 clicks in the 10 months around launch, it will likely sell between 2 and 4 million units in the first four months after release.

Although user click data is a powerful predictor of game unit sales, we readily acknowledge that other factors – such as game quality, TV investment, online display investment, social buzz, and more – must be incorporated into our analysis to create a predictive model that is even more accurate and reliable.

There is a quantifiable link between what people search for and what they buy that enables us to predict game sales.

Note that the study may involve console games only.

At the time when a person conducts a search, that game has the person's attention. This is more true than for Xfire or Raptr which merely measure when a program is running, not what the owner of the computer is doing.

Xfire and Raptr are not measuring attention. As Wilhelm Arcturus reports, they will report a game in use when a computer is left unattended with a game launcher open. Even if the game itself is running, the player may be on a bio break, macroing, or using some other AFK-based exploit where the game is running without the player present.

Google Trends statistics reflect people physically present at their computers actively typing the name of the game into the search box. That's a significant level of engagement with a game, even if they aren't actually playing at the time of the search.

Google Trends does have some issues that may result in it undercounting mature games. As a game matures, everyone becomes familiar with it. Once it's old news the number of people searching for it by name to get general information will fall.

Well after release games may retain high numbers of engaged players conducting searches, but these searches focus on details of the game. Such searches include highly targeted game-specific terms but often not the name of the game as discoverable by us using Google Trends, although Google itself could tie such searches back to games using its data. (See also Tobold's commenter Oscar's comment about "functional" searches.)

Discovery of game-world information for mature games may also move off Google to game-specific database sites, build configurators, or forums, depending on the extent of the game community's ecosystem.

Despite these shortcomings, I think Google Trends is much better suited than Xfire or Raptr to measuring the popularity of a game, due to Google's huge demographic advantage.

SOE president John Smedley validates the use of Google Trends to measure game popularity

In the throes of the recent launch of PlanetSide 2, SOE president John Smedley has been quite active on Twitter. Tonight (or actually last night) he tweeted this:

Happy to see this google.com/trends/explore...

His tweet links to a Google Trends plot much like this one:

The difference being that his plot shows the most recent 90 days, while the above shows an absolute time range of September through November. Both plots are similar now but the above plot won't change after the end of the month. [Update: He tweeted on November 22nd when the most recent data for Google Trends was from the 21st. November 21st is the point where the blue line for Planetside 2 briefly crosses the red line for League of Legends.]

When I first blogged about Measuring MMO popularity with Google Trends a bunch of MMO bloggers were pretty critical. Smedley could have cited Xfire or Raptr, but his tweet is a nice validation of Google Trends from the president of a prominent MMO company.

Measuring MMO popularity with Google Trends

Tobold had a bet going that with the release of the Mists of Pandaria expansion, World of Warcraft would retake the top spot on The Nosy Gamer's rankings of the top twelve MMORPGs. Guild Wars 2 had held the top spot since its release last month.

Nosy Gamer's rankings are based on data from Xfire and and thus reflect its fairly hard core demographic, so with this week's rankings out showing WoW indeed beating GW2, Tobold gets to crow about being right.

But since games are now played by people far outside the narrow demographic of hard core gamers measured by Xfire, I'd like to call attention to another source of data which can be used to measure a game's popularity—Google Trends.

Below I'll show that the Google Trends dataset reveals WoW's long-term decline despite the four expansions released over the years. It also shows that while a few western MMOs have spiked above WoW's popularity for a short time, only one asian MMO has been able to sustain it. There are a few western games that are more popular than WoW but they're not MMOs.

Wannabe crafting monopolists hating on Guild Wars 2

I've been playing and enjoying Guild Wars 2. One of the differences between GW2 and other games is the universal Trading Post, which is an auction house that spans all the servers.

Because spanning all the servers allows the Trading Post to provide a very efficient market, it's been subject to some heavy criticism from MMO bloggers.

While not criticizing the Trading Post, Spinks notes that:

...raw materials (particularly metal, cloth, and leather) typically sell for more than the finished pieces.

How is this new? This was always the case with WoW items crafted from trainer recipes. The issue is the crafting system, not the TP. (I don't mean to imply that Spinks is a wannabe monopolist; I just wanted to point out that selling crafted items below cost is not new with GW2).

Tobold weighs in with a piece titled Crafting regrets in which he says:

I think Guild Wars 2 is extremely bad for making money from crafting, because the auction house spans all servers. There is no opportunity for arbitrage in such a large economy. [my emphasis]

And later:

There simply doesn't appear to be a point to crafting in Guild Wars 2. Are the any "bind on pickup" recipes like in World of Warcraft? If not, I'll just buy everything for cheap on the AH. [my emphasis]

Arbitrage—that's something Wall Streeters do. I've done it to earn gold in MMOs, but I've always felt a bit dirty doing it. I think it's excellent that the efficiency of the Trading Post reduces the opportunity for arbitrage.

I read Tobold's post a few days ago but had forgotten it by the time I read and commented on Elder Game's The Case Against Auction Houses. I'm reproducing my comment here in full since it serves as an excellent rebuttal to Tobold's desire for monopolies via bind-on-pickup recipes. On Elder Game I wrote:

I love crafting, particularly to make items for my main and alts.

Sure, it's nice to be able to sell at a profit, but not if it involves monopolies. What the crafters who whine about profit are really saying is "I want a monopoly." Personally, I wouldn't cater to the wannabe 1-percenters, but hey, it's your game.

In-game shops use geography to achieve local monopolies (perhaps shared with a few others). Not all locations are as good as others, so how would the game decide which players get to have shops in the prime locations of the crowded main city?

The problem isn't with the AH, the problem is with crafting systems. WoW limited the supply of crafted items by making some ingredients rare, but mostly by making recipes random rare drops; anyone lucky enough to get a Pattern: Rich Purple Silk Shirt can make a bundle, either by selling it outright or as a tailor making the shirts.

Worst were the bind-on-pickup recipes that dropped in endgame instances so that non-raiders had no hope of ever acquiring them. As someone more interested in crafting than raiding it was one of the things about WoW I hated most.

The issue is how to limit the supply of crafted items in a fair and equitable manner without resorting to using the random number generator to award monopolies to 1-percenters.

As I see it the problem is with designs that make it possible to create items in the blink of an eye. Why do so many games allow one to "lovingly handcraft" a valuable item in the time it takes to click a mouse? Craftsmanship implies labor over a period of time.

One game I've seen which implemented a crafting bottleneck well was Fallen Earth, where crafting takes real time whether you're playing or logged off. Making crafting use real time limits the supply without resorting to randomness or monopolies. Yes, it's still open to exploitation, or at least mass production, through the use of alts and additional accounts, but the idea of making item creation take real time seems like a good mechanic.

It leaves unsolved the issue of how to provide a sink for all the commodities left unused by a time-throttled crafting mechanic.

There are other ways to improve the profitability of crafting instead of implementing recipe monopolies or making the Trading Post less efficient:

First, others in the thread brought up the old idea of real wear and tear so that items don't last forever but wear out and require replacement.

Second, eliminate gear as quest rewards, and do away with 99 percent of all item drops from mobs so that they only drop salvage items or vendor junk. A useable item of any quality dropping from a mob should be a rare thing. Make the majority of gear in the game player-crafted, except for a handful of items.

The ancient trope of phat lewtz drops carried down from the pencil and paper days of Dungeons and Dragons hurts crafting more than any Auction House. Early D&D's focus on combat at the expense of crafting has persisted in RPGs for years and continues in MMOs.

For the first couple of years of its existence, Minecraft had it right in that mobs only dropped raw materials. I was very disappointed when an update cause mobs to sometimes drop weapons and armor, possibly enchanted.

As I see it, Guild Wars 2's changes to crafting and commerce aren't the problem—the real problem is that the changes didn't go far enough.

Soren Johnson on Minecraft in the cloud

One can’t help wonder what Mojang could do with a cloud-based version of Minecraft, seamlessly updated, playable from any device or browser, that connects every world end-to-end.

-Soren Johnson in GD Column 20: The Coming Storm

Originally published in the Feb 2012 issue of Game Developer magazine, he posted it to his blog today.

Beyond the mention of Minecraft it's an interesting discussion of cloud gaming that I highly recommend.

6502 emulator in Minecraft runs Forth

Notch's new game 0x10c and its virtual 16-bit computer, the DCPU-16, have made news recently. Somewhat eclipsed by Notch's project is the 6502 emulator now available for Minecraft as part of Prerelease 5 of the mod RedPower 2 by Eloraam.

Eloraam calls it the 65EL02, because "it supports all the 6502, 65C02, and part of the 65C816 instruction set" as well as "a set of completely new instructions and two addressing modes."

The emulated CPU comes with 8K RAM, with up to seven additional 8K RAM Modules installable on Backplanes that must be placed on blocks behind the CPU.

Additional components are Monitors, floppy Disk Drives, and IO Expanders, all of which may be located at a distance from the CPU as long as they're connected by Ribbon Cable blocks.

A minimal installation consists of a CPU, Disk Drive, and Monitor. Up to 256 devices (just the Monitors, Disk Drives, and IO Expanders thus far) are supported by Redbus over the ribbon cables.

RedPower 2's 65EL02 computer system in Minecraft. Shown are a CPU with three 8K RAM Modules, a Monitor on the left, and a floppy Disk Drive adjacent on the right. Further right a Ribbon Cable connects to an IO Expander interfaced to a Bundled Cable that carries 16 redstone signals. (Click to enlarge.)

The portion of Eloraam's addon with the computer is known as RedPower Control due to its intended use for industrial process control. Among other things the computer can control in Minecraft are automated mining machines built with RedPower's newly-introduced Frames, which are movable scaffold blocks used to build moving gantries, booms, and cranes.

Eloraam has put an incredible amount of work into RedPower Control and the other parts of Prerelease 5, and it shows. Her textures, while remaining at Minecraft's low resolution, are very attractive and make RedPower's items pleasant to look at in-game.

A example is the front panel displayed when you right-click the 65EL02 CPU block:

The front panel of the 65EL02 CPU. Yes, the toggle switches really work! (Click to enlarge.)

The front panel recalls the look of Digital Equipment Corporation (DEC) hardware really well. The rocker switches, in particular, are great renditions of those on the front panel of PDP-11/35s and /40s. When I complimented her on the appearance of the front panel, Eloraam described the effort it took:

The rocker switches took hours to draw. They were so hard that I ended up resorting to tracing the rough contours and illumination from the one good front-on picture of a PDP-11/35 I could find on Google. Even then, it wasn't a copy-paste job. The exemplar gave me a good map of where the speculars and shadows fall, but it was also a grainy JPEG with a myriad of photographic imperfections, as well as having the wrong aspect ratio. So once I put down the basic outline and marked out how the speculars fall, I tossed the exemplar and drew everything myself.

I must admit I found working on such a large canvas daunting. I'm an engineer, this art stuff is all new to me. I've developed a knack for 16x16 through lots of practice (RedPower has something like 500 sprites now, and I drew them all), but this was something else entirely.

Since the 65EL02 is an 8-bit CPU, Eloraam didn't have as many options for programming environments as we have on today's 64-bit computers. While it's possible to program the 65EL02 in assembly language, for general use she chose to implement a Forth interpreter. Less well-known than BASIC, Forth is far more powerful while retaining a tiny footprint and being easy to understand.

The output of the Forth WORDS command. (Click to enlarge.)

Note the sophistication of the scheduler built into the 6502 emulation as described in this reply by Eloraam to a question about infinite loops:

Infinite loops are fine. RP Control is carefully designed so that you won't screw up your world even if you crash the virtual computer. It's actually not especially hard to crash the virtual computer, since the whole OS is loaded into its RAM and you can easily write to that RAM. Still, since the computer is fully virtualized, it won't hurt your world or even cause a slowdown.

The one thing to note is that you should always use at least one TICK or some TICKS in a loop if you're just waiting for things in the world to change. The computer will run a lot faster if you're not constantly wasting CPU cycles. The CPU only gives you a small number of cycles each tick, but it saves them up (to a certain limit), so if you don't waste them, your CPU can run a lot of code at once when you need it. TICK (and TICKS) calls a special instruction in the CPU which saves up any unused cycles, so it's your friend.

Further technical information about RedPower Control's 65EL02 is available on Eloraam's blog RP Control Internals, and on the RedPower wiki's page for RedPower Control.

For those that prefer video, YouTuber direwolf20 has a video introduction and basic tutorial for RedPower 2's new computer system:

It's interesting to note that Notch considered the 6502 before switching to a custom 16-bit CPU, even going as far as writing a working 6502 emulator. Just two days after getting his emulator running, Notch tweeted:

Starting to dislike the 6502 idea. Spending heaps of cpu cycles to emulate not being able to multiply. Reconsidering custom CPU again.

It appears that Notch intends his virtual CPU to be a major part of his new game and thus decided that the 6502 would be too limited due to being eight bits. Judging by the enthusiastic reaction to the DCPU-16 it appears he made the right choice.

Eloraam took another path by augmenting the 6502 with 65C02 and some 65C816 instructions and implementing MUL and DIV. The goal for the 65EL02 is much lower than for the DCPU-16. While capable of cool things, the 65EL02 seems intended to be a workhorse for carrying out the business of industrial process control in Minecraft.

Much of the excitement about the DCPU-16 has been anticipatory, although this will change soon when Notch releases the DCPU-16 emulator and then the first playable release of 0x10c. On the other hand RedPower's 65EL02 is available now and fully usable in its intended environment.

Let the 65EL02 hacking begin!

Update: This article was linked on Hack a Day, Slashdot, and engadget.

Escalator in Minecraft using RedPower 2 Frames

Eloraam recently unveiled prerelease 5 of her RedPower 2 mod for Minecraft. One of the long-anticipated features in PR5 is Frames, which are movable scaffolding blocks that have had fans drooling since last fall.

On my list of possible things to build with Frames was an escalator. I'm pleased to show that RedPower 2 Frames indeed make it possible to construct an escalator in Minecraft.

Escalator using Frames from RedPower 2 (click to enlarge)

A still picture really doesn't do it justice, one needs to see it in action:

One fly in the ointment: Frames don't move the player horizontally, so you have to hold the Forward key during the horizontal movements.

This escalator somewhat misuses Frames and Frame Motors. In normal projects where Frames are used to construct mobile gantries or booms, the ratio of Frames to Frame Motors might be on the order of 30 to 1. Here it's roughly two Frame Motors for every Frame.

The design turns out to be quite simple. Because the Frame Motors for each direction of escalator travel are in the same plane and are adjacent both horizontally and vertically, no additional wiring is needed for power beyond the the single Blue Alloy Wire providing it at the base. Like all blulectric devices, Frame Motors conduct blutricity to their neighbors.

A Red Alloy Wire on the back of each set of Frame Motors supplies the timing signal to synchronize all the motors. The signal is from a Timer set to an interval of one second. I'm not sure how inefficient it is in terms of blulectric power consumption to activate all the Frame Motors for each move when only half are needed, but it's much simpler to wire.

This escalator just represents the tip of the iceberg. Frames are an incredibly flexible building block that can be used in a countless number of ways in building things that move.

And Frames are just one of the new things that arrived with prerelease 5. When you combine Frames with Control's emulated 6502 CPU, the liquid handling provided by pumps, and the parts added to Logic, the result is a set of components that will keep RedPower 2 players busy for a long time.

Edit: Thanks to Eloraam and others on the Minecraft Forums who answered questions, and to direwolf20 for his YouTube tutorial showing the basics of frames.